Skip to main content

Lies of a Cell

The importance of including cell-only households in phone surveys continues to grow alongside the difficulty of getting accurate results if you don't.
  • Author:
  • Updated:

Pollsters have known for some time that the 40-year-old process of calling someone's home phone and quizzing them about Candidate X or Issue Y has a small but growing bug in reflecting reality — some people, and more importantly, some definable groups, have abandoned landlines for cell phones. (Of course there were always people with no phones not getting polled, but that's another story.)

In 2007, the Pew Research Center and The Associated Press teamed up to ask what was missing from national surveys, and answered it was cell phone-only users. At that time 12.8 percent of U.S. households were cell phone only. "Yet while the noncoverage problem is currently not damaging estimates for the entire population," the authors wrote at the time, "we find evidence that it does create biased estimates on certain variables for young adults, 25 percent of whom are cell-only according to the most recent government estimate."

About a year and a half later, in the run-up to the 2008 presidential election, Miller-McCune weighed in on the issue (and used even more recent Pew statistics to do so).

"The demographic of the cell-phone-only crowd is largely young, ethnic and lower income," our Joan Melcher wrote at the time. "Indeed, a paper released at the beginning of the year by Macro International, a research, management and marketing firm based in Washington, D.C., noted that "certain populations, such as 18- to 34-year-old adults, have all but disappeared in land-line surveys."

At that time, a month-old Nielsen survey was cited noting that 17 percent of U.S. households had gone solely cellular.

A few days ago, Pew again took a look at the question of cell use polluting polling validity and found pretty much nothing has changed, except that a quarter of American households are now cell-only. In their "Assessing the Cell Phone Challenge," the researchers determined that not only is "non-coverage bias" routinely corrupting telephone surveys, but for some populations it's making getting good numbers almost impossible to obtain.

As Pew observed, "Respondents ages 18-29 now constitute just 7 percent of a typical landline sample, less than one-third of their proper proportion in the population," and this gulf will widen as people retain their landline-less ways as they mature. And it's not just kids — Hispanics and renters are also cell-dependent. If that weren't enough, the universe of people with landlines ruffles the seas of homogeneity in other ways — landliners own more guns and desktop computers, while their untethered peers are more liberal. (Some 53 percent favor legalizing pot compared to 37 percent of landline-only users. The actual paper includes lots more fun tidbits on demographic, behavior and viewpoint differences. )

So, just plug in some mobile phone users - an expensive and fraught undertaking as it is — and all is hunky-dory?

Well, no. It seems cell-only users are more likely to answer, and their greater cooperation, rather than being an automatic positive for the polling process, makes analysis that much more fiddly.

So even if pollsters spring for the extra cost of including mobile callers, don't expect more accurate results any time soon.