On being … surprised?

By Ingrid Sapona

We just had a provincial election in Ontario and come October we’ll have a mayoral election here in Toronto. (You may have heard of our mayor – he’s kind of famous, er, infamous. But that’s a whole other story and since I can’t see myself writing  a column called On being … a crack cocaine user – nor can I imagine any of my readers relating to that topic – rest assured, you’ll see no further mention of our mayor here.)

The provincial election was called when the Liberal Party, which had governed since 2011 as a minority, failed to get support from the opposition for the 2014 budget. As you might expect, in the 35-day run-up to the election there was a heck of a lot of political polling going on. I got at least three calls myself, asking who I was likely to vote for. (Yes, you read that right – Ontario provincial election campaigns are limited to five to six weeks – eat your heart out American readers!) On the eve of the election, the polls indicated the race was going to be close. Well, it wasn’t. The Liberals won a solid majority.

The misleading pre-election polling here in Ontario was noteworthy locally, but it paled by comparison to the erroneous pre-election poll predictions elsewhere. In the premier race in British Columbia last summer, for example, pollsters where shocked when the party that was supposed to win by a landslide lost – by a lot! And of course, just this week there was the “surprise” loss by Eric Cantor in the Virginia Republican primary, despite the fact that his internal polls had him ahead by over 30 percentage points.

While I can understand that politicians are anxious to get some indication of how much support they might get in an election, surely they must realize the risks involved in taking the pre-election polls to heart. Sure, pollsters take a lot of factors into account, adjusting for all sorts of demographic factors, like age, wealth, education, location, and so on. And of course, all polls have that asterisk that provides the mysterious margin of error number, which is often so broad as to make the spread between the candidates virtually zero, but never mind.

If you ask me, if there are any lessons to be learned from the numerous pre-election polling misreads, it’s that candidates who put any stock in them do so at their own risk. If polls show you’re a shoo-in (like they did with Cantor), or even just in the lead, you run the risk of being complacent. And if you’re behind in the polls, you run the risk of feeling downtrodden, which makes it even harder to put your best foot forward, and I imagine you run the risk of people being reluctant to support someone others think will lose.

Personally, I love answering political polls, especially the super-efficient automated ones. I don’t mind saying what candidate I will vote for, and I always tell the truth. The way I see it, if the poll is being paid for by the candidate (or party) I support, I’m happy to have them know that they have my support. And, if the poll is sponsored by a party or group that I don’t like, I don’t mind taking a bit of the wind out of their sails. Mind you, my willingness to even respond to such polls is founded on the fact that, unlike many other choices that I have a hard time making, when it comes to politics, I’m never a fence sitter.

But, I think the problem with pre-election polls has to do with a fundamental difference between responding to such a poll and casting a ballot. When someone responds in a pre-election poll, they do so knowing they have nothing truly at stake. Candidates and the pollsters they hire seem to lose sight of this fact. Of course, I guess that’s human nature too, as they, arguably, do have something at stake with such polls. So, until pollsters figure out a way to factor in some of the whims of human nature, I’ll find it hard to take polls too seriously.

Anyway, that’s how I’d explain why pre-election polling has let to so many surprises. What’s your take on it? (I’m just asking…)

© 2014 Ingrid Sapona