social media propagandists aren’t changing minds, and that’s terrible news
A new study of social media from Duke University asked a group of 1,239 Twitter users about their political attitudes and social media habits one month apart in 2017, then checked if they were exposed to accounts outed as being ran by Russia’s social media propaganda arm, the Internet Research Agency. Their goals were simple. They wanted to determine if exposure to the propagandists’ accounts changed people’s minds about political topics and if they didn’t, what would. As it turns out, those propagandists were little more than just background noise and they failed to meaningfully change the test subjects’ views. But while this might seem like good news, the takeaway is actually pretty disturbing when we apply it to the broader context of political debates in the United States and Canada.
Now, the propagandists’ lack of effectiveness isn’t for lack of trying but because relatively few users engage with their content. Just 6% of the study’s respondents mentioned, liked, followed, or retweeted even a single piece of content out of the 288 million impressions trolls generated on Twitter. This tracks with Facebook’s findings that while 126 million Americans might have seen some piece of Russian propaganda on the platform, that content would’ve come up in their feeds once per every 23,000 posts. These drive-by propaganda appeals probably aren’t swaying minds of the relatively few people who see and engage with them, and people don’t change their political views just because a sketchy account shared a meme or a link.
While the Duke study notes that their sample wasn’t truly representative of the American electorate and didn’t track duration of the respondents’ engagement with propaganda, it should have turned up at least some sign that people changed their minds after being exposed to the trolls’ content if the notion that our politics are heavily warped by malicious outside forces was correct. So, what’s happening? Well, this is the disturbing part. The people who share and promote foreign propaganda designed to incite conflict and division already believed in and agreed with its message. It’s not that Russian trolls brainwashed grandma Mildred into bigoted and conspiratorial beliefs, it’s that she had them for years and agreed wholeheartedly while social media’s core algorithms sealed her in an apocalyptic echo chamber by her choice.
Really, a brainwashing operation would’ve been preferable because without it, we’re left with the sad and unnerving fact that American right-wing trolls, Canadian Wexit enthusiasts, and furious QAnon adherents need absolutely no outside help to loathe and despise their fellow citizens so much so that even the rare appearance of foreign propaganda trying to cast their family, friends, and neighbors in the worst possible light is simply a confirmation of their existing opinions. This isn’t hyperbole. Just consider that Trump voters explicitly ask the man they elected president to hurt their fellow Americans and write detailed fantasies in which they get the chance to murder liberals in another civil war. With attitudes like that, how much fuel will a foreigner’s regurgitation of stuff their targets already routinely post online add?
It’s very tempting to look at the scale of impressions and the number of posts and tweets by trolls working day and night to seed strife and disdain between the citizens of countries they consider adversaries or competitors, and point to them as the reason why politics across the Anglophone West is broken and is being given a nasty stress test across Europe. But when we put these figures alongside the billions of tweets, posts, and minutes of video watched every day on social media, these provocations are a drop in the bucket. What we’re seeing are signs that the world order that functioned so well in the previous century is failing us in this one, and we have to look at structural, systemic, far-ranging solutions to massive, globe-spanning issues instead of simply blaming the trolls. They trolls are not helping. But neither are we.
See: Bail, A., et. al., (2019) Assessing the Russian Internet Research Agency’s impact on the political attitudes and behaviors of American Twitter users in late 2017, PNAS, DOI: 10.1073/pnas.1906420116