F15 - Voting
Date: Jun 4 | Time: 08:45am to 10:15am | Location:
Chair/Président/Présidente : Cameron Anderson (University of Western Ontario)
Discussant/Commentateur/Commentatrice : Joanie Bouchard (Université Laval)
Candidate Attractiveness and Electoral Success: Evidence from Quebec: Michael Wigginton (University of Ottawa), Daniel Stockemer (University of Ottawa)
Abstract: Life treats good-looking individuals very favourably: compared to more average-looking individuals, it is easier for physically attractive people to make friends, gain higher paychecks, move up the career ladder faster, and even win elections. Good-looking politicians get a vote premium of up to 7 percentage points in first past the post elections. In this study, we aim to decipher whether it makes a difference for the development of this vote premium whether or not the picture of the candidates is on the ballot. Using elections to Quebec’s National Assembly as a case, we systematically analyse the (additional) effect printing a picture on the ballot has on candidates’ beauty premiums. To do so, we compare the effect of physical attractiveness on candidates vote shares in the 2008 elections, when there was no candidate picture on the ballot, with the 2012, 2014, and 2018 elections, when ballots featured a candidate picture. We also calculate the difference in beauty premium for candidates, who ran both before and after 2012. Using descriptive and inferential statistics, we find that the beauty premium increased by approximately 3 percentage points after 2012; that is before 2012 good looking candidates gained approximately 5 percentage points, if they ran against less good-looking candidates. This beauty premium increased to 8 percentage points after 2012, attesting to the additional impact of physical attractiveness if the candidate picture is on the ballot.
Economic Voting in Canada – Considering the Contextual Effects of Equalization in a Federal System: Cameron Anderson (University of Western Ontario)
Abstract: Economic effects on incumbent support have been well established in the scholarly literature both comparatively and in Canada (Lewis-Beck and Stegmaier 2013). This relationship has also been shown to be contextually dependent on clarity of responsibility whereby effects of economic conditions on incumbent support are stronger as clarity of responsibility increases. While this relationship can be relatively straightforward in unitary systems, federalism adds complexity to how this relationship plays out. In Canada, one way in which the federal system may complicate the relationship is through the system of financial redistribution between the federal government and the provincial governments called equalization. Provinces that economically perform below a national standard receive fiscal transfers from the federal government to ensure the ability to deliver a comparable level of public services. This paper assesses the extent to which the status of being a ‘have’ or a ‘have not’ province influences the strength of federal economic voting. Specifically, it is expected that a greater provincial dependence on federal transfers will increase the strength of economic effects on federal incumbent support. To explore this question, the proposed work will draw on a pooled data set of 1988-2019 Canadian Election Studies combined with provincial-level aggregate indicators of provincial receipt of equalization payments. The paper will contribute to our understanding of economic voting in Canada and how voters cope with the financial and administrative arrangements of federal systems more generally.
The Moderating Effect of Housing on Economic Voting: Alexandra Jabbour (Université de Montréal)
Abstract: Economic voting is one of the most effective mechanisms for explaining electoral choices. While the stability of the mechanism in most established democracies gives economic voting a certain credibility, researchers have recently begun to challenge the association between economic performance and the vote share for the incumbent. This paper is in line with this approach and studies the moderating effect of the local environment on perceptions of the economy. More specifically, I argue that traditional sociotropic and egotropic assessments of the economy are both shaped by local considerations with heterogeneous effects according to the personal situation of each individual. Instead of traditional measures such as GDP, unemployment rate or income, I take into account the cost of housing as a measure. This measure has several advantages: it touches the entire electorate, it is a concrete indicator that does not require high levels of sophistication and it should be less sensitive to the impact of political attitudes such as partisanship. To validate my theory, I proceed in two ways. First, to demonstrate the association between the cost of housing and voting behaviour, I use the LISS panel of 5000 households in the Netherlands that includes measures of the share of the household expenditure devoted to housing. Second, to establish a possible causal mechanism, I use the sudden interruption of rent-control polices In Massachusetts, specifically in Cambridge in 1995. The objective of the paper is to provide evidence that the economic vote might be underestimated.
Climate Change and Carbon Pricing in the 2019 Canadian Federal Election: Erick Lachapelle (Université de Montréal), Kathryn Harrison (University of British Columbia), Matto Mildenberger (University of California Santa Barbara)
Abstract: The Canadian general federal election was held on October 21st 2019. Though few issues really stood out, many polls suggested that climate change was among the top issues during the campaign. Existing research suggests that climate change is only rarely a ballot box issue. Yet research in Canada suggests that climate change attitudes are structured regionally, and the 2019 federal election result reflects these divisions. To what extent did concerns with climate change and carbon pricing translate into votes? Which voters were most likely to vote with climate as a key consideration, and for which parties did they (not) vote? Using a unique four wave panel data set of voters in Quebec, Ontario, Saskatchewan, Alberta and British Columbia, this paper sets out to examine the role of climate change issues in the 2019 Canadian general federal election. It examines the relative roles of issue voting, strategic voting, and economic voting in determining the outcome of the Canadian federal election. It concludes with general reflections regarding the role climate change attitudes and party positions can play in shaping Canadian electoral outcomes.
Do Women Rely More on their Feelings and Leaders’ Character in their Political Evaluation? Evidence from 125 Elections (1942-2015): Jean-François Daoust (University of Edinburgh), Amanda Bittner (Memorial University)
Abstract: Do women and men differ when evaluating politicians? This is an interesting question that has been addressed in a lot of research which concludes that in several cases, there is a gender gap in the extent to which citizens like some leaders. Although very useful at the descriptive level, this information does not improve our understanding of the potential differentiated impact of leaders’ evaluation among women and men. That is, whether women/men are more positive/negative toward particular politicians is not related at all to the weight of that consideration in their vote choice calculus. In fact, women could be more positive in their evaluation, but the weight of this consideration be less important than the weight of leaders’ evaluation for men – or vice-versa. Within a comparative framework, we test whether women rely more on personality traits that relate to a leader’s character (compared to competence) and whether the weight of feeling thermometers is greater compared to men. To do so, we adopt two strategies. In a first study, we focus on five established democracies (United States, United Kingdom, Canada, Australia and New Zealand) providing a longitudinal perspective going back from 1942 to 2015. In a second study, we use the Comparative Study of Electoral System (CSES) dataset which provides substantial variance in terms of context within 100 elections in 40 unique countries from 1996 to 2017.