OTTAWA - A battery of conflicting federal horse race numbers is pouring fuel on a raging debate over the reliability of political polls.
A new poll by Harris-Decima for The Canadian Press added Wednesday to the confusing mix of surveys released over the past week.
It put Tory support at 36 per cent, the Liberals at 28 and the NDP at 15.
The telephone survey of 2,020 people was conducted Feb. 17-27 and is considered accurate within plus or minus 2.2 percentage points, 19 times in 20.
The results are in sharp contrast to other recent surveys which have put the Tories as high as 43 per cent, 16 points ahead of the Liberals, and as low as 32 per cent, statistically tied with the Liberals.
Harris-Decima chairman Allan Gregg said his firm's latest numbers are consistent with the longterm pattern in which the Tories and Liberals have essentially been stuck for five years. And he took a veiled poke at rival pollsters, who've been offering contradictory interpretations of what's behind their findings of seemingly dramatic fluctuations in voter intention.
"While I would love to claim that negative advertising, charges by Elections Canada against Conservative officials or 'the Oda effect' is having a great impact on Canadian voter preference, we believe that within the bounds of statistical error, there has been no change this week," he said.
Gregg has incurred the wrath of some of his competitors since he went public last month with blunt criticism of the methodological problems facing the industry and the penchant of pollsters and reporters to over-hype dubious between-election political survey results. His views were echoed by several other pollsters in a story by The Canadian Press but Gregg, a high profile, veteran pollster, has taken most of the heat.
At least one competitor complained about the "spurious" story in an email to the Market Research and Intelligence Association, the industry's voluntary self-regulating body, and warned: "I would not underestimate its negative impact on our industry." The pollster urged the MRIA to respond to Gregg's charges.
The MRIA fought back earlier this week with a full-page ad in the parliamentary precinct newspaper, The Hill Times.
The ad, aimed at debunking "some myths" about the polling industry, maintained that public opinion research in Canada is "both valid and reliable." As proof, it cited the accuracy of polls in predicting the outcomes of the past two federal elections.
The ad also argued that polls strengthen democracy "by giving voice and influence to Canadians." And it concluded by saying the MRIA stands "behind this message 100 per cent of the time, 20 times out of 20."
But Gregg slammed the association for effectively implying the polling industry is 100 per cent accurate, all the time.
"Their response is we have no problems whatsoever," he said.
"That's like Suncor saying, 'No, those ducks didn't go into the tailing pond or they just got a little dirty when they went in there."'
Gregg said the proliferation of sometimes conflicting polls and the hypeventilating analysis that frequently accompanies them does not strengthen democracy. On the contrary, he said: "Rather than have a public that's informed, you have a public that's misinformed."
He said he's not arguing that polls should be ignored; only that their import needs to be interpreted much more cautiously. Rather than pontificate on weekly fluctuations in individual polls, he said it makes more sense to average the results of various surveys and look at the trends over longer periods of time.
"It's not so inaccurate that we should be just saying, 'Oh, polling is in complete disrepute and the discipline is so fatally flawed that we should never look at polls."
Gregg has drawn most fire for his assertion that the industry's ability to yield accurate results has never been worse, due to plummeting response rates on telephone surveys and yet-to-be resolved problems with online surveys.
Darrell Bricker of Ipsos Reid -- which this week put the Tories at 43 per cent, within reach of a majority -- called those concerns "nonsense" and "factually incorrect."
"It's not like this multi-billion dollar research industry isn't struggling with this every day and coming up with some very clever ways of dealing with it. We are, all the time," he said.
"The market research industry isn't broken, it's in transition, as it has been since we stopped doing door-to-door interviews."
Bricker agreed with Gregg that some pollsters and reporters over-hype poll results. But he added: "Some of the guys who are the worst offenders were the people (The Canadian Press was) quoting in the article. So, physician, heal thyself."
Bricker said he welcomes the proliferation of horse race polls since "the more repetition you have, the more you know if you're getting reliable results" and the easier it is to spot a rogue poll.
Pollster Nik Nanos said it's incorrect to point to the widely varying numbers in recent surveys as proof that polls aren't reliable. He argued that polls from different companies can't be compared because no two companies use identical methodologies and generally aren't in the field at the exact same time.
Timing is particularly crucial, he said, since a lot can happen in a few days to cause a shift in voter intention. Nanos pointed out that the Ekos poll released last week, which put the Tories at only 32 per cent, was actually conducted Feb. 4-9 -- a week or two earlier than other polls which have put the Tories considerably higher.
Moreover, Nanos said, given the margins of error, there's really very little difference between Harris-Decima's latest numbers and those of Ipsos Reid. Both could actually be finding the Tories hovering at about the 40 per cent mark -- almost precisely where Nanos' last poll put them.
