Alex Rand writes:
The recent Canadian federal election had one riding’s result determined by 1 vote, which made me think of your old probability of your vote being decisive paper!
And the plot thickens because one person who voted for the party that lost by 1 vote had their ballot rejected due to an error on the part of the Canadian election authority, who had included a typo in that person’s mail-in ballot, causing it to be rejected. See this article from the CBC.
Thought this could be fun blog material if this topic is still of interest to you.
It is! If you’re interested in the general topic, I also recommend these papers:
2012: What is the probability your vote will make a difference?
2007: Voting as a rational choice: why and how people vote to improve the well-being of others
2004: Standard voting power indexes don’t work: An empirical analysis
In particular, the appendix to that past paper (just click on it and scroll down to page 674) carefully explains why our probability-of-your-vote-being-decisive calculations are not invalidated by considering the possibility of recounts and disputed tallies.
The other thing I sometimes like to say to explain this to people is to look at things from the perspective of the campaign rather than the voter, and consider the probability that a swing of 1000 votes could change the outcome of election. There are elections that are close enough that 1000 votes could make a difference . . . now consider different strategies that a campaign might do to try to effect that swing. Getting those thousand votes might be tough, but various pitches and mobilization strategies might do a little bit of it, and you can imagine some particular mailer or whatever that could cause an expected swing of one vote.
In the August 2025 Alabama local elections, two cities (Guntersville and Centreville) had their mayoral race end in a tie. One was resolved after a runoff election, and the other after provisional ballots were tabulated.
A more interesting resolution of a tie took place earlier this year in Corpus Christi, TX: After a tied runoff for a city council seat, officials had to determine a random method to break the tie. Candidates first rolled a die, then drew numbered beads, with the higher number winning the election. The incumbent won the draw-off, retaining his seat.
https://www.votebeat.org/texas/2025/01/10/corpus-christi-city-council-election-tie-breaker-everett-roy-billy-lerma/
Vote in your local elections! They make just as much difference in your life as statewide or federal elections, if not more.
Although not a psepholigist, I’m an election junkie … and turnouts bother me. As do electoral sytems – which makes it very difficult to compare countries as regards the outcome.
This being said, you might consider studying [in] Switzerland: sensationally low turn-out at federal elctions, but with four country-wide referenda per year; the one canton which has a mandatory vote tends to deliver the same proportions as the rest of the country, where on average, only 45% of voters (if you’re lucky) deign to turn up.
On a personal note, I think that first-past-the-post systems suck … (but that’s just me, I suppose)
I worked in a by-election in Canada many years ago where there were 115 polls IIRC. We lost the election by 115 votes.
As the story is several months old I tried searching to see if that was the final result. It appears it was but initially the Google AI (in a result that I haven’t been able to reproduce) reported that:
“Tatiana Auguste was briefly declared the winner of the Quebec federal riding of Terrebonne in May 2025, but she ultimately lost the seat following a revision of the election results. The outcome was determined by a series of recounts and validation processes that corrected previous errors.”
and
“Current status
As of October 2025, Tatiana Auguste is no longer considered the Member of Parliament for Terrebonne, and the Bloc Québécois holds the seat. ”
So be wary of AI summaries.