Wednesday, 5 September 2007

Outremont by-election: The Problems of Polling in a Cell Phone World

Nobody’s saying out loud if any of the political parties have done polls in the Outremont by-election, although the campaign workers for the four major candidates have been busy canvassing by telephone. If you ask them how they’re doing they’ll all smile and tell you that it's going very well.

That’s part of the campaigning game, but it may also be that finding out how people intend to vote in elections these days is a lot harder than it was 10 or 15 years ago. The difference is the increasing number of people—particularly the young—who don’t have land lines and so aren’t listed with directory assistance. That removes them from the polling universe, in effect. How this affects polling results is an interesting question.

One large study in the US recently found that those people who only have cell phones aren’t that different from people who have land lines. The study’s conclusion was that probably the differences can be factored in when massaging the raw poll data.

One way of overcoming the unavailability of cell phone numbers would be to call various permutations of possible numbers in the prefixes assigned to cell phones. Getting computers to generate and call numbers is not legal in the US, but apparently would be allowed in Canada. Whether it would be well-received is another matter. One campaign worker I know turned absolutely purple when I suggested the possibility. “I’d be furious if someone got my cell number that way,” he said. “it would be completely counterproductive.”

Well, maybe. But there are those who would argue that calling anyone at 6:30 p.m. is counterproductive in at least half the cases. Parties do it anyway, and in fact couldn’t run campaigns without telephone canvasses

No comments: