Seniors and middle-class homeowners are increasingly calling the shots in Canadian politics, as groups who consistently turn out to mark an X on voting day enjoy an influence that is disproportionate to their numbers.
While Prime Minister Stephen Harper and Toronto Mayor Rob Ford were able to win over those key vote-rich demographics, Ontario Progressive Conservative Leader Tim Hudak's campaign failed to resonate with those same voters in an election with a record-low turnout rate of 49.2 per cent.
That Ontario turnout now has pollsters reconsidering their methods and debating whether they should produce two sets of numbers: the traditional survey of eligible voters, and a second set of numbers that gives less weight to groups least likely to vote, such as youth.
Ekos pollster Frank Graves did just that on the night before the Ontario election in a survey for ipolitics.ca. The "most likely" to vote numbers proved to be more accurate. Darrell Bricker, CEO of Ipsos Public Affairs, says declining voter turnout has his firm thinking of using a similar approach.
"It was the middle-class taxpayers around the city of Toronto that elected these guys," said Mr. Bricker, in explaining the key demographic that Mr. Ford and Mr. Harper won over, but Mr. Hudak did not.
Pollsters generally agree that the more people have a personal financial stake in government decisions – particularly by paying taxes – the more likely they are to vote.
Giving added weight in polls to the opinions of seniors and other groups most likely to vote – including homeowners, high-income earners and those with a university education – raises important new problems. Canadians who say they support the NDP and the Green Party are also among the least likely to vote. Surveys weighted by voter turnout risk underreporting public support for those parties, influencing the decisions of voters. Such surveys would also discount the opinions of youth, new Canadians and renters.
Amid growing concern over voter apathy, Mr. Graves acknowledges that pollsters must ensure their methods don't compound the problem.
"We've very strongly made the point that in a situation where now half of Ontario voters or more aren't voting, the answer to this is not to say, 'Oh, well let's not bother measuring them. They don't count,' " he said. "No. If anything, the argument would be that good polling should be very careful to register those missing voices."
Pollster Nik Nanos of Nanos Research isn't sold on the need to release two sets of numbers. He says young voters are quite upfront in telling pollsters they don't plan to vote, so simply asking people how likely they are to vote addresses the polling problem.
As for the larger question of democratic representation, Mr. Nanos said online voting and greater interaction between government and citizens would help address declining voter participation among youth.
"It would be a significant step forward," he said.