Candidates increasingly make use of social media in election campaigns. However, little is known about the effect this increased use of social media has. Candidates themselves seem to believe it makes a difference, but is this the case? This paper seeks to examine whether or not candidates that make use of social media in their personal campaign get a significantly higher number of (preference) votes. Specifically we investigated whether usage of Twitter and Hyves (a Dutch version of Facebook) had an impact on the number of preference votes of a candidate during the Dutch national 2010 elections.
Two effects can be thought of. First of all it may well be that social media have a direct effect: followers and friends of a candidate may be more likely to vote for a candidate because he or she cultivates a more direct and personal relationship with the candidate. Yet it may well be that only those people who follow a candidate already intended to vote for that candidate. But social media can also have an indirect effect. In this case social media usage will have a positive effect when journalists who follow a candidate actually write about a candidate’s messages and the message is thereby presented to a broader audience.
To carry out our analysis, we make use of a unique dataset that combines data on social media usage (number of follower and friends, number of messages, number of journalists following a candidate) and data on the candidates themselves (amongst others: position on the list, exposure to the old media, gender, age and incumbency). Our dataset includes information on all 493 candidates of the ten parties that received at least one seat in the 2010 election.
It turns out that there is a small but significant direct effect of social media usage. This effect remains even after we control for reverse causation and the degree to which a candidate is widely known prior to the election campaign.
