I had a piece in a lot of papers today that reported on the latest results of our Ipsos Reid poll. Ipsos has the Conservatives at 40 per cent and has the Liberals at 23 per cent. (My friend Tonda M. reports on a poll today her paper commissioned that has the Liberals doing even worse: tied with the NDP at 21 per cent)
A grumpy Facebook friend wrote in to say:
“Why isn't anybody doing a story about how wrong the polls were last election, and how seriously flawed they may well be again?”
This complaint comes up a lot and is often followed by accusations from the misinformed that report on polls as a replacement for “old-fashioned reporting.”
Personally, I've written probably 12-15 'old-fashioned' election-related stories over the last two weeks, of which precisely one was a poll story.
Canwest isn't unique: There's a helluva lot of reporting on the issues but at some point we have to rely on an electorate that actually wants to learn about the issues and will seek out reportage on a variety of topics. If all you read are the poll stories and you ignore others, what can the MSM do about that?
Poll stories, it seems to me, are a perfectly legitimate complement to overall election coverage. Political parties have sophisticated overnight polls using large population samples which they would never in a million years share with the public. Why shouldn't media organizations hire pollsters so that we can report on what backrooms of each political organization know but won't tell us?
But back to my grumpy Facebook friend. As I messaged him, in 2006, the pollsters weren't that far off. Nik Nanos, whose firm was then known as SES Research, got it pretty much spot on with a poll that finished its work on Jan. 22, four days before E-Day. But other pollsters were close enough on the main issue, that Stephen Harper was going to win a minority government.