Submitted


  1. A mixed-utility theory of vote choice regret (with André Blais and Jean-François Laslier).


  2. The paper builds upon an original pre- and post-election survey conducted before and after the 2015 Canadian election. Directly after Election Day, Canadians were asked for which party they voted, and whether they regret their choice. We find that 39% of them were not perfectly happy with their decision, and 4% even said that they made a bad decision. We show that the surprisingly high proportion of regrets can be rationalized when we consider that voters are maximizing a mixed-utility, composed of both instrumental and expressive benefits. Our mixed-utility theory, which relies on district-level results and answers to only one survey question (i.e., party liking), explains a substantial part of vote choice regret, even when we control for other relevant variables. This study brings an important contribution to the literature on voting behavior, which usually considers that voters are either instrumental or expressive, but not both at the same time.


In preparation


  1. Does inequality harm democracy? The inclusion and representation of the poor in Europe (with Marco Giani).


  2. Two important challenges regarding socio-economic inequality and democracy have attracted the attention of political observers: the poor abstain more than the rich (problem of inclusion), and would vote differently if they had to (problem of representation). How important are these challenges for modern democracies? To answer this question, we use a novel fully non-parametric design based on Coarsened Exact Matching to impute the vote choice of abstainers in 30 elections and 15 European democracies between 1998 and 2014 (N ≈ 55,000). We show that socio-economic inequality translates into a severe problem of inclusion, but it only weakly affects representation.


  1. Citizen demand as a source of personal representation: a field experiment (with Thomas Gschwend, Thomas Zittel, and Steffen Zittlau)


  2. Personal communication between members of parliament (MPs) and constituents, often labelled as personal representation, is important for democracy. Previous studies show that the electoral and institutional context affects the probability of MPs to engage in personal representation. In this paper, we argue that the type of demand they receive from citizens also plays a crucial role. To test our argument, we conducted a field experiment with German national MPs, in which we randomized the type of demand formulated by constituents. Our treatment is a demand for personal, or party representation. We find that (1) MPs receiving a demand for personal representation communicate more, and that (2) this is especially the case for MPs elected in a single-member district.