I have noticed that about 30 percent of my friends have gone exclusively cell phone in their private lives.
They have “real” phones at work but their home and personal phones are cell phones.
I have never received a cell phone call from a pollster, but have received a number of them on my home phone.
How are pollsters dealing with this change in the population dynamic?
TV
Isn’t there some law against them calling cell phones?? It has something to do with people getting pissed over airtime charges for unwanted calls.
I saw an article on it yesterday, but can’t recall where. Pollsters are not given cell phone numbers and are very concerned about the issue, but don’t know how it will affect anything.
This is a very tough year for pollsters. In addition to the cell phone issue, there’s the question of what are likely voters (previous years’ models may be wrong), how overseas voters will affect things (those in the army are heavily Republican; everyone else is strongly Democratic), the effect of new voting systems and provisional voting. They didn’t do particularly well four years ago, and may do worse this year.
It just occurred to me that most of that 30 percent I know who are cell phone exclusive at home tend to be anti-Bush types.
Hardly a valid poll on my part but if it is mildly accurate it could be an interesting percentage.
TV
I missed a call on my cell phone and called the number back since it was out of state. Turned out to be Gallop, saying that they were sorry they missed me, blah blah.
So pollers do call cell phones.
The people that have switched to only cell phones tend to be in the 18-35 age bracket, which generally tends to lean more democrat. So yes, this may skew the polling results. This is a well-known problem, and people seem to be in a quandary. It was a similar issue (many of the poorer people didn’t own a phone) that led to the whole “Dewey wins” spectacle years ago.
Also, my mom received a survey call on her cell phone (which I answered) about a month ago. I thought it was rather odd.
I thought of this factor during the recent Canadian election campaign (where the polls ended up being very wrong; two major parties were thought to be tied, but the incumbent party did much better than predicted). If cell-only people were uniformly distributed, it wouldn’t be a problem. But people who only have cell phones (myself included) are generally under 35 and tend to be a more liberal group than average.
Poll groups insist that they use representative samples. I would imagine that they would include 18-to-35 people in their sample in the same proportion as in the population, even if they don’t call anyone who only has a cell phone. (However, the 18-to-35s that they do contact may not be representative of the real group, if it doesn’t include cell-only people and people who aren’t home during polling hours.) When the Canadian polls ended up being wrong, the best reason the media could think of is that a lot of people lied, or voted differently than they said they would. The cell-only issue wasn’t raised.
I think the issue with this presidential election is not only that younger voters are more likely than older ones to have only cell phones, but also that these younger voters might turn out to vote in much higher numbers on November 2 than they have in the past.
This would, obviously, screw the pre-election polls up.
But we won’t know for certain until at least November 3.
One of the heads of the major polling companies was interviewed on NPR a couple of weeks ago, and he wasn’t discussing this issue of poll accuracy, but poll accuracy in general, and he stated that even the best conducted polls shouldn’t be taken as gospel.
So while it’s certainly possible that a demographic is being missed, it’s not like the pollsters believe that their poll is going to be 100% accurate.