In 2016, voters in multiple countries around the world were targeted on social media to a degree never seen before. Russian-sponsored fake accounts, misleading social posts and illegally acquired profile data were all deployed to manipulate votes and pull off two of the most unexpected results in decades: Donald Trump elected president of the United States and Britain voting to leave the European Union. The question for voters today: Should we expect anything different in the 2020 elections?
To find out, I’ll be looking at the role social media played in the 2016 elections, and the resulting fallout, before casting an eye ahead to identify what role social media will play in the 2020 US elections.
Part 1: What Did We Learn About Voter Manipulation in 2016?
Part 2: What Changes Have Social Platforms Made Since 2016?
Part 3: What Should We Expect in Upcoming Elections?
Part 1: What Did We Learn About Voter Manipulation in 2016?
When it comes to discussing the impact of a campaign of misinformation on social media, we have to stick to the facts available.
Much of what we know about the role of the Russian Internet Research Agency (IRA) comes from two reports written for the U.S. Senate and released in on December 17, 2018. “The IRA and Political Polarization in the United States” by Oxford University’s Computational Propaganda Project and analytics firm Graphika and “The Tactics & Tropes of the Internet Research Agency” by New Knowledge both document the numerous attempts made by the Russian-backed Internet Research Agency to influence the result of the 2016 election via social media.
In an orchestrated campaign that relied on thousands of real Russia-based users, content was posted on virtually every major social medium, including Facebook, Twitter, YouTube, Instagram, Reddit and Pinterest.
The IRA used the targeting features of each social media platform to send highly personalized messages to specific demographics. One target was Black Americans, who were urged to boycott the election. Another was right-wing conservatives, who were targeted with patriotic, anti-immigration messages and pushed to vote for Trump.
Social Media Was Definitely Used to Influence the 2016 US Elections
But the IRA wasn’t the influential organization on social media during the 2016 election in the US. For their campaigns, Ted Cruz and then Donald Trump hired the services of Cambridge Analytica, a UK-based data analytics company that specialized in political campaigns.
Research by The New York Times and reported by Matthew Rosenberg, Nicholas Confessore and Carole Cadwalladr found the firm carried out a series of services using illegally harvested private information from more than 50 million Facebook profiles that it had collected since 2014.
“Under the guidance of Brad Parscale, Mr. Trump’s digital director in 2016 and now the campaign manager for his 2020 re-election effort, Cambridge performed a variety of services, former campaign officials said,” the NYT reporters write. “That included designing target audiences for digital ads and fund-raising appeals, modeling voter turnout, buying $5 million in television ads and determining where Mr. Trump should travel to best drum up support.”
Voters had no idea that they were handing over so much data to Cambridge Analytica, nor did they knowingly agree for their data to be used to coerce them.
Presidential Voter Manipulation by the Numbers
The reach that agents were able to generate in 2016 using social media alone is staggering. Below are some of the most prominent statistics cited by research.
- More than 30 million people shared IRA content on Facebook and Instagram between 2015 and 2017, the report from Oxford University notes.
- There were significantly more engagements, such as likes or shares, with Russian-created content on Instagram (187 million) compared to Facebook (76.5 million), according to the New Knowledge report.
- The New Knowledge report found that the IRA reached “126 million people on Facebook, at least 20 million users on Instagram, 1.4 million users on Twitter, and uploaded over 1,000 videos to YouTube.”
- Research by Levich Institute professors Alexandre Bovet and Hernán A. Makse found one-quarter of all tweets in the five months preceding the US election were fake or very misleading.
Far from being a mere mouthpiece, the IRA managed to set the political agenda and fuel the flames of digital debate to the extent that regular American citizens shared their messages en masse.
Why Was Voter Manipulation So Widespread?
The misinformation campaigns were so effective because Facebook’s algorithm is programmed to serve up more of the content that a user engages with most.
As Social Media Today’s Andrew Hutchinson points out: “The problem with Facebook is that what works for increasing user engagement doesn’t necessarily work for ensuring balanced coverage. Facebook wants users to stick around, to spend as much time on-platform as possible, and a key way of doing that is by showing them more of what they want to see.”
Social media “worked just as advertised,” tech reporter Kurt Wagner writes — stolen identities and VPNs aside.
“But beyond that, their election interference campaign was a textbook example of how a legitimate U.S. business or publisher would create and dispense its own social media campaign. It all looked rather easy — and familiar — when you read through the indictment. The IRA employed a lot of the same strategies a new-age media company or savvy brand marketer might employ.”
The IRA didn’t disrupt the technology or hack the social media platforms, Wagner notes. They used them just as any legitimate business would.
At this point, we must note that it is impossible to quantify the impact these campaigns had on the actual election result, widespread as they were. Reason editor Nick Gillespie believes that both Senate reports fail to demonstrate convincingly what impact Russia’s meddling had on election results. “That omission—and a corresponding lack of interest in putting such efforts in a historical or contemporary context—means the studies provide little to no actual insight into electoral politics on the ground or online.”
While the social media campaigns were largely used to reinforce a negative perception of Hillary Clinton, Gillespie argues that Clinton also ran a poor race. Black voter turnout in 2016 was as low as it was for John Kerry in 2004, for instance, and Gillespie believes this was more a result of Clinton’s poor campaign strategy than it was a result of Russian influence.
Brexit and Other Campaigns Laid the Groundwork for November 2016
While the US election has taken center stage in the voter manipulation and social media interference scandal, it wasn’t the first time these tactics were used to influence a vote. Both Cambridge Analytica and Russian trolls were involved in the UK’s Brexit referendum in June, 2016.
AggregateIQ, a Canadian political consulting group and Cambridge Analytica contractor, was hired by several pro-Leave groups, including the official Vote Leave campaign, to run a series of targeted Facebook advertising campaigns. These campaigns have been cited as vital to the election result by several of the movement’s leaders.
Evidence unearthed by Chris Vickery, director of cyber risk research at UpGuard, suggests that AggregateIQ developed the election software that Cambridge Analytica sold to clients during the presidential election.
The services of Cambridge Analytica were also used by another pro-leave campaign, Leave.EU. Russian trolls were active throughout the campaign, but particularly so on the day of the vote. According to data released by Twitter, Russian agents were responsible for more than 1,000 pro-leave Tweets that included the hashtag #ReasonsToLeaveEU on June 23.
Britain and the US weren’t alone in suffering voter manipulation, according to a report by Freedom House. “Manipulation and disinformation tactics played an important role in elections in at least 17 other countries over the past year, damaging citizens’ ability to choose their leaders based on factual news and authentic debate.”
The 2016 election wasn’t even the first time social media had been used to impact the vote in the US. Facebook held sway as early as 2012, according to Atlantic reporter Rebecca Rosen. Rosen cites a study led by USC San Diego professor James Fowler that suggests Facebook played a key role in mobilizing the youth vote in the 2012 election. The study asserts that Election Day advertising on the social media platform — which included a notice that it was Election Day, directions to the nearest polling station and buttons to signify that you had voted or were going to do so — led to an increase of 340,000 votes.
Rosen points out that it is likely a large portion of these votes went to the Democrats, as Facebook skewed towards younger voters and women in 2012, two groups who preferred Obama over opponent Mitt Romney.
Nor did Russian meddling cease after the election. In fact, the IRA became even more productive. The report from Oxford University found that post-election activity increased by 238 percent on Instagram, 52 percent on Twitter and 59 percent on Facebook.
Part 2: What Changes Have Social Platforms Made Since 2016?
In the aftermath of the revelations outlined in the previous post, every major social media platform implemented a number of measures aimed at improving their service and winning back the trust of customers.
Facebook, Twitter and YouTube all endorsed the Honest Ads Act, which requires US and European political ads to declare who is behind them. All three platforms also deleted a significant number of accounts. Twitter has claimed that it is deleting 1 million accounts daily. Facebook removed almost 3 billion accounts over the course of a year.
Not all of these accounts were fake, however. On October 11, 2018, for instance, very real American media members were removed from Facebook and later from Twitter. Rolling Stone’s Matt Taibbi notes that these came from both sides of the political divide. “Some, like the Trump-supporting Nation in Distress, had claimed Obama would declare martial law if Trump won in 2016. Others, like Reverb and Blue State Daily, were straight-up, Democrat-talking-point sites that ripped Trump and cheered the blues.”
Both Facebook and Twitter have also implemented updates to their algorithms aimed at limiting the impact of false, malicious or harmful content.
On January 11, 2018, Mark Zuckerberg announced that Facebook had updated its algorithm to focus on “meaningful social interactions” between friends and family. It would also promote news from trusted and local sources over other publications. In May, 2018, Twitter updated its algorithm to hide bad tweets and troll messages.
New York Magazine’s Brian Feldman views Twitter’s update not so much as a censor, but as a way to make sure people stay on the site for as long as possible. “It’s a strategy intended to maximize the amount of time users spend on the site by ensuring they are less likely to see content that upsets them or turns them off. It’s focused on behavior, not on party identification.”
Both social platforms have also invested in AI technology to identify and remove content that is considered fake, inflammatory or harmful. YouTube has also taken conspicuous steps to remove hateful content. In June 2019, the site announced it had updated its hate speech policy to include extremist content, which it said it would be policing. At the time of writing, it’s not clear how YouTube will be doing this.
But Are Their Efforts Genuine?
Despite the actions that the major social media mentioned above have taken, many people still question their motives.
It’s not hard to understand why based on a New York Times report that demonstrates how Facebook allowed companies to access user data after the elections without users’ knowledge.
The report, by Gabriel J.X. Dance, Michael LaForgia and Nicholas Confessore, details how the social media giant allowed companies such as Amazon, Spotify and Microsoft to access users’ personal data and even read their personal messages. Facebook allowed this to such an extent that those businesses were effectively exempt from the site’s own standard privacy rules.
Nor is Facebook taking responsibility for its breaches. In fact, the company is fighting hard to stop it from ever being held responsible by consumers, writes Neema Singh Guliani, a senior legislative counsel at the American Civil Liberties Union.
In response to a lawsuit filed by consumers in Illinois claiming that the company broke privacy laws by using facial recognition software without consent, Facebook is arguing that the law doesn’t grant consumers the right to sue. Facebook argues that the law states consumers have to show additional harm was caused in order to sue. This, Guliani argues, means that suing Facebook wouldn’t be possible even if the company did overtly break the law.
Twitter is not without controversy post-election, either. In 2019, Twitter started inserting popular Tweets from unfollowed accounts into users’ feeds. While this is largely innocuous, it is having unintended and potentially damaging consequences.
Some of these tweets have been shown to be politically motivated and come from people with fringe views, many of whom are verified by the website. “In effect, the practice means Twitter may at times end up amplifying inflammatory political rhetoric, misinformation, conspiracy theories, and flat out lies to its users,” CNN’s Oliver Darcy writes.
YouTube’s recent attempt to crack down on extremist speech has also been met with skepticism. After all, hateful rhetoric has enjoyed a home for years on YouTube. “The platforms have become very good at issuing PR statements about proposed changes that don’t ultimately have much effect,” Stanford tech and media researcher Becca Lewis says. “Any change in the right direction is good, but the platforms keep making promises they don’t ultimately keep.”
States and Governments Take Action Into Their Own Hands
Around the world, governments have been passing laws that put increasing responsibility on social media companies.
On January 1, 2018, Germany enforced the NetzDG law, which promises fines of up to 50 million euros if companies fail to remove “obviously illegal” hate speech within 24 hours of notification. As a result of the law, both Facebook and Twitter have added additional features to the German versions of their platforms and trained additional moderators.
In the UK, the government has outlined a duty of care for social media companies, which will leave them facing huge fines or being blocked completely if they fail to remove harmful, illegal or misinformative content from their platforms.
Australia has also passed a law that holds social media companies responsible for harmful and illegal content posted to their platforms.
Significant action has been taken in the United States, too. On April 10, 2018, Mark Zuckerberg appeared before a Senate hearing to answer questions relating to Facebook’s data-sharing scandal. In a five-hour session, Zuckerberg faced a barrage of questions from the Senate commerce and judiciary committees.
But this hearing was nothing more than a sham, according to attorney, Fordham professor and political activist Zephyr Teachout. “It was designed to fail. It was a show designed to get Zuckerberg off the hook after only a few hours in Washington DC. It was a show that gave the pretense of a hearing without a real hearing. It was designed to deflect and confuse.
“Each senator was given less than five minutes for questions. That meant that there was no room for follow-ups, no chance for big discoveries and many frustratingly half-developed ideas.”
No nationwide law has yet to emerge in the US, but several states have taken action, reports Wired’s Issie Lapowsky: ”Vermont implemented a new law that requires data brokers which buy and sell data from third parties to register with the state. In California, a law is set to go into effect in January  that would, among other things, give residents the ability to opt out of having their data sold.”
In 2018, the U.S. Federal Trade Commission launched an investigation into Facebook to decide whether the Cambridge Analytica scandal meant the platform violated a 2011 agreement. At the time of writing, it is expected that the case will be settled for between $3 billion and $5 billion.
How Could These Changes Impact 2020?
When the 2020 elections roll around, we may find that the changes implemented so far don’t go far enough. In fact, they may already be outdated.
John Hultquist, director of intelligence analysis at FireEye Inc, notes that trolls are already changing tactic to sidestep rules introduced by the social media platforms. Rather than creating their own content, they are amplifying content that’s already out there, he says. “Then it’s not necessarily inauthentic, and that creates an opportunity for them to hide behind somebody else.”
The seeds of doubt from the 2016 race have already been sown. Mistrust is already rife among many black activists, who fear their concerns will be attacked or dismissed by other activists claiming they are fake, writes Buzzfeed’s Ryan C. Brooks. “Activists trying to spread information or opinions on Twitter about specific candidates have been shouted down by other activists who say the hashtags they’re using are overrun by bots, Russians, or trolls — a fight made only more complicated by the actual presence of bad actors across the platform.”
It won’t just be fake users writing fake messages in 2020, either, according to John Villasenor, a nonresident senior fellow at Brookings. Fake videos can now be created that look hyper-realistic. “With continued advances in artificial intelligence-based techniques for performing detailed frame-by-frame editing, it’s easier than ever to create highly convincing depictions of events that never actually occurred,” Villasenor writes.
“With the 2020 election campaign heating up, deepfakes are likely to be part of the landscape, spurring discussions regarding their role in elections and what can be done to minimize their impact.”
Part 3: What Should We Expect in Upcoming Elections?
The presidential election isn’t the only vote happening in 2020. There will be at least 34 seats up for election in the Senate, all seats in the House of Representatives, and the governorship of 11 states and two US territories. The vast majority of states will hold legislative votes in 2020, too, and there will be mayoral elections in a huge number of cities.
It seems all of these elections will be under threat of manipulation via social media. So much so that Michael A. McFaul, Ph.D., in a report from the Stanford Cyber Policy Center, suggests a paramount need to establish international relations that discourage meddling. His report outlines seven recommendations for governments to do so, including fortifying international commitments to human rights, strengthening the international norms that protect elections and leading advocacy against disinformation.
The report also recommends that “timely, tailored, consistent and credible costs” be imposed on Russia by the US government to punish them for their actions in 2016 and to deter them, or other governments, from attempting something in 2020.
Current Vulnerabilities and Risks
The Internet Research Agency has already proved that it is more than willing to peddle influence in non-national elections. One of the first actions ever taken by the IRA was to influence local elections in St. Petersburg. Agents from the IRA went door to door posing as journalists asking residents which candidate they supported. The information collected was then used to work out which candidates would need their vote count to be corrected to ensure that the ruling United Russia party would win the election.
Social media is already being used to tamper with local elections on US soil. In a piece for the The New Yorker, Adam Entous and Ronan Farrow cover how an election for the hospital board in the California town of Tulare was the target of a disinformation campaign.
In a bid to ensure an incumbent board member was re-elected, the hospital’s head physician hired an Israeli private intelligence company called Psy-Group. The company, which had the slogan “Shape Reality,” began creating a series of websites — including Tularespeaks.com, Tulareleaks.com, and Draintulareswamp.com — to smear the other candidate, Senovia Gutiérrez, and her allies.
“Along with the Web sites, online personae, who claimed to be local residents but whom nobody in town recognized, began posting comments on social media,” Entous and Farrow report. “Some of the messages suggested that Senovia took bribes. Others pointed to her Mexican background and her accent and questioned whether she was an American citizen.”
The efforts were ultimately unsuccessful. Gutiérrez won 75 percent of the vote, but if elections as small as a hospital board are at risk of social media-based manipulation, any election is.
U.S. House Speaker Nancy Pelosi recently became a target of trolls when a video of her was doctored to give the impression that she had been drinking. The video received more than two million views on Facebook, which refused to take down the video.
Pelosi commented that this was tantamount to Facebook enabling Russian influence. “We have said all along, poor Facebook, they were unwittingly exploited by the Russians,” she said. “They have proven, by not taking down something they know is false, that they were willing enablers of the Russian interference in our election.”
Social Media Is Changing Campaign Funding
For more than a decade, social media has been changing the way election campaigns were funded.
In a panel with political consultant Joe Trippi back in 2008, Arianna Huffington said: “Were it not for the Internet, Barack Obama would not be president. Were it not for the Internet, Barack Obama would not have been the nominee.”
Trippi added that Obama’s use of digital media allowed his campaign to raise small amounts of money from hundreds of thousands of people, without the need for armies of volunteers. He was also able to leverage YouTube, which politicians (and marketers) at the time were just learning how powerful it is to be on a video platform where people choose what they want to see. Political ads on TV, by contrast, disrupt the user’s viewing experience, and understanding this difference a decade ago was crucial for the Obama campaign.
Subsequent campaigns have used social media to raise significant amounts of money in very short timeframes, notes political journalist Tom Murse. “Money bombs are typically 24-hour periods in which candidates press their supporters to donate money. They use social media such as Twitter and Facebook to get the word out and often tie these money bombs to specific controversies that emerge during campaigns.” In 2008, presidential candidate Ron Paul was particularly successful with this tactic.
Washington Monthly’s Nancy LeTourneau believes the old model of campaign funding, where huge resources are spent on television ads, is dying. In its place are social platforms that allow campaigns to connect with voters at a huge scale and very low cost. “The more personal nature of those connections can help fuel small contributions from a larger pool of donors with a click on a computer screen.”
More Money Than Ever Will Be Spent on Social Media
Ironically, despite the concern over the use of social media in 2020, or perhaps because of it, next year’s presidential elections will see more money spent on social media advertising than ever before.
Democratic candidates and President Trump are pouring money into Facebook ads, writes Business Insider’s Joe Perticone.
“Already, the Trump 2020 reelection campaign is going all-in on Facebook, spending more than $4 million since the start of 2019. Since May 2018, the Trump campaign has spent nearly $12 million on Facebook advertising.”
Elizabeth Warren, Bernie Sanders and Kamala Harris have all spent more than $1 million on social media advertising since January 2019. Joe Biden has spent a combined $700,000 on online advertising, as has Andrew Yang, Perticone continues.
It’s not just the candidates themselves who are investing heavily in social media advertising. One of the biggest Democratic super PACs, Priorities USA, is bringing digital ad buying in-house.
The idea isn’t just to save money, writes HuffPost reporter Kevin Robillard. It will make the super PAC more agile and efficient, too. Should Priorities USA need to target a new or different audience with a particular message during the campaign, it will be able to do so far more quickly than if strategy changes had to be implemented by a third-party agency.
In the end, it looks like the real winners from attempts to combat misinformation on social media won’t be the electorate. It will be the social platforms themselves.