By John Horrigan and Ellen Satterwhite, Office of Strategic Planning and Policy Analysis.

A recent article by Randall Stross in the New York Times calls attention to a study by the National Bureau of Economic Research in a paper titled “Scaling the Digital Divide: Home Computer Technology and Student Achievement,” by Jacob Vigdor and Helen Ladd, which caused quite a few emails and questions in passing here at the FCC.

The study is an analysis of the impact of home computer and broadband access on student achievement, particularly the impact on standardized test scores in math and reading. Following 5th through 8th graders in North Carolina between 2000 and 2005, the authors report a negative correlation between internet access and standardized test scores. Further, this negative effect is more pronounced for low-income students. Simply put: students who gained access to a home computer between 5th and 8th grades tended to see a decline in reading and math test scores, according to the study.

For those of us deep in the weeds on this subject, this finding was not as earth-shattering as some may have assumed. In fact, it is consistent with the findings in the National Broadband Plan: connectivity and hardware matter, but computers and broadband access cannot replace parents, teachers and broader social support as critical inputs into student achievement. Laptops in the home are not a silver bullet--digital literacy training for parents and teachers, appropriate content for online learning systems, and broader community digital literacy efforts are necessary to ensure children benefit from technology.

Like any general purpose technology, the exact economic and social benefits of broadband are difficult to quantify. Yet, a number of research studies (such as those from the Center for Learning Studies in Urban Schools and Information Communication Technology) in the US and abroad demonstrate that instructional gains come about only if schools undertake new instructional approaches tethered to technology and if they adopt new practices to support the technology. The FCC took these findings and made recommendations to support and promote digital literacy for teachers and in the classroom and to support the development of innovative broadband-enabled online learning solutions in Chapter 11 of the National Broadband Plan.

Furthermore, as Vigdor and Ladd point out in their review of the literature, there is evidence that holistic broadband adoption and use programs—those that involve more than simply providing laptops to children—have positive impacts on student classroom performance. Such findings were the support for recommendations in Chapter 9 on Adoption and Utilization. Recommendations like the National Digital Literacy Corps and improved training and support for libraries and other community-based organizations, are ways to build a community’s digital social support structure and help make broadband access beneficial, and not detrimental.

The studies highlighted in the Times’ article are valuable contributions to the discussion of how to make broadband part of educational solutions. The National Broadband Plan recognizes that computers and high-speed connectivity can play an important role in improving outcomes in the classroom – along with the expertise of the educational community, engagement by the technology sector, and involvement of family.

Today, the FCC is releasing the raw dataset that was the basis for two recently released reports. One, released on May 26, was on bill shock and early termination fees. It is entitled “Americans’ perspectives on early termination fees and bill shock.” The second report, released earlier this week, is entitled “Americans’ perspectives on online connection speeds for home and mobile devices.”

The survey that was the basis for these reports covered a lot of ground, and the two reports we recently released did not cover all of it. In the coming weeks and months, the FCC will release findings analyzing other questions from this survey. For that reason, the data released today does not include all data from the April-May 2010 survey, but data only on those questions analyzed in the two reports.

The package of data comes in two files:

1) Raw data: Delivered in SPSS format, which is a popular program for statistical analysis of data. It is a format most other statistical programs can read.

2) Questionnaire: This file explains in detail the structure of the SPSS file, and will be of interest mainly to those wishing to do their own analysis of the data. You will find that the questionnaire contains all the survey’s questions – but not the results.

The FCC has a commitment to transparency in conducting the analysis that helps shape its work. We hope interested members of the public benefit from having access to the data. Enjoy!

Today’s news has several nuggets on the importance of libraries in the broadband access ecosystem. The University of Washington released a study, funded by the Bill & Melinda Gates Foundation, that found that 77 million people – or one-third of Americans over the age of 14 – have used public library computer or wireless network to go online at some point. Some 40% of library computer users sought help with career needs and 37% looked for information about health issues.

Another news item underscored the importance of Internet access at libraries in the current economic climate. The Baltimore Sun has a story on proposed budget cuts for the City of Baltimore since, like many cities around the country, the recession is forcing Baltimore to tighten its belt. However, the city’s libraries are “largely untouched” from proposed cutbacks because library use has increased by 20% in the past year. Much of this increase is attributed to unemployed people using library to look for and apply for jobs. As FCC research shows, 60% of broadband users have gone online to apply or look for a job – and 73% of those not currently employed have done this.

All of this is very consistent with research the FCC commissioned from the Social Science Research Council to better understand adoption barriers among low-income Americans. A key finding of that research is that library resources have come under stress during the recession as people come there to apply for jobs or carry out other important tasks.

The National Broadband Plan recognizes the important role libraries play in access. The Plan recommends that Congress consider providing additional funds to the Institute of Museum and Library Sciences to improve connectivity, enhance hardware, and train personnel of libraries and other community based organizations. Libraries serve not only as a crucial resource for access, but they also provide an environment where new broadband users can acquire the skills to take advantage of the range of online resources.

Today, the FCC is releasing the raw data files that were the basis for “Broadband Adoption and Use in America” working paper. The Broadband Data Improvement Act directed the Commission to “conduct and make public periodic surveys of consumers” as part of the FCC’s efforts to understand who uses broadband, who does not, and, if not, why people do not subscribe. We released the results of the survey on February 23rd, and today we make available to the public the underlying data for the survey. The data (downloadable here) comes in several files:

1. Raw data: Delivered in SPSS format, which is a popular program for statistical analysis of data. It is a format most other popular statistical programs can read.
2. Codebook: This file explains in great detail the structure of the SPSS file, and will be of interest mainly to those wishing to do their own analysis of the data.
3. Cross-tabulations: These files show how different categories of respondents (e.g., those in certain age cohorts, or those with different educational levels) answered survey questions.

If you’d like to see the “topline” survey results, they are already online here.

The FCC has a commitment to transparency in conducting the analysis that helps shape the National Broadband Plan. We hope interested members of the public benefit from having access to the data, and the ability to ask questions of the data that the FCC working paper did not pursue. Enjoy!

Increasing the current levels of broadband adoption in the U.S. from the current level of 65% will not happen automatically. Last week’s release of the FCC’s “Broadband Adoption and Use in America” helped frame the challenges that current non-adopters face. The survey found that, when pinned down on the most important reasons for non-adopting, cost led the way (36%), followed by digital literacy (22%), and lack of relevance (19%).

Yet the survey sought to explore whether barriers to access had more than one dimension. Common sense tells us that barriers to use of any product can have more than one component. 

And so it is for broadband. The survey was structured in such a way so that people were asked about multiple barriers to adoption before the survey asked them to identify the most important reason. This lets us look at a single barrier – such as cost – in several ways. Some 66% of non-adopters identified at least one of four cost-related factors as barriers to adoption. Those factors were: level of monthly fee, affordability of a computer, installation fee, or reluctance to enter into a long-term service contract. Half specifically pointed to the level of the monthly fee as a problem. Notably, however, 85% of those who cited monthly fee as a barrier also cited at least one of the other three cost-related reasons. 

A new study by the Social Science Research Council entitled “Broadband Adoption in Low-Income Communities,” released on March 2 had similar findings. The SSRC research is based on structured conversations with more than 150 non-adopters. The SSRC sample is not nationally representative, but its qualitative research approach has the significant advantage of eliciting stories about the context of non-adoption. The FCC commissioned this research to deepen our understanding of barriers to adoption.

Like the FCC study, SSRC’s research underscored the importance of cost. Essentially all non-adopters the SSRC team interviewed mentioned cost as a barrier to adoption. SSRC also found that cost has multiple dimensions. Low-income people without broadband said that cost to them means not only monthly fee, but also hardware and software costs (including virus protection), installation costs, equipment maintenance fees, and transaction costs for disconnecting. The SSRC research also found that many non-adopters lack the skills to carry out online tasks such as applying for a job – which places great strain on public libraries whose staff often serve as the de facto help desk.

The SSRC study had two additional findings that resonate with the FCC survey.
 

1. Un-adoption: SSRC’s sample of non-adopters included 22% who had broadband once, but “un-adopted,” that is, they had to disconnect service. Reasons given for this included loss of a job, technical problems (e.g., the computer broke or was rendered useless by viruses), billing issues such as unexpected hidden fees, or bundling problems (if, over time, one part of the bundle proved difficult to sustain). In the FCC survey, 8% of non-adopters had “un-adopted.”
2. The need to have service: SSRC found that all the people with whom they spoke understood the need to have broadband access – and these people said the drivers for use for them were education, access to jobs, and access to government services. None of the people in SSRC’s sample had broadband at home, but all were willing to go to great effort to use it at libraries or community centers. Indeed, the SSRC report conveys an anxiety about broadband among the non-adopters interviewed; they know its importance yet face hurdles to getting it at home.

It is worth noting that the FCC survey did find that some non-adopters – mostly older ones – say they do not have a compelling need for service.  The group of non-adopters who say lack of relevance is their main adoption barrier breaks down as follows: 5% say they do not see the need for more speed, 5% say they believe the internet is a waste of time, 4% say there is nothing they want to see online, and 4% do not use the internet very much.

The FCC and SSRC research tell us something important in thinking of cost as a policy lever to address non-adoption. Lowering the monthly cost of access would unquestionably help many adopters get online with broadband at home. But cost has a broader context than just the monthly bill. Cost of ownership (e.g., maintenance, necessary software) and nature of service plan (e.g., length of contractual commitment or, as SSRC finds, bundling) matter too. Moreover, many non-adopters need the basic skills on how negotiate online sessions to carry out tasks that increasingly require broadband access.

These findings point to an implication of the research that has explored reasons for non-adoption: addressing non-adopters’ barriers will require comprehensive solutions to address the multiple cost and skills hurdles people without broadband face.

The great thing about bringing people together to talk about broadband adoption data is that you always get new ideas for questions to pursue in the future. At the "Building the Fact" base workshop last week, this dynamic played out as expected. Susannah Fox, from the Pew Internet Project, reminded listeners of the growth in broadband adoption at home from just 3% of Americans in 2000 to 63% as of April 2009. Link Hoewing at Verizon noted the fast adoption pace of broadband relative to other communications technologies. But he, and others, pointed out how we have likely entered a maturing phase on the adoption curve, meaning that reaching the remaining 37% of non-adopters will be a challenge. Peter Stenberg from the Agriculture Department highlighted particular challenges for rural Americans.

We were also reminded how strongly broadband can impact how people get information. Susannah Fox had a striking statistic: 42% of all adults say they or someone they know has been helped by following medical advice or health information found on the internet. This represents a significant increase since 2006 when 25% of all adults reported being aware of helpful outcomes. As wireless access means become more prominent, as CTIA's Christopher Guttman-McCabe said, these kinds of trends, as well as content sharing, will be reinforced.

How to address the remaining non-broadband adopters in the U.S? Karen Archer Perry and Kate Williams both provided insightful suggestions. Both noted the human element in adoption-promotion efforts, saying that there must be a social infrastructure to promote adoption. That is, non-adopters need nearby training programs, where teachers can not only help build digital skills, but demonstrate the relevance of particular applications. The goal is to motivate people to access broadband, grow their confidence in using it, and allow broadband and online content to enrich their lives more and more over time.

All the panelists' ideas are wonderful food for thought as Task Force thinks through broadband adoption barriers.