.svg)
We often assume that online engagement is strongly related to electoral wins, given how much time many of us spend online. But interestingly, according to our SAMbot data, impressive electoral showings do not require high online engagement. For candidates who already hold significant political capital, engaging in discussions online may not be as important as it is for candidates who do not have equivalent resources and networks.
Though Ana Bailão placed in a close second in the 2023 Toronto mayoral by-election (she received 32% of the vote, while election winner Olivia Chow received 37%), Bailão didn’t draw much engagement on Twitter, particularly compared to her opponents. Chow had more than four times the engagement on Twitter than Bailão, and third-place candidate Mark Saunders had nearly three times as much engagement as Bailão.
Saunders and Bailão shared some similarities — both formerly held significant positions in Toronto (Saunders served as the chief of police and Bailão as the deputy mayor), both were endorsed by major Toronto political figures (Saunders was endorsed by Ontario Premier and former Toronto City Councillor Doug Ford, and Bailão by outgoing Mayor John Tory, along with numerous current city councillors and MPs). Given that there was no incumbency advantage (the theory that voters are more likely to vote for incumbent, established candidates over new faces) in this election, these endorsements likely had a strong effect on polling. In this election with no incumbents, candidates like Bailão, Saunders, and Chow seemed to benefit from some effects similar to an incumbency advantage due to their previous leadership roles in Toronto and influential public profiles.
Both Bailão and Saunders were framed as key front-runners throughout the by-election as the most likely candidates to beat Chow. Polling results ping-ponged between whether Bailão or Saunders was the public’s second choice. If one were to use Twitter engagement as a proxy, however, Saunders would appear to be a significantly more popular candidate.
Saunders received the third-highest amount of engagement on Twitter of the candidates we monitored. Saunders’ peaks in Twitter engagement surrounded the key tenets of his campaign: police funding, safe injection sites and drug use, and bike infrastructure. Saunders’ May 26 tweet, criticizing the existence of a bike lane on a high-traffic street, garnered significant engagement on Twitter, his second-largest spike in engagement throughout the entire election. His largest spike in engagement coincided with a number of tweets regarding volunteer support for Saunders from Doug Ford’s “FordNation'' volunteers on July 23.
Bailão’s peaks in Twitter engagement were related to former Mayor John Tory. Two of her largest spikes in engagement coincided with her May 20 tweet of a photo of her and Tory at a community event, and a June 21 tweet which included a video of Tory formally endorsing Bailão. Her other spikes in engagement coincided with debate clips, graphics sharing her policy agenda and comparing it to Chow’s, and election day.
Whether it was through engaging voters offline, or through using other methods of online engagement, Bailão’s campaign demonstrates that electoral success and online engagement and discussion (at least on Twitter) don’t necessarily correlate. The ground game — the door knocking and in-person events — along with connections and pre-existing relationships continue to be crucial in achieving electoral success.
For more information about our research approach see our methodology section and data release
.png)
-min.png)
.jpg)
.png)
-min%20(1).jpg)
.jpg)
.jpg)
.jpg)
.jpg)
