Sunday, 23 December 2018

338Canada Electoral Projection Update: 2018 Ends With High Uncertainty

Several Canadian polling firms have been on the field since late November to measure the pulse of Canadian voters, and so we end the year with a slate of fresh data to sink our teeth in. Indeed, in the past four weeks, six different firms have publicly released voting intentions of Canadian voters and, while the numbers vary from one firm to another, taken together these numbers show that the Liberals end 2018 on a downward trend.

[Here are all publicly available federal polls that have been published in 2018. Visit this page for the complete list.]

We add these numbers to the 338Canada model and, with the help of these polls' regional and demographic sub-samples, we publish today a brand new federal projection that depicts the current political landscape in Canada at the dusk of 2018.

Popular Vote Projection


Riding just under the 40% mark for most of the fall, the Liberal Party of Canada ends the year down to an average of 35.0% of popular support. The Liberals have their work cut out of them for 2019, because even though they are still leading the pack nationally, such a level of support could prove insufficient to win a second straight majority (as we will see below).


Not far behind the LPC, the Conservative Party of Canada finishes 2018 with an average popular support of 33.1%. When we look back at the CPC numbers throughout the year, we notice that, outside of natural fluctuations, its support has been incredibly stable - never much below 30% nor much above 35%. Thus, when the seat projections show a tightening between the LPC and the Conservatives (as it is the case this week), it is better explained by the Liberals dropping then by a Conservative surge.

Even though the New Democratic Party ticks up a point and a half to 16.1% this week, the left leaning party has little cause for celebration this year. Compared to its 2015 results, the NDP is down in every region of the country and is on the verge of being wiped out of Quebec (where it won 16 seats in 2015, highest number of any province). We will watch the Burnaby South by-election closely this winter, as it could have national consequences: should Jagmeet Singh fail to get elected to the House of Commons, an NDP caucus revolt would not be out of the question, as Don Martin of CTV News wrote in his 2019 predictions. The Green Party has already announced that it will not have a candidate for the by election as a "courtesy" for the NDP leader.

A cynical liberal strategist could think his party should do the same to make sure Mr Singh stays in place for the general election next fall. After all, no federal party benefits more from the NDP's misfortune than the Liberal Party. But that's a whole other story.

Here are the popular vote projections with their 95% confidence intervals:



Seat Projections


With the current level of support, the Liberal Party of Canada wins an average of 166 seats (down from 189 last week), just 4 seats shy of the majority threshold of 170 seats. As you can see from this graph below, the Liberals have the highest average seat count, but their confidence intervals overlap significantly those of the Conservatives - meaning there is a high level of uncertainty on which party could win the most seats.


The Conservative Party of Canada ticks up this week to an average of 133 seats nationally (significantly higher than its 2015 total of 99 seats). As we will see below, such level of support would make the CPC a strong underdog should those numbers hold to next fall. However, notice how wide those confidence intervals spread: from 92 to 172 seats. This tells us once again how uncertain and blurry the current picture is.

The New Democratic Party wins an average of 27 seats, up almost ten seats from last week's numbers. However, such a result would still be quite a setback compared to the 44 seats Thomas Mulcair and the NDP won in 2015.

In Quebec, the Bloc québécois takes advantage of the Liberals' slide in La belle province: the current BQ average is 8 seats (it won 10 seats in 2015). We will watch with attention in January to see whether BQ leader candidate Yves-François Blanchet (the lone candidate so far) will move the needle for the pro-independence party.

The Green Party of Canada takes an average of 3 seats per simulation, all of which are located on Vancouver Island in BC.

Finally, the People's Party of Canada is currently leading in the swing district of Beauce, PPC leader Maxime Bernier's home riding.


Projection of the winner


The 338Canada model ran 100k general election simulations that took into account several weeks of polling, demographic data from the Canadian census and the electoral history of all 338 districts.

With current popular support and seat projections, the Liberal Party wins the most seats in three quarters of all simulations (75.8%). It wins a majority of seats in 45% of simulations.


The Conservative Party wins the most seats in 23.5% of simulations. Although a Conservative majority is not impossible with current numbers, this scenario would be considered an outlier (3.7% of simulations).



Regional Distribution


In the Atlantic provinces, where the LPC swept all 32 electoral districts in 2015, the Liberals are still leading comfortably the seat projection, although another sweep looks rather implausible. The Conservatives could make gains, especially in New Brunswick.



The hopes for a Trudeau reelection have at least one clear path, and it is to sweep Quebec and most of its 78 seats. According to current numbers, the LPC could win as much as 55 seats in Quebec, way up from its 2015 total of 40 seats. Those potential liberal gains come mostly from the NDP, which could be reduced to two seats: Rosemont-La-Petite-Patrie (Alexandre Boulerice) and Berthier-Maskinongé (Ruth Ellen Brosseau).



In Ontario, recent numbers have fluctuated so much that seat projections have a tremendous amount of uncertainty. Look at those confidence intervals:


While the Liberals hold the highest Ontarian seat average with 58, the confidence intervals spread so wide that we have a statistical tie. In fact, when running test simulations, the model calculates that a mere two point swing between the LPC and the CPC could cause at least 25 seats to change colour.


In the Prairies, the Conservatives are poised to win almost every seat outside of Winnipeg and Ralph Goodale's Saskatchewan district of Regina–Wascana.


It is no surprise that the Conservatives are ahead in Alberta, but with the support just under 60% of Albertan voters, the Conservatives are currently projected leading in 33 of the province's 34 seats. This blue wave would wipe the Liberals out of Alberta (where they won 4 seats in 2015) and the only district left would be Edmonton Strathcona (where the NDP is still favourite).



Finally, British Columbia has without a doubt the highest concentration of swing districts and three way races. While the Liberals still hold a narrow lead over its rivals, every party - including the Greens - can hope to make gains in BC.




In conclusion



Less than ten months away from the 43rd Canadian general elections, we end the year 2018 with high levels of uncertainty regarding the voting intentions of Canadians.

While the Liberals are still in the lead nationally, their current position cannot be described as comfortable. Should a cooling in the global economy affect Canadians in a negative way, we could see anger and dissatisfaction towards the Liberals translate into support for its rivals. One point or two in Ontario could swing the projection in favour of the Conservatives and make Justin Trudeau a one term Prime Minister.

Quebec will also have a say on the 2019 election. The Liberals have been leading by wide margins in Quebec since the 2015 election, but a renewed leadership for the Bloc québécois (and what looks like an inevitable nomination for Mr Blanchet) could open up room for an anti-liberal and/or protest vote, especially with the NDP so low in the province.

It will be a fascinating year.

Merry Christmas and Happy Holidays to 338Canada readers!

* * *

If you enjoy this blog, thanks for sharing it with friends and relatives. I believe Canadian media needs data based, spin free and non partisan political analysis, which I intend to provide all the way to the election next fall.

I wish to sincerely thank all those who support this blog and comment respectfully on 338's Facebook page and on Twitter.





Philippe J. Fournier is the creator of Qc125 and 338Canada. He teaches physics and astronomy at Cégep de Saint-Laurent in Montreal. For information or media request, please write to info@Qc125.com.


Philippe J. Fournier est le créateur de Qc125 et 338Canada. Il est professeur de physique et d'astronomie au Cégep de Saint-Laurent à Montréal. Pour toute information ou pour une demande d'entrevue médiatique, écrivez à info@Qc125.com.