Tuesday, October 19, 2010

Calgary 2010 Aftermath, Part 1: The polls were not as far off as you think

The aftermath of Calgary's absolutely crazy municipal election has left many wondering what's happened in this race. How did a veteran councillor and a popular media figure become overtaken by a little-known university professor who had a small, though grassroots, team to support him? Vaulting from very far behind in the game to first place is an accomplishment indeed, and it'll be studied by election watchers and pollsters every which way they can.

But, before the final results come out and we can delve into why Naheed Nenshi's rise truly came about, I want to try and dispel a common myth I've found floating around out there: that the polls got Nenshi's position completely wrong, and that is partially because of a bias based on pollsters that go with landline-only poll calls, rather than going after the cell phone using urbane Calgarian that they assume has voted for Mr. Nenshi.

While the issue of landline-based polling is a serious one that faces today's pollsters, it's best left to someone like Nate Silver to explain the nooks and crannies of this issue. In fact, he already has, and I'd recommend you take a look. But, more to the point: did pollsters in Calgary's race really get the result that wrong?

On the face of it, the numbers certainly indicate that yes, the pollsters got it wrong. The most recent poll out had Ric McIver and Barb Higgins tied at 33% each, with Nenshi not far behind at 30%. That's a far cry from the 40-32-26 split we see with the election results, so obviously the polls got it wrong, right?

Not quite. The Leger poll was a poll commissioned for CTV and the Calgary Herald in October, and it showed a radically different picture than what the same polling company showed back in September, when McIver was ahead with 43% of the vote, to Higgins' 28% and Nenshi's 8%. Between mid-September and mid-October, Nenshi experienced a huge rise in support as candidates dropped like flies, with a couple supporting him, and his campaign gained momentum. For any candidate to make a jump like that is amazing, and showed Nenshi had momentum on his side.

Meanwhile, other pollsters showed a similar tale. Ipsos' most recent poll (October 11, but taken between Oct. 5-6) had Higgins with a large lead of 37%, McIver second with 34%, and Nenshi bringing up the rear with 21%. A few days before that, a Return On Insights poll taken between October 1-2, showed Higgins with a lead at 31%, McIver at 28%, and Nenshi at 16%.

Nenshi was clearly on his way up as election day came ever closer, and all polls were showing increases in votes for him as the days went on. More people had a favourable impression of Nenshi, and more people were paying attention. The fact that this was occurring in the middle of advanced polling means that when those vote totals are released, we'll see Nenshi running very competitively with the two front-runners at the time. The more credibility Nenshi had, the more people voted for him, and the more polls took note. So while the poll's topline numbers were generally incorrect, they all reflected Calgary's trend towards Nenshi fairly well.

From this point, I could launch into an argument about the inherent volatility associated with polling, especially in municipal elections, but I'll leave it at this: the polling in Calgary's election showed a horse race between Barb Higgins and Ric McIver because for the majority of the time during this election, because it was. Nenshi's momentum did not come until late in the campaign, which polls did reflect but not to the point where they could predict Nenshi's final numbers and lead, only that he was gaining ground. Polling is at best a blind way to take the public's pulse, and at worst completely fucking inaccurate. However, when you follow the trends of this election, they come pretty close to what the end result ended up being. For that, pollsters this election deserve some kudos.

As this is "Part 1," Part 2 will cover how Nenshi's momentum built up, and where he ended up getting the votes he did. I hope to make this into a 3-part series, with the last containing my conclusions on how exactly Mr. Nenshi became Mayor of Calgary.

2 comments:

  1. Naheed Nenshi's win reflects Calgary's diversity in a big way. He not only used broadcast and social media to his advantage, but he took the time to reach out to multilingual communities in the city.

    Thus, ending the '1980s Stampede Board-style' of winning streak.

    ReplyDelete