I don’t disagree that it’s very soon to be using the polls as predictors of anything, but that isn’t quite the same thing as saying that they don’t matter at all. Obviously they matter, and if you’re a Liberal you must be looking at these numbers with alarm. The most recent data does suggest a CPC majority if an election were held today.
On the Alberta election, I actually don’t know if we can really say the polls were all that far off. By the end most were showing a UPC majority, and that’s what we got albeit a very narrow one. We certainly didn’t see an NDP landslide but speaking personally I didn’t really expect one. I try not to make too many predictions (I’m mostly wrong…) but my recollection is that before the election we all figured it would be a close race, and it was. There might have been some outliers in the polls but I think that’s where the polling average would lead you.
In this case, even if the CPC manages to lose again, it won’t be because these polling numbers were wrong—it will be something to do with events between now and then causing voter preferences to shift. Sometimes preferences are a bit less “sticky” when an election is farther away, and that’s what can make polls this far out from an election a little less reliable in the long term. But they absolutely still matter.
|