Saturday, July 5, 2014

Public opinion polling in Canada descends into madness, so let's ignore polls and just follow our guts, please!


Something really bad has happened to the quality of public opinion polls in Canada in recent years, particularly those trying to gauge voter opinions during election campaigns. Recent polling disasters - in which the final results bore little resemblance to most final campaign polls - confirm we can no longer rely on polls to tell us much of anything. 

It's true that polls, only when grouped together by aggregate websites like ThreeHundredEight.com, can pick up on general trends.   But even polling aggregates are being undermined by the shoddy work being done out there nowadays.

In 2011, most polls accurately predicted the surge of the NDP into second place, even though almost all of the final campaign polls underestimated Tory strength.  We thought we were looking at a close fight between the NDP and the Tories, but in the end Stephen Harper easily won his first majority government.   Polls in British Columbia in 2013 and Alberta in 2012 proved to be even worse.

The same trend occurred in the recent Ontario election where polls were all over the map.  It was embarrassing.  Some pollsters consistently underestimated NDP strength throughout the campaign, while others overestimated Tory strength in their so-called 'Likely Voters' polls.   

The worst was CTV-sponsored pollster Ipsos-Reid whose polls suggested that Tim Hudak was catching on with the public, even after his disastrous pledge to cut 100,000 public service jobs.  Those of us who were talking to actual people knew differently throughout.  The final Ipsos-Reid joke was on the eve of election day showing the three parties virtually tied.  But Ipsos-Reid's "Likely Voters" formula actually showed the PCs six points ahead of the Liberals and NDP, both allegedly tied at 30%.  

As we know, this final poll by Ipsos-Reid was total B.S.  But most disturbingly, it fit a pattern throughout the campaign for the CTV-paid pollster by overestimating Tory and NDP strength, almost as if the pollster was simply trying to play Tory-NDP cheerleader.  The fact Ipsos-Reid polls were being financed by CTV-Bell-Globemedia speaks volumes.  That private media conglomerate has showed its pro-Tory biases for years in subtle ways.  When the Globe & Mail editorial board decided to call for a Liberal minority government in Ontario, Globe owners apparently overruled their staff and ordered a bizarre Tory minority endorsement instead. 
 
There is hope though as this private media manipulation during the campaign seemed to have ZERO impact on the public, which still voted the way it wanted.   Based on shoddy polls, 308.com predicted a Liberal minority and a tight vote between the Grits and PCs, with the NDP way back.  Most media predicted the same.  Thus, when the Grits won handily due to a PC collapse, most described the results as "shocking."

I, too, was surprised by the Tory collapse.  None of the pollsters had predicted it.  Although my gut instinct that the Liberals would win a majority turned out to be true (although I do admit I had thought that would come about due to a NDP collapse, not a Tory one.) 

Today, Forum says that 35% want NDP Leader Andrea Horwath to step down.  My verdict: this poll is wrong and is of no relevance.  We have no idea what the public wants Horwath to do.   I would say Horwath should ignore the polls and listen to her gut, as should other New Democrats when determining what to do going forward.

It also doesn't bode well for the upcoming Toronto municipal race.  The only pollster doing public polls thus far is Forum, which has shown strange Rob Ford strength, even after being shipped off to cottage country rehab.  I don't believe those Forum polls for a second.

I'm going to start ignoring Forum completely and most other polls and resume listening to my gut about how things are going.  Because my gut has been surprisingly accurate of late.  And my gut tells me a big majority of Toronto voters are angry with Rob Ford and are not considering giving him another chance.  My gut is telling me Ford is heading for a humiliation at the polls in November.  The only question that remains is who will end up on top: Olivia Chow or John Tory.   Thus far, my gut tells me it's going to be Olivia Chow. 

The only pollster with any remaining credibility would be Nanos.  Their one Ontario poll released on May 26, over two weeks before voting day, proved to be the most accurate predicting the final outcome. 

No comments: