Understanding Broadband Un-adopters

In an upcoming paper for Telecommunications Policy, we present findings from our recent study of broadband un-adopters – those who have had their household Internet connections discontinued. These households represent an important piece of the overall broadband adoption picture, since they have experienced the Internet at home but ultimately did not maintain the connection. Given the recent Pew Research Center findings of a slight drop in home broadband adoption in 2015, future policy efforts should benefit from a deeper understanding of why households terminate their service. Our study takes a comprehensive look at these un-adopting households, including assessing how common this practice is across income levels, and also looks at the underlying reasons for stopping their broadband connections.

Using data from the nationally-representative Current Population Survey, we find that these “un-adopters” comprised 12% of all households without broadband service as of 2013. From another perspective, approximately 4% of households nationwide with a broadband connection have become “un-adopters” – though, as expected, this rate of un-adoption is much more common among lower-income households (see figure).

Source: Current Population Survey and Internet Use Supplement, 2013.

In comparison with their “never-adopter” counterparts, un-adopters are significantly more likely to cite cost, the potential to use the Internet elsewhere, and the inadequacy of their computer as reasons for their discontinued use. In particular, our models show that households with incomes up to $40,000 are more likely to select “too expensive” as their reason for stopping service – suggesting that policies focusing on un-adopters may need to cater to more than very low-income households.

Our research also found that:

• Un-adopters who are also retirees are more likely to answer “no need” as their reason for discontinuing service, and
• Un-adopters in metropolitan areas are more likely to select “computer inadequate” as their reason.

Each of these findings has implications for those working on digital inclusion issues.

Our study of un-adopters also uses recent data from the Federal Communications Commission’s Low-Income Broadband Pilot Program during 2012–2013, in which 14 providers implemented various versions of subsidized broadband access. Across these projects, approximately 22% of those signing up were previous un-adopters (comprising from 9% to 50% of the projects’ participants). Descriptive statistics from the roughly 9,000 people who participated in the Pilot Program help shed some additional light on previous un-adopters. Although instances of missing data limited inference in some cases, several meaningful differences between un-adopters and never-adopters were found:

• Across the projects, un-adopters (as opposed to those who never had a prior broadband connection) seemed to have a preference for wired connections, and were more responsive to specific pricing strategies for those wired connections (i.e. their monthly connection costs were lower).
• Un-adopters were significantly less likely to accept an offer to enroll in a digital literacy program upon enrollment; however, they were more likely to already have completed such a program.
• There were no statistical differences between the groups in terms of how many participated for the full 12 months of the project. Importantly, previous un-adopters were less likely to retain the service both one and three months after the subsidized access ended.

The analysis in our paper, based on two distinct data sets, makes three basic points with policy applications:

1. Un-adopters with income levels up to $40,000 would benefit from a broadband subsidy program,
2. Retired and older un-adopters should be targeted for digital literacy/educational efforts that include strategies to assist those who do not feel Internet content and services are relevant to their lives, and
3. Un-adopters favor wired connections and are extremely price-conscious in their decisions to re-adopt.

Each of these points can be built upon by policy makers attempting to increase rates of broadband adoption.

Understanding and engaging un-adopters will be crucial as the FCC’s Lifeline program and other adoption-oriented programs and policies move forward.

Colin Rhinesmith is an Assistant Professor of Library and Information Studies at the University of Oklahoma and a Faculty Research Fellow with the Benton Foundation.

The full findings of this research will be published in a forthcoming issue of Telecommunications Policy.