Political Perspectives is produced by the students and faculty of Carleton University's School of Journalism and Communication, Canada's oldest journalism school.

16th
JUN

Stéphane Dion, Bugs Bunny, and a dead cat

Posted by padams under All, Election 2008, Election 2008 Media commentary, Media Commentary, Political Strategy

Paul Adams

A small, but significant corrective to Andrew Cohen’s column in today’s Ottawa Citizen.

I sympathize with Cohen’s view that CTV, and in particular, Mike Duffy, gave Stéphane Dion shabby treatment when they aired his repeated false starts in an interview just five days before last October’s election.

The Canadian Broadcast Standards Council recently criticized CTV for breaching the industry’s code of conduct for airing the false starts even after it had told Dion that it would not. (CTV has strongly defended its decisions in the incident.)

However, is it also true, as Prof. Cohen argues, that for Dion’s campaign, the broadcast may have been “decisive”? His argument appears to be based on a misreading of the polling data.

“It was late in the campaign and polls suggested the Liberals were gaining on the Conservatives,” he writes. And later: “Last October, the polls suggested the Liberal Party’s ascent stalled after the interview. While we cannot say if Dion’s momentum would have brought his party victory, it isn’t impossible.”

Well, almost anything can happen in an animated cartoon, as Bugs Bunny trenchantly observed. But the idea that Dion was riding some “momentum” that might have carried him to victory until the release of the interviews is mistaken. The evidence of the polls is not what Prof. Cohen suggests.

The daily tracking polls that appeared in the last campaign all used a variation of the “rolling poll” system, whereby a published poll includes three, or in some cases four, nightly samples rolled into a single number. Each day, the oldest day is dropped and a new day is added. That means that the published numbers are to a degree retrospective. A number published Thursday, for example, would normally include samples from Monday to Wednesday, or even Sunday to Wednesday.

In other words, there is a delay between moves in public opinion and their capture in the polls. Moreover, I would say from my general observation that it takes two or three days after that for the media to internalize the news of significant shifts contained in the polls and incorporate it into their narratives of the campaign.

At the time of the CTV release of the Dion interviews, there was certainly a media perception that the Liberals were gaining, but this was old news. In fact, all the tracking polls had begun to show the Liberals dipping once again by then.

Mr. Dion had been fairly effective in the leaders’ debates the previous week. The Liberals did indeed get a bump in the polls in the days afterwards, running through the weekend. However, once the last full week of the campaign began, the Liberals slumped back to the dismal numbers they had suffered mid-campaign and which they carried through to election day.

I remember tearing my hair out the very morning of the day the Dion tapes were aired when I saw a headline on the CBC morning program trumpeting the Liberal revival. The revival had been over for several days, and by that time a careful reading of the polls made that evident, notwithstanding the inevitable delay created by the rolling poll system. The problem was, if I may be allowed a gratuitous comment on the media, that the “tightening race” narrative was a better story, and only reluctantly shucked off.

You can check the daily tracking numbers for yourself here.

In fact what had happened was something different.

The Liberal bump after the debates was a dead-cat bounce which was already over by the time of the infamous interview.

The CTV-Dion incident might have been the last nail in the coffin. It may rightfully be described as a final ignominy.

But decisive, it was not.

Paul Adams teaches journalism at Carleton. He is also executive director of EKOS Research Associates, a polling firm that published a daily-tracking poll in the last campaign.