New Wireless Substitution Estimates Released

May 7, 2009. The National Center for Health Statistics (NCHS) released the latest wireless substitution estimates yesterday, reflecting data from July-December 2008.  This information has been collected in the National Health Interview Survey (NHIS) since 2003 and reported every six months as part of NCHS's Early Release Program.  This is the only nationwide estimate of the prevalence of wireless-only households available for researchers.

The new estimates show that 20.2% of households are now wireless only, meaning that they do not have a traditional landline phone.  This is up from 17.5% in the first half of 2008, and reflects the largest six-month increase in prevalence seen since the NCHS began collecting data.  An additional 1.9% of households have no phone at all (down from 2.5% in the first half of 2008), resulting in 22.1% of households that are not accessible by landline phone.  The report is available here.

The prevalence of wireless substitution is of great interest to survey researchers because so many studies are conducted by phone, yet cell phone samples are typically not included in these studies.  We know the wireless-only population differs from the rest of the population on several factors relating to health and health access; as such findings for studies that exclude this population may suffer bias.  For more information please refer to our issue brief on this topic, "The Impact of Wireless-only Households on State Surveys of Health Insurance Coverage." Click here to download the brief. 

Since 2007 the NCHS also collects information on "wireless mostly" households, defined as those households that have landlines but receive "all or almost all calls" on a cell phone.  This prevalence has increased from 12.6% in January-June 2007 to 15.4% in July-December 2008.  While this increase is not as dramatic as the wireless-only prevalence, the ability to survey this group using traditional phone methodology is in question.

SHADAC researchers are following these trends to stay abreast of the latest information and to develop tools for states to adjust their survey techniques.  Next week several of our staff will attend the annual meeting of the American Association for Public Opinion Research (AAPOR) to share what we know and learn from other survey researchers and the pollsters.  We'll blog about the conference and post presentations by Mike Davern and Kathleen Call on our site on May 19.