TPRC45 has ended
Back To Schedule
Saturday, September 9 • 2:33pm - 3:05pm
Understanding the Trend to Mobile-Only for Internet Connections: A Decomposition Analysis

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

Feedback form is now closed.
Household internet access via a mobile-only connection increased from 8.86% in 2011 to 20.00% in 2015, more than doubling in only four years. Understanding the driving factors behind this trend will be important for future iterations of broadband policy. The drivers of the diffusion of broadband access are well documented, but little has been done on the topic of the mobile-only connection. Many demographics have embraced mobile-only connections over this period, including Hispanics, Asians, older Americans, and Americans in non-metro areas. An open question, however, is which relationships are driving this shift to mobile. For example, is the shifting relationship between age and a mobile-only connection a more important driver than that for race or non-metro status?

Data for this paper comes from the Current Population Survey (CPS) supplemental survey on Computer and Internet Use for July 2011, 2013, and 2015. Beginning in July of 2011, the CPS began including a mobile only connection as an explicit option when asking how the household connects to the internet. Combining this information with demographic variables from the survey allows for examination of how the relationships have changed over this four year time period.

To answer this question, the propensity of an individual to adopt the internet through a mobile only connection is modeled using a logistic regression for each of the three years. Following these regressions, a non-linear Blinder-Oaxaca decomposition is used to determine which of the shifting relationships are most responsible for the trend to a mobile only connection over the time period of 2011 to 2015. Changes due to shifts in the demographic characteristics of the population themselves are not likely to be responsible, since the nationally representative CPS data did not change very significantly across the years. The results bear this out, with changing characteristics accounting for less than 1% of the full 11.14% increase in mobile-only adoption. However, behavioral relationships across demographics did change dramatically as specific groups were much more likely to be in the mobile only category in 2015. The leading behavioral relationships impacting the trend are those associated with age (50.55%), race/ethnic background (4.75%), and non-metro status (1.88%), indicating that these demographic groups are becoming more willing to adopt the internet via the mobile only connection. By understanding which demographics are driving the shifts to mobile-only, programs focused on bridging the digital divide (such as Lifeline); can be more focused in their efforts.


David Gabel

Queens College


Brian Whitacre

Professor, Oklahoma State University

Saturday September 9, 2017 2:33pm - 3:05pm EDT
ASLS Hazel Hall - Room 329