Monday, November 4, 2013

Polling in the Quebec municipal elections

The Montreal municipal election was setting up to be a very interesting race after the turmoil of the last few years, and when the mid-campaign CROP poll showed Mélanie Joly making big gains, it seemed inevitable that there would be more surveys out to capture the changing landscape. But no, instead that mid-campaign poll which was out of the field on Oct. 15 turned out to be last poll of the campaign.

It left the results of yesterday's election really impossible to call, as Joly had momentum but Denis Coderre looked solid. In the end, Coderre won as the polls had said he would, but with only 32.1% of the vote. That was a far cry from the 39% to 41% support he was pegged to have earlier in the campaign. Joly did indeed finish second with 26.5% of the vote, up slightly from the 24% where CROP had her, while Richard Bergeron finished third with 25.6%, up from the 21% in the final CROP survey. Marcel Côté's campaign could not recover from the early setbacks, and took 12.8% of the vote, up from the 11% of the CROP poll.

The campaign was badly served by the polls in terms of presence, but they did not portray an incorrect picture of the race. As you can see, the polls did capture the drop in Côté's support and the momentum of Joly's campaign. But it gave no hint that Coderre's vote would drop so significantly.

Coderre's party won the most seats in council with 28, but Joly could not get many of her candidates elected, including herself. Her party took just three seats, while Bergeron's Projet Montréal took 20. Even the Coalition Montréal of Côté won more seats than Joly, with six. Another eight were won by smaller parties.
The three mayoral candidates had their geographic blocks, as the map above shows (the basic map was edited from the Wikipedia page on Montreal boroughs). Coderre won the eastern and northern parts of Montreal, along with LaSalle, while Joly and Bergeron split the central part of the city (Joly won some of the more anglophone parts such as Pierrefonds and NDG, while Bergeron won the more left-wing parts of the city such as the Plateau and Mercier). 

Coderre won seven boroughs in all, with Joly winning seven as well and Bergeron taking five. It was a much closer race than it seemed to would be only a few weeks ago.

Elsewhere in Quebec

How did the polls do in other parts of the province? There were some mixed results.

In Quebec City, the final poll from Léger (Oct. 18-22) gave Régis Labeaume 78% of the vote to 21% for David Lemelin. The results were 74.1% for Labeaume and 24% for Lemelin, so a good job by Léger here.

Laval was less successful, as Léger's last poll (Oct. 18-21) gave Marc Demers 36% of the vote to 24% for Claire Le Bel and 17% for Jean-Claude Gobé. Demers did win (with 44.2% of the vote), but Gobé finished second with 24.3% and Le Bel was well behind at 12.4%. Half marks for Léger in Laval.

The mid-campaign poll by Segma Recherche in Gatineau did not reflect the final results very well at all. It was taken Oct. 15-19, so there was some time before yesterday's vote, but the poll gave the incumbent, Marc Bureau, 51% support to 34% for Maxime Pedneaud-Jobin. Instead, Pedneaud-Jobin defeated Bureau with 51.6% to 37.5% of the vote.

Cible Recherche's poll in Trois-Rivières (Oct. 22-27) put Yves Lévesque in front with 41% to 33% for Sylvie Tardif and 20% for Caterine Dufresne. It was a decent estimation, but by no means perfect, as Lévesque took 49.2% to 31.2% for Tardif and 14.1% for Dufresne.

And in Lévis, Léger's mid-campaign poll (Oct. 13-16) painted a different picture than the result - but still had the winner - with Gilles Lehouillier at 42% to 22% for Antoine Dubé and Isabelle Demers apiece. Instead, Lehouillier won in a squeaker with 38.9% to 34.2% for Demers. Dubé took 16.4% of the vote.

This list is not exhaustive, and I may have missed a few municipal polls in these cities (and in other major cities) in my cursory search. But this gives a broad overview of how the polls did: quite well in some places, badly in others, and generally not close enough to the election date.

Hopefully the mayoral campaign in Toronto will be better served in terms of both quantity and quality of the polls. If the 2010 campaign is any indication, as well as the circus that is Rob Ford's tenure, we should at least be assured of the former.