Investigating the Role of Social Bots During the 2019 Canadian Election
19 Pages Posted: 6 Mar 2020
Date Written: March 2, 2020
Threats of social media manipulation during elections have become a central concern for modern democracies. The 2016 US election raised awareness of the problem and triggered a vigorous response from public agencies and social media companies around the globe. We contribute new evidence about the phenomenon by investigating the behavior of suspected social bots during the 2019 Canadian election, using a collection of 19.3 million messages posted by 1.8 million unique users on the Twitter platform. We ask three questions of interest. Have social bots influenced public sentiment toward party leaders during the campaign? Do social bots post content that differs substantively from the messages posted by regular, "human" users? And finally, is there evidence of foreign interference during the 2019 Canadian election? We find that social bots were disproportionately hostile to the incumbent prime minister Justin Trudeau. Our evidence suggests that social bots attempted to prolong the 'blackface' controversy afflicting the Trudeau campaign early on, but ultimately had no influence on public sentiment toward party leaders. While we detect clusters of social bots that are most likely foreign in origin, the general pattern appears more consistent with the idea that national partisan groups were the ones primarily making use of social bots for political influence.
Keywords: 2019 Canadian Election; Social Bots; Political Communication; Foreign Interference; False News; Justin Trudeau; Social User Embeddings; Community Detection; Twitter
Suggested Citation: Suggested Citation