Although the 338Canada Electoral Projections are designed to project the result of the next general election, data is data, so let's compare the results of those by-elections with the riding projections published last Sunday, February 24th 2019.
By-elections are usually far more difficult to project, because 1) very little local data is available (reliable data that is, not "internal polls"), 2) the turnout rate is more often than not much lower than in a general election, so there is greater uncertainty on who will go vote (and who won't), and 3) the small amount of coverage makes it hard to estimate "hyperlocal" effects.
Nevertheless, let's take a look at all three by-election results and compare them with 338Canada's latest projections:
Burnaby-South, BCJagmeet Singh has finally become an elected MP with his victory in Burnaby South. Because the projection was designed for the general election, the Green Party was still included in the graphs. My intentions were to create a side projection for the by-election, but the data available was unfortunately too limited.
Take a look:
The LPC and CPC projections were right on the mark. The model slightly overestimated the PPC's performance (although getting to double digits is still considered a fairly good result for the PPC).
The NDP's results tells me that, with an effective ground game and get-out-the-vote efforts, the NDP was able to consolidate most of its support and a fair fraction of Green voters. With a Green candidate on the ballot, it is safe to assume Mr Singh's margin of victory would have been much narrower.
The departure of Tom Mulcair (or, rather, his ousting from its leadership) has hurt the NDP a great deal in voting intentions, especially in Quebec, where the NDP won over 40% of the vote in 2011, about 25% in 2015, and is currently polling at barely 9%-11% in the province.
Therefore, the clear favourite going into the Outremont by-election was the Liberal candidate. Indeed, the LPC used to win Outremont easily (24 out of 25 times between 1935 and 2006) before Tom Mulcair's arrival on the federal scene in 2007.
Here are the projections and results:
The Liberal, Green and Bloc projections were almost perfect, but the model underestimated the NDP candidate (Julia Sanchez) by 9 points. It also missed the mark with the Conservative vote - the CPC finished 5th with a disappointing 7.3%.
York-Simcoe was projected as a safe conservative seat, and the results confirmed it clearly. Conservative candidate Scot Davidson won 53.9% of the vote, just barely above the projection's confidence interval.
The Liberal and NDP projections were dead on. The model did overestimate the results of the smaller parties (GPC and PPC).
* * *
It would be unwise to brag about those results (doubt is safer than certainty), but it gives me great confidence on the model's performance and the quality of the poll aggregate so far.
The next 338 Projection Update will be published on Sunday, as per usual.
Have a great day!
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.