The takeaway from the new poll from Toronto-based Forum Research seems rather remarkable.
If an election were held today, Forum says, 39 per cent of the country would vote for Andrew Scheer’s Conservative Party and 35 per cent would vote for Justin Trudeau’s Liberals.
Forum’s president Lorne Bozinoff doesn’t stop there, though. He then applies his poll result to a seat distribution model and concludes that, if an election were held right now, the result would be a Conservative minority — can you say Prime Minister Andrew Scheer? — where the Tories would have precisely half the seats in the House of Commons. The Tories would win 169, while the Liberals would win 130 seats, the NDP 26, the Bloc Quebecois 12, and the Green Party would keep its single seat.
“Trudeau enters the fall legislative session with his popularity slipping,” Bozinoff said in a statement accompanying the release of the poll on Sunday morning. “The primary beneficiary of Trudeau’s decline is Andrew Scheer’s Conservatives.”
This is just the second poll since early October 2015 that finds anyone but Justin Trudeau’s Liberals in first place. (The other one, from March, was also from Forum.) So, the question quickly posed on social networks is does this poll represent the reality of Canadian public opinion or is it an outlier?
Rival pollsters and statistics scientists will have their own opinions but the only real answer is that we truly cannot know if this is an outlier or reality unless we had an actual general election and were able to compare Forum’s poll against actual results.
Some pundits and political operatives will dismiss Forum out-of-hand because of some polls it has published in the past, where it tries to track political preference in just one riding. (It swung and missed by a wide margin, for example, during a 2013 by-election in the Manitoba riding of Brandon Souris.)
But when Forum takes the national pulse, its record versus the actual results has been as good as its peers and, in some cases, much better. For example, its final poll before the 2015 general election found 40 per cent support for the Liberals, 30 per cent for the Conservatives and 20 per cent for the NDP. The actual results? A 39.5 per cent lead for the Liberals, 31.9 per cent for the Conservatives and 19.7 per cent for the NDP.
Get daily National news
In the 2011 general election, Forum was, by some measures, best among its peers when it came to its final poll of that campaign versus actual results.
But in this case, of course, we will not have actual results with which to compare Forum’s. The next best thing then is to take a look at several recent polls to see if Forum may have picked up on a trend.
The Forum poll, which was in the field on Sept. 13 and 14, finds the Conservatives up by four points over the Liberals. Meanwhile, the weekly tracking poll from Nanos Research, for the week ending Sept. 15, finds the Liberals with more than a 10-point lead over the Tories. For the week ending Sept. 8, Nanos had the Liberals were up by 12 points on the Tories.
Campaign Research was in the field Sept. 8-11 and it found the Liberals with a 12-point lead. Abacus Data, polling from Sept. 1-3, also found a 12-point Liberal lead . Mainstreet Research, polling Aug. 28-31, found an 11-point Liberal lead.
So you be the judge: One pollster, Forum, finds a four-point Conservative lead while five other polls from four other pollsters done around the same time find the Liberals up by 10 points or more.
Forum, for the record, uses an interactive voice response telephone survey technology and polled 1,350 for its most recent poll. Forum says the margin-of-error is three percentage points, 19 times out of 20. Campaign Research and Abacus cannot calculate a margin-of-error because their survey population is not selected on a truly random basis. (That’s not to say they may not be accurate; it’s just a statement of statistical fact.) Mainstreet, which also uses a telephone survey method and, as a result, can calculate a margin-of-error, surveyed 2,000 Canadians, reached on both landline and cell phones, and claims it is accurate to within 2.19 percentage points. Similarly, Nanos uses a telephone survey of cellphones and landlines and factors in 1,000 results collected over a four-week rolling sample. It says its margin-of-error is 3.1 percentage points.
And yet, if you take all the polls published since the 2015 election and plotted the results on a graph, as a Wikipedia contributor has done, there appears to be some evidence the spread between Liberals and Conservatives has been tightening somewhat since mid-summer. Still, there’s no denying the big picture that shows Liberal dominance in all polls since the 2015 election.
So what to make it of it all? On social media, the response among political partisans is predictable: Liberals dismiss the Forum poll as an outlier while Conservatives hold it up as proof of how public opinion has swung against the Liberals, likely as a result of their recent misadventures in tax reform.
Outlier or not, thoughtful partisans on either side see it as a reminder that, in politics, odd things can and do often happen. In Canada, we need only point to the circumstances through which B.C. and Alberta ended up with NDP governments or the United States ended up with Donald Trump.
So this Forum poll may remind some Liberals of the dangers of complacency and the damage that may be done to their brand if not enough promises are kept on everything from Indigenous issues to climate change.
For Conservatives who may have privately doubted that new leader Andrew Scheer can topple Trudeau, this Forum poll may give them some heart. The Conservatives have a solid base of support of at least 30-32 per cent no matter the pollster and that party continues to dominate when it comes to political fundraising.
As for the country’s New Democrats, they’re busy wrapping up a leadership race and should have a new permanent leader early next month. That will introduce a new dynamic into the federal political scene and it will be at that point, that all these polls – will start to take on a new importance.
Comments