Two reports about the use of social media by the Russians to interfere with the 2016 elections were released yesterday. The reports give us greater insight into how social media platforms were used to influence the election in favor of Donald Trump. Much of this isn’t new to most readers, however the details prove that the future of elections is in social media.
The first report, commissioned by the United States Senate Select Committee on Intelligence, was produced by New Knowledge. New Knowledge used data provided to it by the Senate, who received it from Facebook, Twitter and Alphabet, Google’s parent company. The posts reviewed by the authors of the report were created by the Internet Research Agency, a Russian-connected company currently under indictment by the United States government for illegally interfering in the elections.
The focus of the report were activities between 2014 and 2017 that targeted American voters in order to influence them politically.
According to the New Knowledge report, besides the usual social media platforms, other channels were abused by IRA. Among them were Gmail, Google Voice to create “false personas” and other social media companies like Vine, Live Journal, Pinterest, Reddit and others. Pokémon Go, a game, was also used to influence American voters.
The “false personas” were needed in order to give the false appearance of trending social media posts by having fake users push the content through the various social channels. The more engagement on a given post, the higher the ranking it achieves within the individual platforms reaching larger audiences.
The algorithms expect organic user engagement to push the content to larger audiences. What the Russians achieved with the false personas and various accounts on different platforms is to provide their influence driving content the illusion of organic growth increasing the reach it has artificially.
Additionally, the Russian troll factory needed to give the illusion that the content they created had an acceptable source from where it came. In other words, if the content alluded to some nefarious event, it needed to give the illusion that it was not derived from an individual’s conspiracy theory, but rather from some respectable source. Thus, fake “respectable” source websites were created to give the illusion that the details came from mainstream America.
The IRA used about 1,000 individuals to create content for their influence project. The Russian-linked company reached 126 million Facebook users, about 20 million on Instagram and 1.4 million on Twitter. According to the federal indictment of IRA, it had a budget of $25 million.
As the Russian infiltrators gained experienced, they moved away from stale platforms, like Facebook and Twitter and increased their activities on Instagram as it outperformed the other platforms in engagement. Instagram’s value was in meme warfare, the use of images as memory triggers.
The report states that the IRA targeted Black-American communities to be used as “assets” for their operations. The report also accuses the Russian-linked company of running “voter suppression operations”. These were voting scams, such as text-to-vote or misleading voting rules. The report also alleges that the operation was designed to support the Trump candidacy while running a comprehensive anti-Hillary Clinton scheme.
Another takeaway from the New Knowledge report is that the major social media companies “misrepresented or evaded in some of their statements to Congress.” The reason is obvious, advertising and the number of active users were revenue generators for the companies. Whether they are organic users or Russian trolls does not matter because the more of each, the better the revenues are for the companies.
The other report that was released is from the University of Oxford.
This report also looked at IRA’s election disinformation campaign. According to the Oxford report, between 2015 and 2017, “over 30 million users…shared the IRA’s Facebook and Instagram posts with their friends and family.”
Assuming at least an additional user influenced by the IRA content shows that at minimum, 60 million individuals were influenced by misleading content created by IRA. The researches also noted an increase in advertising that coincided with important election dates. Oxford researches add that IRA’s activities “were designed to polarize” Americans by dissuading African American voters by urging them to boycott elections and for “Hispanic voters to distrust US institutions.”
Both reports make it clear that millions of American voters were manipulated into giving credence to fake news items designed to influence a positive view of Donald Trump and a negative one of his opponent, Hillary Clinton.
The obvious question remains, was it enough to distort the true electoral will of the American people.
The Computational Propaganda Research Project; “The IRA, Social Media and Political Polarization in the United States, 2012-2018” can be downloaded here.
The New Knowledge “The Tactics & Tropes of the Internet Research Agency” Report can be downloaded here.