Just days after Canadians elected Liberal leader Mark Carney as the country’s next Prime Minister, the mix of victories, upsets and expected outcomes across the nation reflects deeper trends that University of Toronto political scientist Semra Sevi has spent years analyzing.
An assistant professor in the Department of Political Science, Sevi has built a groundbreaking dataset tracking every candidate in Canadian federal and Ontario provincial elections since 1867 — over 62,000 candidates in total.
She spoke with A&S News about what her research reveals about representation in Canadian politics.
Your data set focuses on incumbency, among other elements. Do candidates have an advantage if they already hold office?
Incumbents in Canadian federal elections once held a strong electoral advantage, with 78 per cent seeking re-election and 76 per cent winning. However, my research shows this edge has declined over time. The probability of winning dropped from 15 percentage points pre-1972 to just 2 points post-1972, while the vote share advantage fell from 5.4 to 1.5 points. This shift likely reflects electoral reforms, increased competition and a larger number of candidates per riding, such as Pierre Poilievre facing 90 candidates in the 2025 federal election. Both Liberal and Conservative parties saw their pre-1972 advantage diminish. These changes suggest a move toward party-centered voting, reducing the influence of individual incumbents.
Does a candidate’s party affiliation matter more to voters today than their individual personality?
Yes. Before 1972, ballots included only a candidate’s name and occupation, which made personal qualities more influential. Today, however, my research shows that party affiliation plays a much larger role. At the leadership level, Prime Minister Mark Carney had never held elected office — his party’s incumbency likely helped him out here. We also saw a shift after Justin Trudeau resigned. When he was still in the race, voters who preferred Pierre Poilievre may have done so mainly in contrast to Trudeau. Once Trudeau stepped down and Carney became the Liberal leader, voters had a new option, one that didn’t force them to choose between two unpopular figures. This revealed that Poilievre’s support was weaker than it appeared when Trudeau was still in the picture.
When candidates switch parties, does it hurt their chances of winning?
There are many reasons why politicians would switch parties — they could have been involved in a scandal and the party kicked them out, or they see that their party is going to be defeated in the next election, so they run with one that might have a better shot. Either way, candidates who cross the floor are typically punished by voters.
I've been interested in whether women who switch parties face more backlash than men. Anecdotally, it seems they do — figures like former Justice Minister Jody Wilson-Raybould, Belinda Stronach and Eve Adams were often labelled as opportunistic. However, because so few women have made such moves, it's been difficult to gather enough data to draw statistically significant conclusions in our research.
There is a perception that women receive fewer votes than men — what does your data say?
There’s a common perception that women receive fewer votes than men, but my data does not support this. I find no evidence of voter bias against women candidates. In fact, one of my earlier works shows that since 1921 — when women first ran for Parliament — women have received nearly the same vote share as men, with only a negligible half-percentage-point difference. In a more recent article, I show that women incumbents are just as likely to run again as men. However, among those who lose, women are less likely than men to run in the next election.
This points to a different issue: recruitment. My current research suggests that the real bottleneck is getting more women to run in the first place. Political parties in Canada serve as gatekeepers, controlling who gets on the ballot. To increase women’s representation in Parliament, we need more women in the candidate pool, which starts with better recruitment efforts by parties.
Do certain occupations lend themselves well to political success?
Yes — especially lawyers. Our research shows that candidates with a legal background have consistently received more votes than those from other professions, a trend that dates to 1867. This may be because skills like public speaking, debating and understanding laws and policy are directly applicable to politics.
Central bankers do one thing — very well — but not many countries elect them because of that narrow focus. To me, it’s apparent that the trade war and economic uncertainty were top of mind for Canadians when voting for their candidate of choice.
Want to see more of Semra Sevi’s work?
Watch the video supported by her SSHRC Talent Award.