Advertisement

B.C. pollsters will revise methods after unexpected Liberal majority

VANCOUVER – Polling companies in British Columbia say they’re planning to make changes so they can predict future votes more accurately as they face a backlash from the public and their clients following the unexpected election of another Liberal majority government this week.

But one outspoken pollster wondered whether the changes would be worth it.

The Liberal win Tuesday night surprised pollsters, many of whom foresaw a win by the New Democratic Party in the weeks and days leading up to the vote.

Decision BC: Election night in 3 minutes

Polling companies have been trying to figure out what went wrong, offering up factors such as undecided voters, last-minute decision making at the ballot box, and low-voter turnout.

Story continues below advertisement

Angus Reid, chairman of Angus Reid Public Opinion, said he knows what his company must now do in B.C.’s complex, multi-ethnic voting environment: invest in multilingual polling and connect with younger voters who typically make the difference between victory and defeat for left-of-centre parties.

It must also get a better handle on NDP voters, noting the correlation between “what they say and what they do” is not necessarily tight.

But at the same time, Reid questions whether the cost of making the changes is justifiable, especially when the media companies that use the service pay a pittance for the work.

Breaking news from Canada and around the world sent to your email, as it happens.

Reid noted there was a time in the mid-1980s when news organizations might spend $200,000 in today’s dollars for polling services, but today they’re lucky to make a “symbolic payment” of $2,000 to $3,000.

“What if I end up on my desk with a proposal to spend half a million dollars to develop some additional techniques and methods to make sure this never happens again?” he asked. “The big question that my board of directors is going to ask, is, ‘Well why should I bother? Who are we doing this for?’ And I think that’s really the question.”

“And frankly if the entire public-polling business ends up simply shutting down or being left to, you know, really a bunch of charlatans, then I think it’ll be a sad day for Canada.”

Story continues below advertisement

Reid said he finds it ironic that the media is now jumping all over those who have been largely donating their services to understand elections.

Reid noted his $100-million, research-and-software company employs about 600 people worldwide and about 300 of them are in Vancouver.

“When we’re right we get zero credit, and when we’re wrong it’s as if we lost the election,” he said.

Meantime, two other companies, Insights West and Ipsos Reid, which conducted two and four polls, respectively, during the campaign, said they’re already planning to make changes.

“We’ve got a lot of work to do at better replicating the actual voting decisions of voters,” said Kyle Braid of Ipsos public affairs, adding he wants to accept responsibility for the error that happened.

Braid said part of the problem was that while his company polled members of the public, only half of the electorate actually cast ballots.

“As an industry, we need to get a better understanding of who actually is going to turn out and how we can ask questions that really get a good forecast as to who’s going to turn out,” he said.

Doing so, he said, would allow the company to issue two sets of figures: one that will focus on the general public, and another that will focus on those who are likely to vote and not just intend to vote.

Story continues below advertisement

Ipsos also needs to learn more about late-decision makers, undecided voters and how they may impact voting results, he said.

Meantime, Steve Mossop, president of Insights West, said polling companies need to create a model that will help them predict who’s actually going to show up at the ballot box.

“Because right now in all the polling that’s done, you’ll find with any pollster, that probably 85 to 90 per cent of people we speak to in an election campaign will tell us that they’re going to show up at the poll, and we know better it’s only 50 per cent,” said Mossop.

Companies also need to engage younger citizens, specifically in the 18-34 age group, who are traditionally underrepresented in the polling but are also the most difficult group to reach and engage.

“We need to do a better job in figuring out how to attract or retain and speak to that core audience,” he said.

Insights West plans to work on incorporating mobile devices like tablets and texting into their methods, allowing them to engage a larger audience, he added.

Both companies indicated they polled online, emailing a panel of Canadian households who have agreed to participate in email surveys and are reflective of the population geographically and demographically.

Story continues below advertisement

Both companies also say they have experienced some form of backlash.

Mossop said some of his clients have questioned the use of online polling.

Braid said the backlash has not come from clients but from the public and media who are now skeptical of polls.

But he also said those who spent more money on their polls, specifically the Liberals, got more accurate results.

Braid said he understands the party polled daily by telephone in swing ridings, allowing officials to talk to a random sample over time, and he believes they also conducted focus groups.

While Braid doesn’t know what it cost the party, he imagines the bill won’t be small.

“It is an indication that if you spend more money, you can get better results,” he said.

Sponsored content

AdChoices