By the time you read this column, I'll be settling into my new home. I'm going a short distance geographically, but a significant distance financially (from renting to owning).
I will be trimming one expense when I move: I'm not getting a new landline phone.
This could have huge implications for the next provincial election.
No, not my personal phone use. Just the fact that I'm representative of a broad movement of unplugging. Why should I, and my girlfriend, pay for three phones?
Why should we pay for a phone that plugs into the wall and can't go anywhere with us, and which seems to mostly attract calls from telemarketers?
Oh, and lately, we've been getting a lot of calls from pollsters.
Aye, there's the rub. Remember last year, when the Wildrose Party was poised to upset the long-reigning Alberta Conservatives? Every poll showed Wildrose with a lead - bar a few at the end that showed the two parties neck and neck. And what happened? Wildrose got walloped by 10 points.
The pollsters sort of shuffled their feet and looked sheepish and tried to shift the blame onto voters, claiming that tens of thousands of Albertans had suddenly changed their minds in a day or two.
Not likely. A few analysts did point to the cellphone conundrum, the issue that pollsters have no good, reliable way of reaching folks without a landline.
I looked it up. Guess which province has the highest number of households with a cell-phone? According to StatsCanada, in 2010, Alberta was leading the pack, at 87 per cent.
Number two was Saskatchewan (83 per cent), number three was B.C. (82 per cent).
Nationwide, among households in the 18-34 age bracket, 50 per cent didn't have a landline at all, up from 34 per cent from just two years earlier!
So my hypothesis is this: polls are getting increasingly inaccurate, but they're getting inaccurate in predictable ways.
Look at your poll. Assume that you have over-polled certain groups, and under-polled certain others.
Which makes this coming May election an interesting test for my hypothesis. We'll get to see if I (might) be right about this.
Here's the idea: young people lean a bit to the left, older people a bit more to the right.
Urbanites lean centre-left, rural folk to the centre-right. The NDP and Greens will attract more younger and urban voters, the Liberals and Conservatives older and rural voters.
Who has dropped their landlines? Overwhelmingly, it's the young, and the urban.
Therefore, after election day, we should find that the Tories and Liberals were polled at slightly higher than the rates we actually see at the voting booth, and that the discrepancy was worse in urban ridings (with heavy cellphone coverage) than in rural ones (with spotty coverage).
Of course, right now, this is just a hypothesis. I might be wrong. Maybe not enough young people vote to make a difference.
Maybe any slip at the polls will be within the margin of error.
Or, maybe modern polling is now entirely useless.
There are methods of polling other than the phone questionnaire. And if I'm even partially right, every polling firm is going to have to start dusting off those techniques.
@ Copyright 2013