Last-minute voter changes are probably more common than you might think. In 2018, 36% of respondents to the post-election survey that we conducted with my colleague André Blais said they had decided on their vote during the last weekend and even on voting day or when voting. The proportion of those who had made their decision in the last few days, however, rose to more than 70% among those who had changed their preferences or who were discreet during the campaign.
Posted at 6:00 p.m.
Where are we in this election?
At least three polls will be released today or tomorrow. As for the polls published since the end of August (18), movements in voting intentions seem to be taking shape. If we take into account all the published polls, the more “traditional” parties, namely the Liberal Party of Quebec (PLQ) and the Parti Quebecois (PQ), seem to have gotten better at the expense of their closest opponents over the past days, after the Radio-Canada debate on September 22. Support for the PLQ now exceeds that of the Conservative Party (17% compared to 15%) and the PQ appears tied with Québec solidaire at nearly 13%. The ACQ is relatively stable at 40%.
This rise of the PLQ and the PQ is similar in the greater Montreal area. The PLQ seems to be on the rise and is currently at 25%. The Conservative Party (PCQ) and Québec solidaire (QS) are down and are tied with the PQ at 12%. Finally, support for the CAQ is relatively stable around 35%.
Can we trust the polls?
This election is particularly difficult for the poll analysts since there are notable differences in the estimates of the various firms. On the one hand, there is a significant difference between Leger and Mainstreet, two firms that use very different methodologies — web panel for Leger and automated telephone surveys for Mainstreet. Compared to Mainstreet, Leger rates downforce at QS four points higher (16.2 vs. 12.1), PQ, 2.7 points higher (12 vs. 9.3), and PCQ, 3.5 points lower. In addition to these firms, there are four other firms with different methodologies and samples: Segma (telephone survey), Ekos (automated survey), Angus Reid and Research Co. (web surveys). Far from helping us to see clearly, these firms are distinguished by different estimates. Segma and Ekos estimated support for the PQ at 14-15% and this, before the last debate. Segma and Angus Reid have the same estimate as Leger for QS (16%), but Ekos estimates support at 21%. Finally, Research Co. and Angus Reid have the same estimate as Mainstreet for PCQ (18-19%), but Ekos estimates it at 12.3%.
How to see clearly?
We must first remember that the more polls there are, the more we can think that the biases of some can compensate for the biases of others. Under the circumstances, the weighted average of the polls, taking into account as much as possible the time when the poll was carried out, is the most reliable solution… pending the results of the elections. On the other hand, if there is a lesson to be learned from the last Quebec elections, it is that we must watch the polls until the end of the campaign to find out if the post-debate movements detected seem to be confirmed.