CBC Ombudsman Says Focus Should Be on Issues, Not Polls During Elections

The CBC logo is projected onto a screen in Toronto on May 29, 2019. The Canadian Press/Tijana Martin

The CBC’s ombudsman says he doesn’t like media-commissioned election polls, but the public broadcaster didn’t violate policy by releasing its own election poll three days before the 2023 Alberta election.

Ombudsman Jack Nagler made the comments in response to a viewer’s complaint over a May 2023 CBC article about a CBC poll showing the United Conservative Party (UCP) of Alberta in the lead.  The article was published with the headline “CBC News poll suggests United Conservative Party headed for victory in Alberta.”

Complainant John Bozocea argued that publishing the article three days before the election “circumvented the democratic process.”

The poll of “only 1200 Albertans by a biased and subjective pollster” was paid for by Canadian taxpayers and declared “falsely and indirectly” that the UCP would win a majority, he said in the complaint.

The UCP, led by Premier Danielle Smith, was re-elected with a reduced majority.

In his review of the complaint, Nagler added a personal note, saying he does not like media-commissioned polls and news organizations should spend more time on election issues than on the political “horse race.”

Nagler said his dislike of media polls is not because they are anti-democratic but because they take away from more important elements during an election period.

”I prefer news organizations to devote less time to the horse race, and more time to the issues in a campaign - which can range from the policy options on offer to the personal attributes of the candidates,” he said.

Nagler also laid out the CBC policy on election polling, saying CBC’s Journalistic Standards and Practices recognizes polling as a legitimate aspect of political journalism.

The policy says CBC commissions surveys to reflect Canadian opinion on important issues of the day. Journalists are required to provide relevant poll information such as sample size, methodology and, if possible, margin of error.

Because polls can affect the outcome of an election, said Nagler, approval of the CBC editor in chief in consultation with the research department is required.

“I did not see anything in this instance which was out of the norms when reporting on elections,” Nagler said.

Contextual information was included in the article about the poll results, he said, making clear that the election results were not predetermined.

“It is fair for you to disagree that this poll was constructive. However, there was no violation of policy,” Nagler wrote.