Across the globe, pollsters’ election predictions consistently underestimate the electoral gains of the far right. At first glance, the problem may appear trivial. After all, it is easier to predict the election results for larger, established parties than smaller, more volatile fringe parties. However, even as the far right has grown tremendously in recent election cycles across Europe, pollsters have continued to underestimate their performance—sometimes by as many as 11 percentage points. FiveThirtyEight’s Nate Silver and other statisticians have since called for the adoption of newer polling technologies to ensure that conservative voters are adequately represented in polling samples.
This dissertation chapter posits that the underestimation of far right electoral performance is not a function of sampling bias, but instead a psychological feature. The objective of this project is to understand the circumstances under which a voter’s preference for the far right is falsified, and then pin down the mechanism by which far right identification becomes ‘legitimized’ and perceived as socially and politically acceptable. This study can help shed light on the mechanisms (e.g., anonymous and pseudo-anonymous legitimizing Internet platforms) by which the far right agenda becomes perceived as legitimate.
The project utilizes Facebook’s ad platform to recruit approximately 3,000 European subjects from Hungary, Germany, and France to participate in a 10-minute survey experiment. Priming experiments are used to determine whether there is a significant difference between subjects’ willingness to identify with the far right when such parties are primed as politically legitimate versus when they are primed as politically illegitimate.
Subjects from these European countries are expected to identify with the far right at significantly higher rates when such identification is framed as politically legitimate than when it is not framed as such. In the Hungarian sample (n=1,006), rates of far right identification increased by almost 5% under a randomly assigned ‘legitimizing’ frame. People who self-identify as ‘centrists’ are the least likely to identify as far-right sympathists under politically ‘illegitimate’ framing (when compared to politically ‘legitimate’ framing).
List experiment is also used to examine whether anonymity matters to rates of far right political identification—and for whom anonymity matters. In the Hungarian sample (n=1,006), women under 35 years of age were most likely to identify with far right sentiment ‘anonymously’ but not ‘publicly’. These results are statistically significant at p=.01 level. Younger generations, in general, are more likely to falsify their support for the far right.