Go to the Globe and Mail homepage

Jump to main navigationJump to main content

Having failed to predict Tory majority, pollsters turn questions on themselves Add to ...

The polling industry prides itself on its ability to look into the future, but it’s going to have to take a hard look at the past as it tries to understand how it completely missed a Conservative majority.

While the nation’s pollsters correctly predicted that the New Democratic Party would see record support on election night and came close on popular vote percentages, it underestimated how support for Stephen Harper’s Conservative party would translate into seats.

If they want to maintain a sense of trust among the electorate, they’ll need to figure out why. Some suggest voter turnout played a role, others suspect that the vote split unpredictably because of the unprecedented popularity of the NDP.

“The industry might want to consider doing some navel gazing,” said Richard Johnston, a political science professor at the University of British Columbia. “I’m struck by the range of predictions being so uncomfortably large – they need to ask themselves why.”

To understand why his EKOS Research failed to catch the Conservative majority, president Frank Graves said the company would need to study the numbers and try to contact some of the people it polled to find out why they didn’t vote the way they said they would. All pollsters should do the same, he said, but once the election is over, the companies move onto other business.

“We need to go back and do some work to find out what went wrong, but nobody has the resources to do this,” he said. “This is done on a shoestring budget, and while we do our best to try and understand our failures, it requires analysis and time and nobody is really interested in that.”

The industry has been under scrutiny since the late days of the campaign as poll numbers veered from expectations, because results seemed wildly divergent at times. Pollsters blamed their frequent gaps on different methods – some used the Internet, some an automated phone system and some did traditional phone interviews.

Nanos Research posted final polling numbers that closely mirrored final results in terms of popular vote on its last day of polling Sunday. President Nik Nanos said his success is proof of the validity of polling, even if some other companies missed the call.

“The main difference for us was polling right up to the last minute,” he said. “I don’t think we need to standardize polling or anything – in general I’d say the results were pretty good.”

* * * * *

While pollsters were close on popular vote predictions, none forecast a Conservative majority. Here is a rundown of the percentage of the popular vote, with results compared to final polls.

TORY Results, as of midnight ET: 39.9%

  • Angus Reid: 37%
  • Ipsos Reid: 38%
  • EKOS: 34.6%
  • Nanos: 38.7%

NDP As of midnight ET: 30.8%

  • Angus Reid: 33%
  • Ipsos Reid: 33%
  • EKOS: 31.4%
  • Nanos: 30.5%

LIBERAL As of midnight ET: 19.1%

  • Angus Reid: 19%
  • Ipsos Reid: 18%
  • EKOS: 20.4%
  • Nanos: 20.9%

BLOC As of midnight ET: 5.8%

  • Angus Reid: 6%
  • Ipsos Reid: 7%
  • EKOS: 6.3%
  • Nanos: 5%

GREEN As of midnight ET: 3.6%

  • Angus Reid: 4%
  • Ipsos Reid: 4%
  • EKOS: 6.3%
  • Nanos: 3.7%


Follow us on Twitter: @GlobePolitics

In the know

Most popular videos »

Highlights

More from The Globe and Mail

Most popular