By Steffen Zittlau, PhD candidate, University of Mannheim
What is the story?
The results of last Sunday’s European election in Germany are not surprising. For the most of it, they are very similar to those of the German federal election held last September (see our previous blog entry): The party of the federal chancellor Angela Merkel, the Christian Democrats (CDU/CSU), won by a comfortable margin over the Social Democrats (SPD).
In the German media, a specific attention was given to the right-wing Eurosceptic party Alternative für Deutschland (AfD). The party failed, by a very small margin, to pass the 5% representation threshold in the last federal election. In this European election, the party obtained 7% of the votes and 7 seats in the European parliament.
To explain the good score of the AfD, political commentators in Germany pointed the role of low turnout. According to them, this low turnout benefited far-right and Eurosceptic parties in all European countries. During the campaign, the mainstream parties also used this claim to encourage their supporters to vote. This hypothesis constitutes now the main explanation of the electoral success of the AfD (and other similar parties in other European countries). Is it really the case though? In this blog entry, I offer a first test of this hypothesis.
A first test of the ‘low turnout benefited the AfD’ hypothesis
The rationale behind the hypothesis positing that the AfD benefited from a low turnout rate is that the supporters of the party are more likely to vote at a European election than the supporters of other parties. The AfD was founded in 2013 in reaction to the decision of the European Union to bail out some member states (Greece, Ireland, Portugal, Spain and Cyprus) to save the European monetary system. The party strongly criticises this policy, and in general the functioning of the European institutions. It is then reasonable to think that its supporters consider European issues as highly salient. Therefore, as the European election is (allegedly) more about European issues than any other election, supporters of the party are supposed to be more likely to turn out at this election than the supporters of other parties.
To test this hypothesis, I rely on official data of last Sunday’s election and check whether the score of the AfD was greater in regions where turnout was low.
First, we observe substantial variations in turnout between regions, ranging from 41% in Bavaria to 57% in Rhineland-Palatinate. This large variation is partially explained by the fact that municipal elections were also held last Sunday in 10 out of 16 German regions. On average, the turnout rate was 5 percentage points higher in regions where municipal elections were held simultaneously.
Before running the test, I correct the data by taking into account the fact that turnout is usually higher, and some parties traditionally stronger, in some regions. I use the last federal election as the baseline, and calculate the ratio between turnout at the European election and turnout at the last federal election. The same correction is used for the AfD’s vote share. For example, a ratio of 1.8 indicates that the AfD’s vote share was 80% higher at the European election than in the federal election.
Figure 1: The Relationship between Turnout and AfD’s Vote Share (click to enlarge)
Figure 1 clearly shows that AfD’s vote share at the European election was much greater in regions where turnout was low. If we were to believe the predictions of a simple OLS model, the score of the party would have been roughly equal to its score at the federal election if turnout had been the same (i.e. at a turnout ratio of 1). Bavaria deviates from this finding though: The AfD was able to nearly double its vote share compared to the last national election. The low turnout only partially explained the score of the AfD in the region.
Concluding remarks
The findings of my first test of the hypothesis positing that the AfD benefited from a low turnout rate suggest that specific parties might benefit from low turnout, especially if they own the issue related to the election of the day. This gives them the possibility of mobilizing their supporters more than other parties. My analyses also show that holding other elections simultaneously does not only influence the turnout rate, it can also alter the result of an election.
