In our two previous posts we used maps to show that there is remarkable continuity in the geography of support for different types of candidates across the 25 years and eight elections since Toronto’s 1997 amalgamation. Support for progressive and conservative candidates is geographically concentrated. These patterns suggest that we can think more systematically about the relationships between candidates’ support bases across time.
In this post we visualize, across the eight elections, which mayoral election candidates are most and least alike based on the geographic distribution of support they received. We do this by calculating correlation coefficients, which indicate how much each mayoral candidates’ neighbourhood vote shares are similar to, or different from, every other candidate’s. The values range between –1.0 (least alike) and +1.0 (most alike).
After picking a mayoral candidate below, the app shows a map of their electoral support across the city and a list of the other candidates ranked by how similar they are to the selected candidate. Unsurprisingly, candidates who ran against each other in two-way races score least alike to each other – Hall and Lastman in 1997, for example. You will also find that the candidates we highlighted in previous posts as coming from more progressive or more conservative ends of the political spectrum tend to have strong positive scores. Especially interesting, however, are the similarities and differences between repeat candidates. For example, check out John Tory’s geographic support bases across the four elections in which he ran. Hall, Lastman, and Miller each ran twice. Rob and Doug Ford can also be compared. See what you find!
