Researching the Cell-Phone-Only Crowd: As Some Give Up Their Land-Line, Radio Researchers Must Adjust

by Rob Farbman, Senior Vice President

The Wall Street Journal weighed in on the cell phone only topic this past week with an article titled “Cutting the phone cord isn’t as popular as once predicted”. Quoting data from Forrester Research and others, the piece suggests that going cell-phone-only may not be an unstoppable trend. Some highlights:

  • While 820,000 people have made the move to cell-phone only, The Journal states, “This is only a fraction of what was predicted”.
  • While a previous Forester study of consumer intentions suggested that the number of cell-phone-only users would have doubled in the past year, the number actually grew by a comparatively low 25%. This suggests that many who had found the idea of going wireless-only appealing never followed through in dropping their landline.
  • Reasons cited for sticking with a landline include: Poor signal coverage for cell phones, the fact that many use a landline for their internet access, lack of directories for cell phone numbers, and potential problems with cell phone service during a crisis where cellular networks become overloaded.
  • The article also notes that some consumers are “reconsidering their decision to go wireless and are reconnecting to a landline”. The article mentions several reasons for cell-only people going back to a landline including signal quality and the problem of running out of allotted minutes.
  • The Journal also quotes a major landline phone provider as reporting an increase in primary consumer landlines for the first time in five years, further suggesting a slowing in momentum for the cell-phone only trend.
  • Many service providers are bundling landline and wireless plans in “one price covers all” packages (that’s sometimes also includes television service). Some consumers are finding that sticking with both a cell phone and a landline is actually less expensive than going cell-phone-only.
  • The Wall Street Journal article also notes that the vast majority of cell-phone only users are younger consumers and that many of these people have actually never had a landline of their own.

While the importance of the cell-phone-only trend among young consumers has been well documented, The Journal article suggests that it is far less likely that the practice will make the inroads into the population as a whole that some had predicted.
Nevertheless, as cell phone usage continues to grow, the effect on survey research and in particular Arbitron sampling methodology has moved to the forefront of the radio industry. Of particular concern is the increasing portion of the population that is giving up landlines altogether and joining the “cell-phone-only” community. It is also an issue of importance for radio researchers like Edison Media Research, and we are striving to stay on top of the issue. There have been a variety of stories in the press on this topic and we want to make sure that everyone has the facts.
According to the latest national data on the topic, between 6% and 7% of the adult population can only be reached on a cell phone. This figure has been reported in national studies by the U.S. Census Bureau, MediaMark and the Edison Media Research/Mitofsky International November 2004 Exit Poll. While growing, these numbers have only recently surpassed the number of people with no phone access at all which has historically been between 3 and 5% of the population. As a result, most telephone surveys are still minimally affected statistically by the exclusion of the cell-phone-only population and often remain the most statistically valid methodology to measure opinion. However, cell-phone-only consumers are predominantly young – Edison’s 2004 Exit Polls showed that 19% of those voters age 18 to 29 were cell-phone-only (as opposed to only 4% of those age 30 and older.) So, the issue of accurately sampling younger respondents is of primary concern.
Given the legal and technological restrictions that survey researchers must currently work within, these cell-phone-only consumers are effectively removing themselves from the household “universe” that telephone surveys sample. Current federal regulations place restrictions on the use of “automatic dialing” devices (a standard instrument of most random digit dialing telephone researchers) to place calls to cell phones. While the regulations were originally meant to restrict telemarketers, the wording has effectively ended up including calls from researchers. This restriction has the effect of making telephone research among cell phone users both cumbersome and expensive.
Of course, even without any legal obstacles, conducting research with cell phone users would remain difficult. There are many sampling issues that make reliable cell phone user surveys very challenging to conduct. These include the fact that many consumers have a cell phone number with an area code unrelated to where they actually live, while others might forward landline calls from one area code to a cell phone of another area code, both of which make it difficult to integrate cell phone respondents with sampling based on households.
Then there are the practical challenges of calling people on their cell phones. Many consumers are on restrictive cell phone service plans where they are reluctant to use up their minutes. How welcome will a research call be when agreeing to participate could literally cost the respondent money? While ideas have been tested including providing “free minute incentives” to respondents, organizing and implementing a respondent payment system like this would be a monumental accomplishment in itself.
There are even potential moral issues involved in interviewing people on their cell phones. We know that people are likely to use their cell phones almost anywhere their daily travels take them. This includes walking in the street or driving in their cars. Should survey researchers use a pre-interview screening question measuring the likelihood of a traffic accident before allowing a respondent to participate in the survey? Also all survey researchers promise each respondent that their answers will remain confidential. That becomes more difficult when the respondent is answering the questions on a cell phone while in a public location such as on a train or in a mall.
These are some of the challenges that have led survey researchers including Arbitron to exclude cell phone interviews for now. Our goal at Edison Media Research is to stay in front of this issue, and to insure the most statistically reliable research possible. But beyond that, it is vital that telephone research match the sampling methodology of Arbitron. So long as the Arbitron “universe” includes only household based samples and not cell phones, we will continue to emulate this methodology for surveys that attempt to replicate radio listenership.
The primary unanswered question for the radio industry is this: “How do cell-phone-only consumers differ from other consumers in their listening habits?” In an interview on the Jacobs Media Web site (, Arbitron statistical guru Dr. Ed Cohen discussed various research studies that Arbitron is carrying out to investigate this very issue. Cohen states that a project slated for the summer will specifically investigate the listening habits of cell only people and compare them with non-cell only respondents
Edison has also conducted research on the characteristics of cell phone only consumers. While our 2004 National Exit Poll did not specifically investigate issues of music and radio, it did offer some fascinating findings on how cell-phone-only voters were and were not different from those who had not given up landlines.
When looking specifically at 18-29 year olds in our Exit poll we saw little differences in most categories based on cell phone usage. The most noteworthy difference we see is by income, with the cell-phone-only voters having significantly lower household incomes. The fact that the primary difference was income suggests that going cell-phone-only is more often a financially influenced decision especially among those 18 to 29 year olds who have “left the nest”:
The data does offer some insight into the often hard-to-understand issues involved in statistical sampling. We cannot assume that the fact that a person decides to give up a landline means they are substantively different than an otherwise similar respondent. The exclusion of cell phone respondents will not lead to the under-representation of younger respondents in Arbitron data (nor in Edison studies) because of the use of representative samples that insure that the demographics of the sample match the population.
So the pertinent question is how are these people different, if at all? Is a 25 year-old cell-phone-only male more likely to listen to an Alternative station than a Hip-Hop station? Are cell-phone-only consumers more likely to have reduced their radio listening time for their iPod? These and many other questions require further research. We eagerly await the release of studies by Arbitron and others on the issue and Edison plans to continue to play a major role investigating the cell-phone-only topic in the future.
The following were used as sources in this article:
“Telephony Challenges: What we know and don’t know about cell phones“, Linda Piekarski, Survey Sampling International
“Household Telephone Service and Usage Patterns in the United States in 2004“, Clyde Tucker, U.S. Census Bureau
“10 Questions with Dr. Ed Cohen”,
“Cell Phones Displace Landlines In Record Numbers, According To Latest Mediamark Research Analysis”, Mediamark Research Inc.
National Election Pool 2004 Exit Poll, conducted by Edison Media Research and Mitofsky International