Assessing the Impact of Twitter on the Outcome of the 2017 UK General Election

Abstract

Prior work on digital campaigning has found a significant and positive relationship between candidate website use and their vote share. How and why these effects occur and whether they continue through to the social media era, however, remain unanswered questions. This paper seeks to address this gap in the literature by examining the impact of candidates’ Twitter use in the 2017 UK General election on their level of electoral support. In particular, we look at the relationship between the pattern and content of candidates’ campaign tweets and their vote share, controlling for a range of rival factors. We design measures that capture key aspects of candidate tweet behaviour that we expect to increase their overall online popularity and thus potential offline support. These are the overall frequency and intensity of candidates’ Twitter use along with more nuanced indicators of the extent of interactivity and the issue emphasis of their tweets. Our expectation is that candidates who made frequent use of the medium and in a way that connected with voters’ concerns would be more popular with voters than those who did not. We test the expectation using a unique database of candidate tweets which we combine data from the Electoral Commission and the 2017 British Election Study (BSS). The paper is significant in that it updates and increases our understanding of the role of digital campaigning in elections and particularly whether it matters to the outcome. We also make an important methodological contribution to the field by developing and testing a new multi-dimensional measure of candidate Twitter use.

Keywords: Campaigns, General election, UK, Twitter, Candidates