What voters want : identifying voter preferences for candidates

This dissertation is comprised of six standalone articles that provide insights on what type of candidate voters prefer with a particular focus on gender, age affinity, occupation, and political experience (i.e., incumbency). The question of who is elected is one of the most fundamental questions in political science as it pertains to the issue of descriptive representation. The first article presents two novel datasets that I collected. These datasets include information on all candidates in Canadian federal and Ontario provincial elections from 1867 to 2019, and they are the basis for four of the remaining articles in this dissertation. The second article examines whether women get fewer votes in Canadian federal elections. Using the novel data I collected, with over 21,000 unique candidates since 1921 (when the first women were allowed to run for seats in Parliament), we are able to compute precise estimates of the difference in the electoral fortunes of men and women candidates. We demonstrate that while there was a gender gap in the past, the difference between male and female candidates’ vote shares is now statistically indistinguishable from zero. The third article investigates whether women get fewer votes in the Ontario provincial elections. We again estimate the effects longitudinally, using the novel data I collected, from 1902 onwards. The results are very similar to those found for Canadian federal elections. This is important because it shows that our estimates are robust: regardless of the level of government, female candidates are not being discriminated against by voters. While these results might rely on Canadian data, finding similar results at different levels of government enhances the generalizability of my conclusions. The fourth article uses cross-national data from the Comparative Study of Electoral Systems project, covering 853,414 individual voters, 51 countries, 126 elections, and 639 unique leaders. Using this dataset, I test the hypotheses that a leader is more popular among voters closer to them in age and that such voters are more likely to vote for them. I find some support for both hypotheses though the effects are substantively very small. The fifth article asks if candidates who are lawyers get more votes compared to non-lawyers. This paper also leverages the novel data that I collected at the federal level, which includes the occupation and electoral performance of every candidate who ran for office between 1921 and 2015. Our analysis shows that lawyers get more votes than non-lawyers, but that their electoral advantage is very small. The sixth article asks whether incumbents have an electoral advantage and if such an advantage differs across gender. This paper once again uses the novel data that I collected to estimate the electoral advantage enjoyed by incumbents during 9 Canadian federal elections, in 2,739 ridings, from 1990 to 2019. Using a regression discontinuity (RD) design, I compare men and women who have very narrowly won or lost elections on their probability of running again, vote share and probability of winning in the next election. I find that there is an electoral advantage of being an incumbent but that the differences across gender are, with the exception of vote share, not significant. Incumbents are more likely to run again in the next election than their non-incumbent counterparts. Furthermore, women do not suffer an electoral penalty across the three different outcome variables, suggesting that voters are not discriminating against women once they run for office.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.