What Do The Polls Tell Us?

Political junkies love to look at poll results. Are they reliable predictors of what will happen?

They can be. In the USA, Nate Silver aggregates the results from all the published polls in making his predictions. In the 2012 presidential election, he correctly predicted the winner in every state, including a forecast that Florida would be the closest race and that Obama would win it.

Silver’s counterpart in Canada is Éric Grenier, whose analysis is published by the CBC. His results in past elections have been mixed, with a substantially wrong predictions of in the 2013 BC election and the 2012 Alberta election. He seriously underestimated the Quebec Liberal vote in 2012. He did much better in the 2014 Ontario and Quebec elections.

Silver has also had challenges, predicting a near tie in the 2015 UK election. His final forecast was for the Conservatives to win 278 seats (or somewhere in the range of 252-305 seats), Labour to win 267 (240-293), the Scottish National Party 53 (47-57), and the Liberal Democrats 27 (21-33). The actual final results were 330 seats for the Conservatives, 232 for Labour, 56 for the SNP and just eight for the Lib Dems. The Conservatives led Labour by 6.4%. Virtually every pollster in the UK got it wrong.

Both Silver and Grenier are appropriately humble about the mistakes, and continue to refine their methodology.

All this is context for considering the polls leading up to Monday’s vote. When the election was called, the polls showed the NDP leading with 33% of the vote, followed by the Conservatives at 31% and the Liberals at 26%. The NDP peaked at 37% on Aug 24 and have been going down ever since, reaching 24% on Oct 14. The Conservatives wobbled a bit and are slightly higher at 30%. The Liberals were unchanged on Aug 24 but have since risen to almost 36%.

Given the major variations in poll results since July, it can be said with confidence that almost all of them will prove to have been poor predictions of the final result. If the election campaign had been 37 days long (as it was in 2008 and 2011) we would have voted on Sept 8—at which time, the NDP still held the lead and the Conservatives were in third place.

If the most recent polls prove to be accurate (by no means a sure thing), there will be a clear Liberal plurality and they will form the government. The other two parties will be in some disarray, so the Liberals can look forward to a period of tepid opposition in the House of Commons. They will certainly be able to count on NDP support to overrule any obstructions from the Senate.

The NDP will have every reason for dismay having received a third less than their peak poll result. Pretending to be centrist did not work well for Mulcair, and he was unmasked by the party’s late reversion to type: opposing the TPP with talking points taken verbatim from union leader Jerry Dias; promising to legislate bank fees; pushing for universal pharmacare. Mulcair will regret having promised to wipe the floor with Justin Trudeau.

The Conservatives will, of course, also be disappointed—but perhaps not so much if they take the long view. If they achieve a plurality, they might wait till February to meet Parliament. The Speech from the Throne will likely result in a vote of non-confidence. Mulcair (no longer smiling) will tell the Governor General that he will support Trudeau, who will invited to form a government in March. The Conservatives will face the same future and will be blamed for a pointless five months of uncertainty.

They will be better off facing the music in October. Harper will almost certainly step down, and the party can get on with the business of finding a new leader. There are two reasons why Conservatives thinking strategically should prefer that outcome. First, the voter hostility to the party is almost entirely focused on Harper and should not much attach to a new leader. Secondly, the Liberals having tacked to the left, there is a lot of room left just a bit to the right of centre.

Canadians like politicians to balance the books. It is disaffection with the other two parties, not the Liberal plan to run $10 billion deficits, that will have earned them the chance to govern. Their platform’s doubtful math will likely cause the deficits to be higher, possibly much higher. Fertile ground for a renewed Conservative party.

If the NDP wins the most seats, or come second to the Conservatives, or the Conservatives win a majority, the entire polling industry will have to hide in embarrassment. In almost any other scenario we will sooner or later have a government led by Justin Trudeau.


Related Articles

Who Represents You + Show all articles

Reference Material