Russian trolls cost Hillary Clinton the election

Special counsel Robert Mueller'sindictment of 13 Russians for meddling in the 2016 US elections has led Democrats to revisit a theory of how disinformation could have flipped the result from Hillary Clinton to Donald Trump, by changing fewer than 80,000 votes across three states.

Yesterday (Feb. 16), president Obama's former senior adviser David Axelrod tweeted about Green party nominee Jill Stein, the most prominent third-party candidate in the 2016 race:

Jill Stein received 50,700 votes in MI. Trump won by less than 12,000. She outpolled Trump’s margin in WI, as well.
If Russians sought to redirect alienated HRC voters to third parties, as indictment alleges, it was a shrewd ploy.https://t.co/1igEg7PV4S

— David Axelrod (@davidaxelrod)February 16, 2018

James Fallows, a correspondent at The Atlantic and former speechwriter for Jimmy Carter, mooted the same theory, as did Obama's former Ambassador to Russia, Michael McFaul. Does it make sense?

The scenario Axelrod is referring to goes as follows: While Clinton won the popular vote by around 3 million ballots, she still needed less than a 1% swing in each of Michigan, Pennsylvania, and Wisconsin to win the electoral college. The Russians targeted voters in swing states, and it’s not impossible they tipped the balance if their engagement led voters to stay home or vote for third party candidates like Stein or Libertarian Party nominee Gary Johnson.

The Russian propaganda campaign reached 126 million people (paywall) through Facebook alone. Mueller's indictment says the trolls targeted swing states (though the indictment says they focused particularly on Florida, Colorado, and Virginia) and Michigan was one of nine states senior figures visited on an intelligence gathering trip in June 2014.

In late 2016, the trolls began targeting ethnic minorities in an effort to persuade them not to vote, or to back Stein instead of Hillary Clinton, according to the indictment. It cites the following posts on accounts controlled by the troll farm as examples:

"[A] particular hype and hatred for Trump is misleading the people and forcing Blacks to vote Killary. We cannot resort to the lesser of two devils. Then we'd surely be better off without voting AT ALL."

— Post on "Woke Blacks" Instagram account, October 2016

"Choose peace and vote for Jill Stein. Trust me, it's not a wasted vote."

— Post on "Blacktivist" Instagram account, November 2016. (The trolls paid Instagram for the post to be promoted for further reach.)

"American Muslims [are] boycotting elections today, most of the American Muslim voters refuse to vote for Hillary Clinton because she wants to continue the war on Muslims in the middle east and voted yes for invading Iraq."

— One of several similar messages posted on "United Muslims of America" social media accounts in early November 2016.

What effect did these posts have on actual voters? We don't know. The indictment says the trolls managed to persuade real people to attend rallies about all kinds of issues in various states, including Pennsylvania, by promoting them on their accounts, with paid ads, and by messaging individuals to encourage them to attend.

They even got one US citizen to send the troll farm's oligarch funder a happy birthday message from outside the White House.

This shows the digital operatives were able to influence some American citizens.  On the other hand, political scientists aren't even convinced the data shows Stein siphoning enough votes from Clinton to be blamed for her loss. We don't have anything beyond anecdotal evidence to determine if the Russian campaign actually changed people's minds, or whether their messages simply appealed to people who already held pro-Trump or anti-Clinton views.