Opinion:

People in Norway, Sweden, and Denmark have high English-language proficiency, strong historic ties to the US, and, according to Jessica Yarin Robinson's research, they tweeted more about the American elections in 2016 than the Russian bots accused of trying to impact on election outcomes.

Ordinary Scandinavians tweeted more about the American 2016-elections than the Russian trolls accused of interference

OPINION: “Foreign interference” in US elections isn’t just for Russian trolls. Social media have made American politics accessible to ordinary users.

In the final weeks before the 2020 US presidential election, the following tweets were posted to Twitter from accounts outside the United States:

which countries will biden invade to prevent war? #NeverBiden

I'm not a huge trump supporter, but it's so clear that the MSM [mainstream media] treats Biden different than trump

We need a real leader. Don't let Trump win again

On election day, write in #BernieSanders*

These tweets were not sent by a Russian troll factory, however. They were sent by ordinary users from Norway, Sweden, and Denmark.

These are not the usual figures we think of when we talk about “foreign interference” in elections. Yet social media, and particularly the growing role of Twitter and Facebook in the American public sphere, have opened up U.S. elections to involvement by a wide range of actors.

These include agents of government-backed propaganda campaigns, but also ordinary foreign users.

The ability of foreign citizens to engage with American politics in the same spaces as voters is completely new, and data suggest this form of foreign engagement is likely more common than government-run “interference” campaigns.

By the numbers

The 2016 American presidential race put a spotlight on the use of the internet by foreign governments to try to influence the outcome of elections. One of the forms this took was the creation of fake social media accounts that would post messages intended to influence public opinion. In an investigation after the election,

Twitter found 3,841 accounts tied to Russia’s so-called Internet Research Agency, a “troll farm” aimed at influencing online discourse.

From a different perspective, I have researched Scandinavian Twitter users’ engagement with international politics, including American politics.

People in Norway, Sweden, and Denmark have high English-language proficiency, strong historic ties to the US, and according to Google search data, were highly interested in the race between Hillary Clinton and Donald Trump.

Drawing on a collection of tweets from fall 2016, I’ve used geographic and linguistic markers to identify Twitter users in these countries. This has turned up 67,551 Scandinavian users who sent 350,562 tweets about the election.

By way of comparison, 8,116 tweets by accounts from the Russian Internet Research Agency appear in the same data.

Like the Russian accounts, Scandinavian users encouraged voters to choose one candidate or the other, or a third-party candidate. They used the #MAGA and #ImWithHer hashtags.

Sometimes they tweeted about the election in Scandinavian languages; more often in English.

They retweeted news — and also fake news. “Hillary AND Bernie likely both had secret ties to the Islamic state” declared one Swedish user to nearly 190,000 followers.

And the Scandinavian users weren’t just tweeting into the ether – they engaged directly with American users just like the Russian “trolls.”

Trump in the global village

This pattern of engagement with American politics has continued over the last four years in the networks I study, perhaps thanks in part to President Trump’s own prolific use of Twitter. @realDonaldTrump is consistently one of the most mentioned users – in English and in Scandinavian languages – even during the covid pandemic. We’ve seen hashtags like #MeToo and #BlackLivesMatter take on global resonance. In a sample taken over the summer, almost a quarter of Scandinavian users tweeted about the George Floyd protests.

Rather than seeing this through the lens of foreign interference, this phenomenon might be better understood as the logical reaction of citizens confronted with globalization in the age of digital media. Since the 1990s, media scholars have argued online communication would enable people to engage more directly with events they see as affecting them, regardless of origin.

In interviews with Twitter users, I have found that while some hope to have some sort of impact on Americans, more often, they see a confluence between US politics and their country’s politics.

“The election of Donald Trump was a political wake-up call for me,” said one Swedish user. She says she saw elements of Trump in Sweden, and took part in political organizing for the first time in the 2018 Swedish parliamentary election.

Donald Trump’s brand of politics has also helped connect users on the right, with Trump-supporters – a small minority in Northern Europe – forming particularly active networks with likeminded Americans.

Stopping foreign interference

Communication researchers learned long ago not to mistake intended effect with actual effect. It’s hard to say what effect ordinary foreign users might have on American elections. So far there’s little evidence that organized foreign propaganda campaigns have actually changed electoral outcomes.

Even so, in advance of the 2020 American presidential election, platforms facing political pressure have taken measures to curb manipulative activities associated with foreign interference. Facebook, Twitter, and YouTube have become more reactive to disinformation, impersonation accounts, and automated posting aimed at manipulating algorithms.

Notably, however, these steps seem to be largely agnostic toward where activity is coming from. This is appropriate. The fact is, some form of foreign interference is now part of the public sphere as it exists online, and trying to draw a line between citizen and foreigner would undermine one of the key democratic promises of the internet.

*Note: The wording of all tweets has been altered to protect user privacy in accordance research data standards.

Powered by Labrador CMS