In Canada, we count seats, not votes. Just like in American Presidential elections, you can win the popular vote and lose the election. In fact, that happened to Justin Trudeau’s father in 1979.
The general way analysts address that is to look at seats by region. However, seats in the same region can move differently. This release tries to get closer to reality in the seat-by-seat contest. Our analysis combines two projects:
- An analysis of federal election districts (which we call “seats”) that groups them into 14 clusters based on which parties are most competitive in those seats. Given the shifting dynamics of Canada’s party system, we have relied on only the results of the past two elections. However, to assess where the parties stand in these 14 seat clusters, we need a lot of data.
- A merge of the most recent three national surveys that include federal votes, creating a unweighted total of 6,420 respondents and a weighted total of 5,900.
In each of our surveys, we collect postal codes. That allows us to create a riding variable for almost all our respondents and to group their responses by the riding they are in.
The deck below shows the output from the analysis.2019 Federal Election - Seat Clusters