Quebec’s general election will be 7 April. From my cursory reading of the news prior to the election call, I had the impression that a majority for the Parti Quebecois (PQ) was all but in the bag. (The PQ currently heads a minority government.) However, sometimes funny things happen during campaigns.
Looking at the projections at ThreeHundredEight.com, which are based on aggregating and weighting polls, one can see that the Liberals are now ahead of the PQ in the vote projection. As of today, they are even ahead outside the projection tool’s estimated confidence intervals*, which range 41%-47% for Liberals and 32%-36% for the PQ.
Seat projections, however, remain closer: 61-76 for Liberals, 46-59 for PQ. If the race tightens even a little bit, the prospect of a plurality reversal becomes real, given the near overlap in project ranges for seats despite the existing minimum projected gap of five points in votes. If we convert the projections into advantage ratios (%s/%v)**, we get a range for the Liberals of 1.18 to 1.28, but for the PQ of 1.16 to 1.31. We can see that either party would benefit from the over-representation expected from a plurality electoral system, but the PQ benefits slightly more–from the model’s projections–when it is at the higher end of its vote range, even though its current maximum projection would not give it a plurality of the vote. Extrapolating from these figures, the PQ might be able to win a majority of the seats on only around 38% of the vote and with the Liberals still slightly ahead.
A reversal–and not the first in the province–happened as recently as 1998, when the Liberal party had 43.55% of the vote and 48 seats, while the PQ had 42.87% but 76 seats. Yes, 76, for a really large majority despite losing the province-wide vote.
I think I will start paying more attention to the Quebec campaign now.
* Taking the confidence interval to be the range from “low” to “high” rater than minimum-maximum.
** Using low-end or high-end projections for both seats and votes in all cases.