Go to the Globe and Mail homepage

Jump to main navigationJump to main content


Lysiane Gagnon

The ‘undecided’ can make or break a party Add to ...

From Alberta’s Wildrose Party to Quebec’s Liberal Party, there’s a common thread of badly flawed election polls.

In both cases, the pollsters – and the media who take their cues from the polls – got it wrong. In the recent Quebec election, the Liberal Party was underestimated in part for the same reason that the Wildrose Party was overestimated in the spring Alberta election: The pollsters didn’t take into account the voting patterns of the self-proclaimed “undecided” and greatly overemphasized the popular appetite for “change.”

In Alberta, the media fed on polls that predicted a Wildrose victory. Party leader Danielle Smith was a fresh face. Change was in the air. The 41-year-old Tory dynasty was about to fall!

The Progressive Conservatives won handily with 44 per cent of the vote. Embarrassed pollsters attributed their mistake to a last-minute, massive change of heart among the electorate – a turnabout caused, they said, by the inopportune comments of a couple of bigoted Wildrose candidates. But this analysis seems a bit short.

Similarly, in Quebec, the polls and their faithful media followers (myself included), traumatized by their failure to forecast the fall of the Bloc Québécois and the rise of the NDP in the 2011 federal election, assumed that the same desire for change would manifest itself at the provincial level. The conventional wisdom was that the newly formed Coalition Avenir Québec would form the Official Opposition and that the governing Liberals would be sent off to near-oblivion.

But on election night, stunned observers discovered that the CAQ, the so-called game-changer, came third and that the Liberals, far from being punished, came within an inch of winning a fourth mandate. They won 50 seats (only four fewer than the Parti Québécois) and 31.2 per cent of the vote, a mere 0.7 points behind the PQ.

After such a performance, the Liberal Party is still an interesting career choice for ambitious politicians. Its upcoming leadership race (outgoing Premier Jean Charest resigned after losing his riding of Sherbrooke) promises to be quite exciting, as several personalities, including high-profile former health minister Philippe Couillard, are already vying for the job.

Quebec’s two major polling firms, CROP and Leger Marketing, underestimated Liberal support by five and four points, respectively, because they distributed the undecided proportionally to the decided, even though undecided voters tend to vote for the Liberals.

During the 1980 referendum on sovereignty, several polls predicted a Yes victory, but polling firm Sorécom and its renowned adviser, distinguished McGill University sociologist Maurice Pinard, correctly predicted the result by allowing a larger portion of the undecided and the “discreet” to the No side.

Unfortunately, this method fell out of use in recent years. Claire Durand, a Université de Montréal sociologist, was the only analyst to foresee the strength of the Liberal Party because she went back to the old, proven method of Prof. Pinard.

The pollsters also ignored another factor that worked for the Liberals: a formidable organization that “got the vote out,” an advantage the CAQ’s fledgling organization was unable to emulate.

All in all, Quebeckers were not in the mood for change. They elected the PQ, a logical alternative, but made sure it would be tightly controlled by refusing it a majority and flanking it with a forceful Liberal opposition.

Report Typo/Error

Follow us on Twitter: @GlobeDebate

Next story




Most popular videos »

More from The Globe and Mail

Most popular