Pollsters Debate Including Cell Phones In Polls

This story was written by Colin Kavanaugh, Daily Pennsylvanian

For pollsters, cell phone static may be getting in the way of good polling.

Across the country, pollsters attempting to accurately reflect the public's choice for president are facing a big - and unprecedented - problem: cell-phone dominance among youth combined with historic young-voter turnout in the primaries.

"There's a great fear that the traditional methods of polling," which utilize only land lines, are not "representative" of younger voters who tend to only use cell phones, said Richard Johnston, a Penn Political Science professor.

In 2004, just 5 percent of Americans were cell-only users, according to the Gallup Organization. Pollsters believed that group was an accurate sampling of the population at large.

But now, 13 percent of Americans only use a cell phone, according to Gallup, and that number disproportionately includes younger people.

Eric Nielsen, senior director for media strategies at Gallup, said the organization has been watching the impact of youth voters and overall cell-phone-only use this election cycle, and he considers it an important issue.

For its election polling this year, 13 percent Gallup's sample is made up of registered voters who only use cell phones.

Nielsen said he realizes that it remains to be seen how much of the expressed young voter interest will turn into results on election day.

For that reason, not all polling firms agree that cell-phone-only voters should be included in their sampling.

Ann Selzer, president of Selzer & Company, a public-opinion research group famous for its accurate polling in the Iowa caucuses, said there is "very little difference" in demographics between cell-only users and land-line users.

As a result, she is not convinced that cell-only voters' political views trend a particular direction distinct from land-line voters.

Selzer has thus opted not to include cell-only users in her sample. However, she does weight younger voters more heavily to account for increased youth interest in this election.

Selzer said polling firms can no longer use "the past" as a reference point for how to weight younger voters in their models given the historic youth turnout in the primaries.

In her polls, she said, she thinks nothing is lost by not polling cell-only users so long as each age group is weighted accurately.

One thing is certain: With historic youth turnout and the increase in cell-phone use nationally, pollsters have been forced to reexamine their methodology, and construct new ways to accurately represent the population.

"We're all going to have to start involving cell phones" in our polls, Nielsen said. "Cell-phone-only groups are becoming a very different sample, and I'm fairly convinced that it will have to become a part of the method we use."