Will the media and pollsters please stop saying the Ontario election is over before it’s even begun?

By Steve Paikin - Published on Mar 19, 2018
Polls are merely a snapshot in time. They cannot predict the future. (Justin Tang/Frank Gunn/CP; Fred Lum/G&M; A.V. Elkaim/CP)



​I’ve never understood why there seems to be an unholy alliance between some pollsters and some media outlets, designed to end election campaigns before they even begin.

Have you been reading the headlines lately?

“Ontario voters to cast ballot for Doug Ford, poll says,” screamed one headline a couple of days ago.

Except that, actually, the poll said nothing of the kind.

“Ontario PCs poised for majority despite unpopularity of Doug Ford: poll,” said another headline. Well, I suppose. Except replace Ford’s name with the last three Tory leaders and you could have used the same headline. And none of that came to pass.

“Progressive Conservatives driving towards majority government,” said a third headline.

Sort of. Except that the election won’t be officially called until May 9, and then there’s something called an “election campaign” that lasts nearly a month after that. And if there’s anything we should have learned about elections in this province and in this country over the past several decades, it’s that polls are not predictive. Read that again: polls are not predictive. They cannot tell you what people will think three months from now. They are usually pretty accurate about indicating what people thought two weeks ago.

See the difference?

Before the 1985 Ontario election campaign, the Progressive Conservatives, led by their new leader Frank Miller, looked like a pretty safe bet to extend the party’s 42-year dynasty. But Miller ran a lacklustre campaign, and on election day, it was David Peterson’s Liberals that actually won the most votes and formed the government when Miller lost a confidence vote on the floor of the legislature.

What the political gods give, they also taketh away. Five years later, three months before election day, Peterson seemed to be on easy street, as the Liberals were preparing to cruise to a second consecutive majority win. But mistakes were made. The campaign mattered. And Bob Rae became Ontario’s 21st premier instead, in one of the most shocking election upsets ever.

Five years later, again at this point on the political calendar, the Liberals looked to be a sure bet to return to power. They were polling at above 50 per cent. But a guy named Mike Harris had an excellent leaders’ debate, ran on his Common Sense Revolution, and became the first provincial party leader in more than seven decades to go from third to first in one election. Three months before election day, the polls and newspaper headlines were not predicting that outcome.

Want more? Three months before Liberal Dalton McGuinty sought re-election in 2007, he and PC leader John Tory were tied in the polls. Then Tory announced his solution to what he considered the inequity of Catholics having a fully funded school system of their own — a right no other faith enjoyed. The PCs’ numbers cratered as the Liberals successfully exploited public fears about an even more religiously divided school system, and McGuinty won his second majority.

Kathleen Wynne has her own history of defying the polls. In 2014, three months before election day, the PCs again looked set to deprive the Liberals of another win. But leader Tim Hudak unveiled his deficit-fighting plan, which included eliminating 100,000 jobs in the broader public service. The Tories never recovered, and Wynne captured the Liberals’ fourth consecutive victory.

It’s often been the same story federally: Paul Martin was guaranteed to win 200 seats in 2004 (except he barely hung on with a minority government); Stephen Harper was completely unelectable (except he won the next three elections); and, of course, Justin Trudeau was as close to becoming prime minister three months before the 2015 election as the Maple Leafs were to hoisting the Stanley Cup during the Harold Ballard years.

The pattern is clear: polls cannot tell you how people will vote, and yet in every election, there seems to be this unhelpful need by some pollsters and media outlets to declare elections over well before they’ve even begun.

In spite of all the evidence to the contrary, why do they still do this? For one thing, it drives clicks. Second, it can help drive a convenient narrative. One can imagine a publication such as the Toronto Star publishing poll after poll of the Tories in the lead because they’re trying to awaken progressive voters to the possibility that Doug Ford could be premier, something I suspect the Star would have big problems with, if its editorial pages are any indication. One can also imagine an outlet such as the Toronto Sun publishing polls that help make Ford look like a safe choice many Ontarians are already comfortable with. Plus, in many cases, the media get the polls for free in exchange for giving the pollsters publicity.

But I’m not sure it helps the media’s credibility to insist that polls are saying things that they’re simply not saying, so remember this: all the polls today are saying that if the election were held today, the PCs would win a majority government. Fine. But the election isn’t being held today. It’s being held on June 7. And as we’ve seen in just about every election over the past three decades, it’s more likely than not that something will happen over the next 80 or so days that will move the numbers one way or another. Yes, the Tories may in fact finally get back onto the government benches with Doug Ford as their leader. But we don’t know that today. We do know that campaigns matter. So how about, for the heck of it, we just pay attention to the one that’s going to start in about a month and a half and see what happens?

Because in this business, “you just never know” actually happens most of the time.

Thinking of your experience with tvo.org, how likely are you to recommend tvo.org to a friend or colleague?
Not at all Likely
Extremely Likely

Most recent in Ontario Election