January 15, 2008
Polling: It's Off The Hook
Have you given up your old landline phone and gone completely wireless, but miss the company of political pollsters calling between the hours of 5 and 9 p.m.? Then Gallup has some good news for you. Gallup's editor in chief, Frank Newport, announced Monday that the nation's oldest polling company began including cell phone users in its surveys as of Jan. 1, 2008. The move means Americans will now be able to voice their support for Duncan Hunter or their disapproval of Congress while driving across town or dining in a crowded restaurant.
In the past, Gallup, along with most pollsters, typically called only traditional landline phones in its research, a practice that has sparked criticism from some who argue that the practice undercounts certain demographics, particularly young people. Now that Gallup has begun polling Americans who have only a cell phone but no landline, the results will be combined with data collected from landline users.
Contacting cell phones presents a whole range of methodological problems: Respondents might be out or otherwise distracted when a pollster calls; they often require rebates to cover the cost of airtime; and it's expensive and time-consuming to compile new databases of cell-phone-only users.
But the decision to start calling mobile phones could eventually pay off by giving pollsters a better idea of where the electorate stands on a given issue or candidate. The impact of Gallup's move is likely to be muted in the immediate future, though, because the number of unplugged phone users is still fairly small. In addition, many polls already weigh their results by age, which helps to cancel out an under-representation of young people.
As Gallup's Newport points out, "Study after study has shown that in general, the effect of excluding from the interview process those who only have cell phones has not seemed to affect the overall marginal results of political studies." He notes that Gallup will "be analyzing the implications of" its recent "shift in methods particularly carefully."
Pollster.com's Mark Blumenthal, who now also lends his polling expertise to National Journal, sees the change as largely symbolic for the time being. "However, with cell-phone-only population growing rapidly this change may be much more important and consequential down the road," he told The Gate via e-mail. "Some speculate that given the current rate of growth, the cell-phone-only population could be close to 25% by next November -- it was estimated at 6% of voters in November 2004."
Blumenthal offered a more thorough examination of the possibility of surveying cell users in his online column last summer.
