A Canadian Press piece that ran in several places today, including the Calgary Sun, made the claim that “experts of all stripes say there is little evidence to suggest any upsets will occur in this election” in Alberta. Bad news for the progressives in Edmonton-Strathcona, right? Not so fast.

First, let’s have a look at who these “experts of all stripes” are. The Canadian Press interviewed the following three people for the article:

  • David Taras, a professor at the University of Calgary
  • Conservative MP Peter Goldring (Edmonton East)
  • Conservative MP Rahim Jaffer (Edmonton-Strathcona)

The latter two “experts of all stripes” are easily dismissed. Of course Peter Goldring and especially Rahim Jaffer want to downplay any fears the Tories might have of losing Edmonton-Strathcona–any excitement among progressives that they could actually unseat a Tory increases the likelihood of a united progressive front.

But what about this David Taras fellow? He’s a professor at the University of Calgary–that means he knows what he’s talking about, right? Well, he may well know what he’s talking about in general, but he’s also shown an unfortunate propensity in the past for giving interviews to the media that have less to do with available facts and more to do with whatever David Taras feels like spouting off about at any given moment. He is, after all, the same David Taras who claimed with a straight face that an iconic phrase from the Canadian national anthem was a) in the American national anthem and b) printed on U.S. license plates.

So since those “expert opinions” are problematic, let’s have a look at the actual evidence. First, there’s the data from past elections that suggests that if trends continue the way they have in 2004 and 2006, NDP candidate Linda Duncan can actually win handily. And as for the current election, we don’t even have to look beyond democraticSPACE itself to come up with far better data on what’s going on in Edmonton-Strathcona than the mere speculation these “experts” provide. The current democraticSPACE projection, based on a mathematical model that translates regional polling data into seat projections, puts the Conservatives at 35-37% and the NDP at 33-35%. By Greg Morrow’s own estimation, this makes Edmonton-Strathcona the only Alberta riding that’s “too close to call.” For that matter, the ordinarily notoriously accurate democraticSPACE actually underestimated the NDP vote in Edmonton-Strathcona last election by six points. Assuming Morrow is using the same model to calculate his projections this time, there’s every likelihood he’s off by the same amount again, and that would actually put Duncan ahead.

Please note that I’m not saying Duncan will win. In fact, I put a lot of faith in the democraticSPACE projections, and based on the current one, if I had to bet a million dollars one way or the other, I’d bet against her. But I’m also awfully glad I don’t have to bet a million dollars, because if you look at the actual data rather than just talking through your hat, every bit of it suggests that she could win. And for that matter, if the NDP’s numbers continue their current upward trend across the country, this “Conservative win likely but too close to call” riding will quickly turn to “NDP win likely but too close to call.”

If the Canadian Press really wanted to claim that “experts of all stripes” had told them that there is “little evidence” to suggest any Alberta ridings could go anything but Tory, they should have actually a) spoken to experts of all stripes, and b) examined the available evidence.site stats