The effect of voting advice applications on Estonian voters' voting behaviour

Abstract

In the age of digitalisation and information overflow, it might be difficult for people to decide which information to consume and trust. However, next to all of that, we still have to make important decisions and one of them taking part in elections. In recent years, Voting Advice Applications or VAAs, as they are usually abbreviated, have become increasingly popular among voters in many countries. It is not an exaggeration to say that they are now the real part of how a voter makes his or her voting decision. The typical reason for this tends to be the fact that using VAA is rational for a voter: s/he gets the information from one place, and what is even more important, based on his or her opinions on certain matters, the program matches and also compares voter’s views to the ones of political party or candidate. Yet, quite little is known about the impact of VAAs. The aim of my master thesis is to contribute to the investigation of VAAs and have a look at their impact on Estonian voters’ voting behaviour. To do so, I will be using two datasets: panel data from European Parliament elections in 2014 and survey data on Estonian national elections in 2015. It is worth noting that in Estonia’s case panel data has not studied before in order to see the possible effect of VAAs on voting behaviour. I will pose three questions in my thesis: how do VAAs influence voting turnout, how do they change voter’s choice set, and finally, if VAAs have any effect on final vote choice. From statistical point of view, the results of my analysis are mostly insignificant. Leaving this aside, I found that based on those two datasets VAAs act more like a control mechanism: those people who already have decided to go to vote also tend to use VAA, also there is mostly no change in their choice set. One reason for this can be found from so called bottleneck theory – people who are more exposed to information, internet usage and different possibilities are therefore also less affected by VAA as it is one competing information source. From the other hand, those people, who are more likely to be affected by VAA, are not exposed to it and therefore I could not see effect of it on voting behaviour. In terms of final vote choice, I noticed that those people who used VAA were also more likely to change their final vote choice. However, as it was statistically insignificant, I cannot say that it translates to the whole population.