More on Mobile Phones

Legacy blog posts Mobile phones Sampling Issues

I want to clarify a few points from Friday’s post on mobile phones, based on some of the very thoughtful reader comments. 

First, it is tempting to try to project a growth rate for mobile-phone-only adults based on the three different studies I cited, but keep in mind that each used different methodologies and asked different questions.  The estimates are not strictly comparable with each other.  We do have pretty good estimates of the size of mobile-phone-only population as of last spring and clear evidence that the population is growing.  So it is safe to assume the mobile-phone-only population has grown since the spring, but we do not know precisely how much. 

Second, keep in mind that we really cannot know for certain (a) how many mobile-phone-only adults will vote and (b) who they will vote for.  My point was that the available demographic estimates of the mobile-only population suggest a Democratic orientation, but the same demographic pattern also suggests they are historically low turnout voters.  Even if we assume comparable turnout and a plausible pro-Kerry margin, their relatively small size still implies a very small effect.  Of course, this inference is a matter of opinion.  Yours may differ.  Strictly speaking, we do not know for sure. 

Third, many agreed with commenter Haim Goldman, who asked, "what about people who have both home landlines and cell phones but rarely answer the landline?" 

Good question.  We do have some recent data on adults with both mobile and landline phones.  It comes from another study reported earlier this year based data from large in-person surveys (approximately 2000 interviews per year) conducted over the last three years by the Roper ASW:

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

RoperASW Data* – Percentage of All Adults
2001 2002 2003
Percentage who own a cell phone 50% 55% 60%
Use cell phone to make…
…ALL calls 3.0% 3.9% 6.6%
…three quarters of calls 6.0% 7.7% 9.7%
Uses
  only cell phones to make
 and receive calls na 1.9% 2.5%

The last category (those who use only cell phones to make and receive calls) shows a smaller number of mobile-phone-only adults (2.5%) than suggested by the number of households in the same category reported at about the same time by the Current Population Study (6.0%).  The difference is likely a function of different question language.  The Roper number (2.5%) is a byproduct of a question that asked respondents about the likelihood of abandoning their landline phones.  On that question, another 4% of all adults said it was either "almost certain" or "very likely" that they would give up regular phone service. 

However, the data in the middle rows are most useful to our discussion.  The 6.6% who said they use the cell phone to make "all" calls is close to the CDC estimate of cell phone only adults.  The main point is that the Roper data show that another 9.7% who say they use their cell phones to make most of their calls.

What is the difference?  A household without a landline is totally out of reach of conventional telephone surveys.  A household with a rarely used landline is not.  In terms of survey methodology the first is an issue of "coverage," the second of  "non-response."  There is little a pollster can do about a coverage problem, but there are strategies for reducing non-response.  I will discuss this at more length in a few days, but it is mostly about persistence, calling respondents several times over several days to get those who are more difficult to reach, whether it is because they seldom use the landline, are rarely home, sometimes screen their calls with an answering machine or caller ID, or are simply on the phone a lot (yielding a busy signal).

Don’t get me wrong:  The non-response issue (which also includes those who simply hang up) is definitely a growing problem for polling, and a potentially much bigger problem this year than the relatively small number cell-phone only households.  Persistence is not a magic answer.  Non-response just a different category of problem, and one that I will take up in a few days.

*Source:  Peter Tuckel and Harry O’Neill, "Ownership and Usage Patterns of Cell Phones:  2000-2004.  Paper Presented at the Annual Meeting of the American Association for Public Opinion Research, Phoenix, Arizona, May 13-16, 2004

Mark Blumenthal

Mark Blumenthal is the principal at MysteryPollster, LLC. With decades of experience in polling using traditional and innovative online methods, he is uniquely positioned to advise survey researchers, progressive organizations and candidates and the public at-large on how to adapt to polling’s ongoing reinvention. He was previously head of election polling at SurveyMonkey, senior polling editor for The Huffington Post, co-founder of Pollster.com and a long-time campaign consultant who conducted and analyzed political polls and focus groups for Democratic party candidates.