Analyzing voter influence

(Photo by Avery Gales)

With the 2012 American presidential election coming to a close, it is somewhat relevant to examine what kind of factors influence voters in their decision making processes. Jason Roy, a political science professor at Wilfrid Laurier University, recently conducted a study that looked at several of these factors.

The study was conducted on 2,727 Laurier students in the form of an online platform survey, and there were three separate studies within the main study that the participants were randomly assigned to.

The first study looked at whether individuals changed support for the incumbent party based on whether election is called under a majority, minority or coalition situation.

“[For example], how would the incumbent vote share change if the other two parties voted against it and forced an election?” questioned Roy. “Would voters be more likely to support the incumbent party because they would see them as potentially the victim of the other parties? Or would the other parties gain support?”

The second study looked at whether there were any gender effects in voting.

This study posed the exact same candidates as the first study, but simply replaced one or more of the candidates’ names with female names. Previous research on this topic has suggested that females are more likely to vote for a female candidate.

“What we’ve discovered in a very, very early review of our data is that there is a similar effect. There are cases where the same candidate, when you shift from being a male to a female, receives almost a 20 point jump from female voters,” said Roy.

However, this effect was not clearly distinct in males voting for males.

The third study looked at whether voters were more influenced by their party affiliations or the individual candidate’s qualities regardless of their party.

To do this, four main Canadian party names were used and a preliminary survey gathered information on strength of party affiliations of the participants. Then each party was randomly assigned one of four candidates where three were similarly mediocre and one was a “super candidate” that was clearly more impressive in most aspects.

“If it’s the candidate [voters vote for], it shouldn’t matter regardless of where he runs,” explained Roy. “The initial analysis suggests that he boosts the vote share by about ten points regardless of party.”

Although initial analyses have been performed, Roy stated the results will still take four to six months to properly go through and analyze. But potential implications of the future findings remain intriguing for Roy.

This information could possibly be useful to individuals, political parties and policy makers.

“Depending what the results suggest, it can certainly be of interest to certain parties. Now, whether or not parties use the information to try to improve the system — that remains to be seen,” explained Roy. “And it could certainly have impact on policy makers. It could certainly influence how they want to change the existing system. Or whether they want to change the existing system.”

The first round of data for the study will be presented April 2013 at a conference in Chicago.

Leave a Reply