The Tyee

Angus Reid poll: BC NDP up by 20 points

Tom Barrett2 Comments

There’s a new poll from Angus Reid Public Opinion today that suggests the B.C. New Democrats are 20 points up on the governing Liberals.

The poll has the NDP with 48 per cent support of decided voters, the Liberals at 28 per cent, the Green party at 11 per cent and the B.C. Conservatives up two points at 11 per cent.

All changes from the last Reid poll, taken Feb. 21-22, are within the margin of error for a poll of this sample size.

The poll was conducted Monday and Tuesday of this week. Reid surveyed 809 B.C. adults drawn from an online panel. A sample of this size drawn at random from the general population would have a margin of error of plus or minus 3.5 per cent, 19 times out of 20. (See note below.)

Here’s an updated chart of the polls taken so far this year:

Company Date Lib NDP Cons Green Other Method Sample +/-
Angus Reid 17-18 Jan 31 46 10 10 3 Online 802 3.5
Mustel 11-21 Jan 33 43 11 11 2 Phone 509 4.3
Justason 25 Jan – 1 Feb 26 48 12 11 3 Phone-online 600 4
Ekos 1-10 Feb 27.4 39.0 14.6 13.5 5.5 IVR* 687 3.7
Angus Reid 21-22 Feb 31 47 9 10 3 Online 803 3.5
Ipsos 8-12 Mar 32 51 9 7 1 Online 1,000 3.1
Angus Reid 18-19 Mar 48 28 11 11 2 Online 809 3.5

* Interactive Voice Response

Note: The sampling error margins given here are those provided by the pollster. While online polls have been very successful at predicting recent elections, there is a methodological controversy surrounding the citing of margins of error for online polls. Some experts hold that it is inappropriate to quote a margin of error for an online poll because participants in such polls are drawn from volunteer panels, rather than chosen at random from the general population. For more on this issue, see this story.

(I did a version of this item for The Tyee today that can be found here.)

How tanking turnout makes for ugly elections

Tom BarrettLeave a comment

Wedge issues and fear politics win when voters stay home, pollsters say
By Tom Barrett
TheTyee.ca

Declining turnout actually affects how elections are fought. Photo: Shutterstock.
Declining turnout actually affects how elections are fought. Photo: Shutterstock.

Chances are the next government of B.C. will be chosen by approximately half of all eligible voters. That means that even in a landslide the winning party will probably have the support of no more than a quarter of all the province’s citizens.

Like other Canadians, British Columbians are becoming less and less inclined to vote. In the 2009 provincial election, only 51 per cent of all estimated eligible voters bothered to turn out.

A lot has been written about falling turnout rates and the precise causes of the decline are still largely a mystery. Some blame a dwindling sense of civic duty in young people. Some blame negative, fear-based campaigns.

But there’s another side to it. Declining turnout also affects how elections are fought. Ironically, the more that voters stay home because of politicians’ bad behaviour, the more likely politicians may be to behave badly.

“Turnout is increasingly becoming more important than persuasion in elections,” pollster Greg Lyle, managing director of Innovative Research Group,  said in an interview.

Lyle said that in the old days, when turnout was in the 70 per cent range or higher, campaigns concentrated on wooing ambivalent voters, or what Lyle referred to as “cross-pressured” voters.

“They’re concerned about balanced budgets, but they’re concerned about waiting lists at hospitals and they want to hear a balanced argument,” he said. “They don’t want someone to just focus on one [issue] or the other.”

But as the turnout gets lower and lower, the people who show up to vote tend to be the ones who aren’t cross-pressured — “people that have very consistent points of view, very ideological positions,” said Lyle, who was Gordon Campbell’s campaign director in the 1996 B.C. election. That’s led parties to build their campaigns around strong differences “that will essentially scare their voters into getting up and going to vote,” said Lyle. Continue reading →

How turnout trips up pollsters

Tom BarrettLeave a comment

By Tom Barrett
TheTyee.ca

While declining turnout poses problems for democracy, it also creates practical problems for pollsters.

“If you do a poll that represents 100 per cent of the population, that’s fine,” said Bob Penner, president and CEO of Stratcom.  “But we know 100 per cent of the population doesn’t turn out. Maybe 50 [per cent] turns out. Well, you’ve got to know which 50 per cent that is. Or at least you have to have a pretty good prediction of which 50 per cent that is.

“Because those that turn out usually differ substantially from those that don’t.”

Penner, a veteran campaign pollster, said “the better pollsters” use turnout models that try to predict who’s going to vote. One way to do that is to ask respondents if they’re going to vote. Those answers aren’t always reliable, though.

People who aren’t going to vote may tell pollsters they intend to do so because that’s the socially acceptable answer. Or respondents may intend to vote but then change their minds.

So pollsters can also look at who has tended to vote in recent elections — older people and homeowners are two such groups — and weigh their data accordingly.

Unfortunately for pollsters, the last election won’t be just like the current one, Penner said. “So you kind of have to put it all into the mixer and kind of have to figure it out. And the more times you do it, the better you get at it and the more likely you are to be correct.”

While all good campaign pollsters have some sort of turnout model, most pollsters who make their work public through the media don’t, Penner said. And that affects their accuracy, he argued.

Mario Canseco, vice-president of Angus Reid Public Opinion,  said his firm does have a method for predicting likely voters.

In last year’s U.S. presidential election, the Reid firm noticed that supporters of Republican Mitt Romney were less committed to voting than those of President Barack Obama, he said. The difference meant that the election wasn’t as close as some pundits were predicting.

“Turnout really hasn’t affected any of our election calls,” Canseco said.

BC’s coming election: Myth of the demon vote splitter

Tom Barrett1 Comment

Libs and NDP sow fear of third parties siphoning votes. But political life is complicated.
By Tom Barrett
TheTyee.ca

When push comes to shove, do third parties 'steal' votes from larger parties more ready to govern? Image: Shutterstock.
When push comes to shove, do third parties ‘steal’ votes from larger parties more ready to govern? Image: Shutterstock.

We’re going to hear a lot about third parties and vote-splitting as we approach the May 14 election. That’s because everybody who knows anything about B.C. elections knows one big thing.

As a Young Liberal delegate said during last fall’s annual party convention: “The only time the NDP wins is when the free enterprise vote is fractured.”

The Liberals and Social Credit before them have been saying the same thing for more than half a century. When Martyn Brown was campaign director for the BC Liberals in the 2001, 2005 and 2009 elections, he worked that line like a government mule.

“It is a powerful argument, no doubt,” he has written, “one that I helped elevate to an art form in my long time in B.C. politics. It certainly helped elect Gordon Campbell’s three successive majority governments.”

There are, however, a couple of problems with the argument. For a start, as Brown now concedes, it misses the point by ignoring why people vote for third parties. It’s based on an outdated Cold War mentality. It also ignores how voters shift allegiance in elections. And it oversimplifies history.

As political scientist Norman Ruff wrote after the 1996 election — one of the key events in free enterprise vote-splitting mythology — “there has never been a monolithic free enterprise vote in British Columbia.”

Continue reading →

Ex-Liberal voters: where will they go?

Tom BarrettLeave a comment

By Tom Barrett
TheTyee.ca

It’s still too soon to write the Christy Clark Liberals off, but for some time they’ve been showing the signs of a coalition on the verge of a breakup.

One indication is the number of people who are telling pollsters that they voted Liberal in 2009, but would vote for someone else today.

Angus Reid Public Opinion vice-president Mario Canseco said in an interview that the Liberals have a retention rate of 66 per cent. That means one-third of self-identified former Liberal supporters say they are going to vote for someone else. (See main story.)

“Some of them go to the Greens, some of them go to the Conservatives,” Canseco said. But roughly half of the former Liberals who are going elsewhere are going to the NDP.

“That is the big issue that we are looking into as we get closer to the election,” he said. “If you continue to have that shift of BC Liberal voters for Gordon Campbell in 2009 becoming NDP voters for Adrian Dix in 2013, then it’s going to be very difficult for them to turn the numbers around.”

One ray of hope for the Liberals is their retention rate has increased from a low of 50 per cent in the fall of 2010, when Campbell quit. And the number of former Liberals who say they would switch to the BC Conservatives has dropped.

But not everyone moving away from the Conservatives is going back to the Liberals, Canseco said.

“They’re saying, ‘Maybe I’ll vote, maybe I won’t. Maybe I’ll vote NDP.'”