Cell Phone Bias

Have you ever wondered how pollsters manage to get an unbiased sample by calling people on their "land lines"? I don't know about you, but I know quite a number of people, especially young ones, who only have cell phones.


Until now, I've been soothed by the pollsters' answers that they oversample the populations who would otherwise be excluded. But now that reassurance appears to be going to hell in a handset.

Mark Blumenthal, writing for the National Journal, published a piece called Dial a Cell, Reach a Dem: Exclusion of Cell-Phone-Only Households Means Pollsters Are Probably Undercounting Obama's Support.

Blumenthal writes, "When some members of the population go uncovered by the sample, a bias will creep into the results to the degree that two conditions are present: The people who are missed have to be significantly different from those not covered, and the missing segment has to be big enough to make a difference."

Up until this election, those two conditions have not been met, but this year, research seems to indicate that they have. Pollsters from the Pew Research Center, Gallup, CBS, New York Times, NBC, Wall Street Journal, Time and AP have done some cell-phone sampling this year, and when they compare those results with their land line results, they find that the cell-phone-only respondents are significantly more likely to be Obama supporters than their land line counterparts.

Blumenthal quotes the Pew Research Center's findings: "In each case, including cell-phone interviews resulted in slightly more support for Obama and slightly less for McCain, a consistent difference of 2 to 3 points in the margin."

This effect was strongest among those under 30: Young people in cell-phone-only households preferred Obama by a 35-percentage-point margin (62 percent to 27 percent), while Obama's lead was only 13 points (52 percent to 39 percent) among young people in households with land lines.

So it's not their age that is most significant about them -- it's their technology adoption itself that appears to correlate with their difference in attitude.

Maybe that couple of points missing from the polls will offset the missing points in the other direction from the Bradley effect. Hmmm.