A new Mainstreet Research poll in Marie-Victorin’s driving, where a by-election will be held Monday, puts CAQ candidate Shirley Dorismond ahead of PQ Pierre Nantel and shows a slump in support for others who have disappeared.
When asked, “As you may know, there will be a by-election in your riding of Marie-Victorin to elect a representative to the National Assembly of Quebec. Who do you plan to vote for in this by-election? “, the candidate of the coalition Avenir Québec receives 40% support among the survey’s entire sample of 431 respondents. Pierre Nantel, of the Parti Québécois, has the support of 33% of those polled. The liberal (8%), Solidarity (5%) and conservative (5%) candidates are far behind the frontrunners. In total, 8% of the respondents say that they have not yet decided.
Among the determined voters, the CAQ receives 43% support, against 36% for the Parti Québécois.
Last week, a local survey by Repère Communication in Marie-Victorin (published in News) gave Pierre Nantel a 12-point lead over his CAQ rival.
According to Mainstreet, even though the two surveys were held a week apart, Dorismond would prefer to lead by seven points, a difference just within the combined margin of error (plus or minus 4.7 percentage points somehow) for both candidates. Mathematically speaking, it wouldn’t be wrong to say that this is a statistical draw, but it’s still remarkable progress.
The question now is which of these political formations will perform best on the ground – both on the first weekend in April, during the pre-vote, and on April 11 (Monday), on Election Day.
The scientist in me is particularly interested in the outcome of this race, not only because of the consequences for the dynamics in the National Assembly or even for the possible arrival of an experienced politician like Pierre Nantel who represents the PQ, but also for differences in the measurement of public opinion in that constituency.
In the current context, Repère’s figures were surprising. Recall that in 2018 the sovereign party had narrowly kept its stronghold with Catherine Fournier. And since then, national support for the PQ has waned: an average of six to eight points since the last general election. Is Marie-Victorin an enclave that stands out from the rest of Quebec? A political microclimate? Who knows.
Besides the usual caveats that this was an internal poll and that polls for a by-election are always a perilous exercise (because of the lower turnout than in general elections), one has to rely here on the methodological principle of Occam’s razor, according to which one must first prefer the simplest hypothesis to explain a phenomenon.
In this case, the relative strength of the PQ in this drive is the candidacy of Pierre Nantel. Pierre Nantel, known for his years as MP for the federal riding of Longueuil-Saint-Hubert (which partially overlaps Marie-Victorin’s borders), has remained in public space since his departure from the House of Commons, including as a regular contributor at The game on LCN and other Quebecor media platforms. His fame is undoubtedly in his favour.
Still, one of the two polls is wrong, and the correct one will be the poll that perhaps best measured the approval of the other parties. Indeed, the two probabilistic polls give the same support to the Parti Québécois: 37% for Repère Communication and 36% for Mainstreet.
Repère Communication, on the other hand, estimates the voting intentions for the other parties to be higher than what Mainstreet measures: 15% for the PLQ, 12% for QS and 9% for the PCQ, a combined share of 36%. Mainstreet gives these three parties a total of 20%. And that may be where this election will play out.
Different hypotheses can explain these differences. As mentioned above, the participation rate in midterm elections is generally much lower than in general elections (in the 2016 Marie-Victorin midterm elections, 26% of voters bothered to vote, compared to 63% in general elections) .
However, the low turnout could be even more pronounced among voters of parties who have no real chance of winning. For example, could it be that Mainstreet has measured a decline in enthusiasm among liberal voters? Since the formation is not competitive in Marie-Victorin, could it be that a fraction of the PLQ voters decided to back the CAQ candidate to defeat Pierre Nantel?
Conversely, while tensions between the PQ and QS have been well documented over the years, the measured decline in support for QS in Marie-Victorin might well increase the chances of Pierre Nantel, a former member of the federal NDP (party with which QS shares many voters in Quebec according to the polls)?
Still, the poll analyst in me is particularly curious as to how the results will compare to these two polls. Maybe the pear will be cut in half and we’ll have a near draw at the finish. Perhaps the Mainstreet sample is representative of Marie-Victorin’s electorate, but not those who will participate in this by-election. And perhaps also that the Repère poll grossly underestimates the CAQ: in 2018 it reached 28% in Marie-Victorin, so it is certainly unlikely that it will be below this level at the moment – if we rely on numerous regional and national polls in the past months.
A CAQ victory would undoubtedly reinforce the aura of invincibility surrounding the CAQ, which had also won the other two by-elections of this legislature, in Roberval and Jean-Talon, by comfortable margins. A victory for the PQ would provide Paul St-Pierre Plamondon with rare good news for his party ahead of the fall elections (and would close the lid, at least temporarily, to those predicting the disappearance of the PQ), as my colleague Guillaume points out. Bourgault-Côté in the May issue of News†
So what’s my prediction? The real, honest, scientific answer is “I don’t know”, but I can’t wait to compare the numbers. We’ll know on Monday evening.
The data for this Mainstreet Research survey was collected from April 2-4, 2022, from 431 Quebec voters ages 18 and older who lived in Marie-Victorin’s carriage. The margin of error for the results of the full survey sample is ±4.7%, 19 times out of 20. Respondents were reached through robocalls via land and cell lines. To view the research report, please visit this link.