An article by Sanne Kruikemeier from Amsterdam School of Communication Research ASCoR that has been published on Elsevier Journal 2014. View Article

Goal

The paper aims to find out how much using a social platform; in this case, Twitter would affect a politician preferential votes. Their presence on Twitter includes interacting with their potential voters and sharing their private persona. Further, it investigates the effect of candidates’ personalized communication on the number of preferential votes they receive.

Data

They have collected all political tweets by candidates in the period leading up to the Dutch national elections on June 9, 2010, and 2 months after. Aggregated data contained 40,957 tweets, posted by 177 politicians from 8 political parties. They have analyzed the collected dataset using both computer-assisted and manual analysis.

Results

• The number of tweets posted by political candidates increased during the election campaign and decreased after the Election Day.
• Interactivity and personalization are indeed key characteristics of online political communication.
• The significant effect of Twitter use by candidates will cause the number of votes they received.
• The more reciprocal interaction a political candidate uses in their communication on Twitter, the more preferential votes this political candidate will receive.
• No significant relationship between the amounts of retweets or hashtags used by the candidate and the number of preferential votes
• Being more popular on Twitter has positive effects on electoral support, although sending out more tweets does not lead to significantly more preferential votes.
• Using a more personalized style of online campaigning on Twitter had no impact on electoral support.
• using Twitter during the election campaign resulted in more votes, than not using Twitter

My Personal Takeaway and possible future expansions

• Twitter effects politicians’ electoral success, and they should interact with their voters to be popular.
• Twitter is an excellent source of a dataset. Especially on narrowed audiences such as politicians, musicians, or other famous people.
• Expanding the dataset timeframe from 3 months to the whole in-the-office period can induce more in-depth conclusions about politician online behavior.
• If this process becomes automated with a little flavor of machine learning and artificial intelligence, we could scrutinize politicians’ thoughts and actions more than ever.
• It can be a glimpse or hint at the future of governments that people tend to interact directly with the decisions politicians are making.