Mainstreet identifies error in polling that failed to predict Nenshi victory

A pollster that repeatedly predicted a Bill Smith win before Calgary’s municipal election, released findings of a review Monday into why it failed to point to the real winner.

Mainstreet Research’s investigation was led by VP of Analytics Dr. Joseph Angolano and vetted by a panel of independent pollsters. He said the errors could be blamed on the “Calgary Effect” which in several cases led to almost a 25 per cent deviation between pre-election polls and what happened on election day.

“The ‘perfect storm’ of as he calls it of factors that led to the error, but all of them were made worse, amplified, by the fact that the ratio of young people in Calgary to seniors is so dramatically different than the rest of Canada,” said President Quito Maggi.

The review also said failure to poll in non-official languages and a tendency for Mayor Naheed Nenshi voters to not participate in Mainstreet’s polls were also factors. In the last of the Post Media commissioned polls, Mainstreet called for a 13-point victory for Smith, but Nenshi ended up winning by about seven points. The firm defended its figures leading up to election day and Maggi and other staff sparred with political scientists and critics online. The arguments also prompted Mainstreet to review its communications strategy.

“And we took it personally. I was hurt at some of the accusations that got quite outlandish, (and) engaged on social media that was completely inappropriate. I completely take responsibility for that,” Maggi said.

Mount Royal University’s Duane Bratt pointed to errors in the pollster’s methodology early on and did not accept an apology from Maggi after the election.