There has been a lot of discussion (and hand wringing) over what seem to be concerted Russian “influence operations” aimed at US politics. Our Russian friends appear to be trying to influence voter behavior, to exacerbate divisions within American society and sow distrust in our political system in general. Much of the content that has been identified as part of this campaign has been identified as originating from the blandly named “Internet Research Agency” in Russia.
The IRA has been hard at work – in 2016, they produced 57,000 tweets, 2,400 Facebook posts and 2,600 Instagram posts and these numbers have been on the rise ever since. But is their effort having a measurable effect on voter attitudes and behavior? Researchers at the Polarization Lab at Duke University have recently published a paper which tries to answer this question.
The Duke study was very limited in scope, limited to 1,239 US voters who identified as Democratic or Republican, used Twitter 3 times a week or more, and who were willing to provide the researchers access to their accounts. The tweets examined covered the one month period between October and November 2017. The findings, while limited in scope, do provide some reason for hope. Here are some of the salient points:
While approximately 2% of Twitter users interact with known IRA accounts, the study participants did so at a much higher rate – 19%.
The participants whose follows were limited to accounts which reinforce their stated opinions were more likely to have seen IRA posts. This is significant, as these are the people who are least likely to be influenced by political messages which don’t match their beliefs.
Surveys of participants before and after the study period showed no difference in political opinions after exposure to the IRA tweets or those of accounts which did not line up with the participants’ beliefs at the beginning of the study.
While this study was quite limited in time and venue (Twitter only) and excluded independent voters, it does offer some indication that Russian influence operations may not be as effective as the Kremlin would hope and that Americans are better at resisting propaganda than we think. I hope that there is further research in this area; we need to know just how serious a problem social media influence operations are.