Working Papers

Campaign Effects on Strategic Voting for Minor Parties: Evidence from Germany

This study uses a novel campaign effects model of strategic voting for minor parties in multiparty democracies that compares the relative importance of coalition insurance voting, compensatory voting, and a hybrid strategy. To do so, it uses data on the 2013 and 2017 German federal elections from campaign-period surveys, polls and an original dataset of the candidates’ tweets about policy issues. The results show evidence of policy-driven voters using a hybrid strategy in 2013 and a compensatory strategy in 2017. There is no evidence of coalition insurance voting in these elections. These findings reaffirm the view that the campaign plays a crucial role in the decision-making process of voters, in this case by allowing them to target coalitions based on policy. Depending on the case, 5% to 11% of the voters for the minor parties targeted by these strategies are estimated to be strategic.

What Role Does Affective Proximity Play in Canadian Elections? (with Edana Beauvais)

Voters’ feelings toward social and political groups can impact their political behaviour. How does group-based affect— voters’ relatively warmer or colder feelings toward different groups —impact their vote choice? The goal of our present work is twofold. Our frst task is descriptive: What does the “affective space” look like in Canada? To answer this question, we use data from the Canadian Election Study on feelings toward ethnic groups, sexual minorities, political parties, and national groups (e.g., Canadians, Quebeckers, and Americans). We use unsupervised machine learning techniques to reduce the dimensionality of affective space in Canada. Our second task is inferential: How does group-based affect impact vote choice? To answer this question, we regress vote choice on two, uncorrelated dimensions of group-based affect— ideological and ethnocultural —in seven different elections from 1993 to 2019 in Quebec and Canada outside Quebec. Our results suggest that voters’ group-based affect contributes to polarized pluralism in the party system (Johnston, 2017), but that fault lines are also emerging.