Skip to content

This is the Life: Time to shove the pre-election polls aside

If we learn any lesson from last week’s election, it’s that pollsters can no longer be considered a reliable source of information...

If we learn any lesson from last week’s election, and from the one in Alberta last year, it’s that pollsters can no longer be considered a reliable source of information. So it’s time for the media to stop treating poll results as news. They aren’t, any more than your drunk uncle who spends his afternoon in the pub talking about politics to anyone who will listen.

I’m not as obsessed with politics at the provincial and national levels as I once was, but I’m still a curious guy. I heard the consistent and unwavering message about polls putting the NDP ahead — a day before the election I saw a chart detailing the results of the last 20 or so major polls. Every single one had the NDP in the lead, never by less than about five per cent.

It’s unlikely that there is a single reason why polling has become an inaccurate reflection of the true intentions of the public, any more than there is only one reason why a pathetic 52 per cent of registered voters made the effort to cast their ballots. Maybe, like me, people are fed up in general with the bombardment of telephone calls wanting to know my opinions on any number of topics. Unlike many others, I have an easy out — I identify myself as a member of the news media and I am immediately disqualified, for reasons I am not entirely clear about. Not that I’m about to question them.

So it is entirely possible that voters have just started to lie as a way of fighting back. It is also possible that people who had stated their intention to vote NDP got cold feet as the reality of an actual NDP government became more likely. It’s one thing to elect a scrappy defend-the-ordinary-joe-and-the-environment party so that the governing Liberals have a strong opposition, but it’s an entirely different kettle of fish when it comes to giving them the keys to the legislature.

More likely, I think, is that the statistical calculations that are made to gather opinions of a few and extrapolate them to extend to the entire population are no longer valid. In earlier days, polls were conducted by real people going door to door. They were much more likely to get truthful responses from face-to-face interviews than they are now that technology allows the entire poll to be conducted using a recorded voice and getting responses punched into a telephone keypad. Conducting a poll robotically is no better a way of collecting information than listening to radio call-in shows — the usual suspects provide their input and the majority simply don’t participate.

Ancient is the argument that people can be swayed by what they hear from opinion polls. Some like to jump on bandwagons and some can hardly wait to jump off. Some might not bother to vote because their party of choice is ahead and they don’t feel the need to make the effort. On the other end of the scale are those who hear their part is lagging behind and think their vote won’t make a difference. When very nearly half the people who are entitled to vote don’t, I think it is incumbent on everyone involved to wonder why, and to make changes to improve that statistic. If democracy works because large groups make better decisions than small ones, surely we would be better off if 90 per cent of those eligible actually voted.

In our small newspaper, we don’t work to sway people politically. Our election coverage was scrupulously balanced because we think our job is to be a conduit between candidates and voters. With our precious editorial space we aren’t tempted bring polling results into our coverage. But if we were, I would be promising our readers that the practice would no longer be tolerated.

As a follower of the news, in the future I am going to be leery about news sources that continue to report on polls as though they are somehow meaningful. They aren’t. It’s time we all admitted it.

Lorne Eckersley is the publisher of the Creston Valley Advance.