Facebook likes and public opinion: Predicting the 2015 Finnish parliamentary elections

Abstract

Predicting an electoral outcome using ‘big’ social media data is a new research theme that has emerged due to the exponential growth of social media. Mainstream research in this field focuses on Twitter and Facebook. This study examines the extent to which Facebook Likes can be used to predict electoral outcomes. The study was conducted in Finland, where 2146 candidates across the country competed for the 200 seats in the Finnish Parliament in 2015. The primary objective of this study was to analyze the degree to which Facebook Likes could be used to predict the outcome of the 2015 Finnish parliamentary election. A complete candidate-level dataset was compiled from Facebook and other sources to analyze the relationship between Facebook Likes and the vote shares of candidates. The initial sample contained 2.7million Facebook Likes, which were gathered directly from candidates' official Facebook pages. The prediction based on Facebook Likes was less accurate than using incumbency and traditional polling on whether a candidate would be elected. However, Facebook Likes and votes were found to have a significant positive relationship. Hence, social media data, like Facebook data, might be a significant but weak indicator of electoral success.