Cellphones make political polling tricky

The rise of the mobile phone is casting a shadow over the reliability of traditional telephone polling.

There is no directory of mobile phone numbers, and more and more young people are opting to go without a landline. So are their opinions being overlooked?

Bearing in mind most political polls are conducted via home phones, 3 News asked 97 18- to 26-year-olds across Auckland, Wellington and Christchurch whether they have one where they live.

Sixty-one percent said no, their household didn't have a landline, and 39 percent said yes, they did.

"It's a big issue that needs to be addressed but also it's one with no easy solutions," says UMR pollster Gavin White.

Mr White says the problem shouldn't be overstated. Historically, the groups less likely to vote are the young, the poor and the less educated.

"If we assume the people who are going to cellphones only are predominantly young, then we have to ask would they cross the hurdle anyway and vote in an actual election."

In fact, he says it is not just young people who are rejecting landlines. The latest census data shows 86 percent of households have a landline, down from 92 percent in 2006.

That means 14 percent of households don't have a landline and because there is no directory of mobile phone numbers those people are essentially off the grid to pollsters.

So why is there no directory for mobile phones?

Telecom says two-thirds of its customers have prepaid phones and it is hard to keep accurate records of who owns or uses one.

Vodafone says customers don't want a directory. They view their numbers as more private. There's no information about prepaid customers and there are portability issues.

"Realistically I think the longer-term solution is going to be online," says Mr White.

But Mr White says pollsters need to get a representative sample and UMR keeps making calls until they do. It's all very technical, but in essence the pollsters adjust their results to take into account the fact they can't reach mobile-only households.

"We use quotas and weights, which take everything back to census," says Mr White.

The political polling code says political polling should have a sample of 500. In our, frankly unscientific, poll of 97, we found 17 percent of young people said they intend to vote National, 14 percent Labour, 12 percent Greens, and the rest, a whopping 57 percent, didn't know.

Apart from those "don't knows" we found, Mr White says young people are hard to survey but not just because many don't have landlines; it's also because they don't like being surveyed.

3 News

source: newshub archive