Who's still offline and why? The real reasons

In its most recent survey of US Internet adoption, the Pew Internet and American Life Project found that 85 percent of all American adults now use the Internet, up from 80 percent a year ago.

While that is good news, that still leaves 15 percent, or 24 million adults, offline. Who are the holdouts? The US still has a "digital divide," but it's not the one most people would imagine. According to the survey, the most significant factor is age: Nearly half of non-users are age 65 or older. Education is the second-most important factor -- more than 40 percent do not hold a high school diploma. Other factors, including sex, race, income level, and geographic location, are less significant, and continue to decrease.

The reasons given (PDF) by those who remain offline, however, provide the most important and perhaps surprising insights. In this year's survey, released Wednesday, the most frequently reported reason given by Americans who do not to use the Internet is that it isn't relevant to them. Usability was the second-most cited reason. Together the two accounted for 66 percent of those who are not online. Price and availability were the least-important reasons.

Misreading the Data

The Pew survey provides important guidance for industries and government agencies whose goals are to get the US as close to universal Internet adoption as possible.

But some advocates are twisting the new data to fit a pre-existing political agenda it simply doesn't support. In an op-ed for Slate.com, for example, researchers at the New America Foundation's Open Technology Institute badly misrepresent the Pew data, reading it to support their conclusion that "high cost and lack of access keep millions of Americans offline."

But the Pew survey actually shows that cost and access are the least significant factors for those who do not use the Internet. Together they were cited by only 26 percent of the survey respondents.

"If you torture the data long enough," Nobel prize-winning economist Ronald Coase famously wrote, "nature will always confess." Worse, misreading the data also leads to misplaced public and private resources. Internet adoption is a key priority for the US economy and for the White House in particular. But with limited resources available to encourage older and less-educated Americans to go online, ignoring the actual findings will lead to counterproductive regulation of Internet access and even slower adoption. Those who need the most encouragement to go online won't get it.

The New America researchers, for example, conclude from the relatively low number of respondents citing cost and availability as justification nontheless for their long-standing goal of mandatory price controls and government regulation of network infrastructure and equipment. But these policies would do nothing to educate older Americans on the value of Internet adoption or improve the usability of existing technology. And similar policies have been tried in the EU and failed, as regulators there acknowledge.

The researchers also rely on a recent report from the Federal Communications Commission that suggests some 19 million Americans, the vast majority in rural areas, do not have access to any broadband Internet provider at home. (The FCC defines broadband speeds [PDF] as 4Mbps download and 1Mbps upload. Internet access via dial-up and DSL speeds below that threshold is simply not counted.)

For the New America researchers, that data suggests the need for substantially expanded government intervention in the Internet ecosystem, principally to build new infrastructure and to subsidize individual access. The FCC's multibillion-dollar Connect America Fund, for example, which is paid for by taxes on consumer telephone service, now offers grants to private network providers willing to deploy broadband infrastructure in high-cost rural areas. The troubled Lifeline program, likewise, is shifting from subsidizing basic telephone service to providing the poorest Americans with broadband Internet.

Before escalating these and other government Internet programs, however, policy makers should once more take a closer look at the numbers. Again, the New America researchers seriously misread the data. There are not 19 million Americans without any broadband access at home.

For one thing, the FCC report they cite ended its data collection in June 2011. But network providers and their investors have spent billions in infrastructure improvements over the last two years, as they have every year since the broadband revolution began. Between 2010 and 2011 alone, when the FCC's data runs out, the number of Americans without access to a single home broadband provider fell by 7.4 million, a number almost certain to have dropped further since 2011.

More important, the FCC report simply ignores the availability of mobile broadband. Although 4G and other networking technologies can and do deliver speeds that exceed the FCC's broadband threshold, the agency excluded mobile entirely from its statistics on access, citing a lack of "reliable" data on precisely how many of the 19 million could or even do get service from mobile broadband providers.

Rather than estimate, the FCC simply counted the entire wireless industry as zero. Yet with the explosion in smartphone use, as every consumer knows, wireless broadband growth far exceeds that for wired broadband. Mobile carriers in the US, according to the same FCC report, invested more than $25 billion in network improvements just in 2011, compared with $18.6 billion in the 15 largest European economies combined.

Including data on mobile broadband access provided in the FCC's report but left out of its calculations, the number of Americans without any home broadband provider falls as low as 5.5 million. That's less than 2 percent of the population.

To put that number in perspective, consider that landline telephone service never achieved more than 95 percent (PDF) adoption in American homes. Indeed, according to the US Census Bureau, more than 3.5 million Americans still lacked complete indoor plumbing as recently as 2011. Yet universal telephone service has been the policy of the US since the formation of the FCC in 1934. And public efforts to improve household sanitation predate the founding of the Republic.

So the fact that broadband Internet access and adoption are still less than 100 percent less than two decades after its invention, while unfortunate, is hardly a crisis. Even excluding mobile broadband, the FCC report found that 95 percent of American homes have access to a broadband connection, in most cases with two or more choices of provider.

What Can Governments Do?

Still, complete connectivity is a worthy goal. And governments can play a critical role in encouraging the few remaining Americans who have so far failed to see the benefit of Internet adoption. But in allocating limited public resources, it's essential to read the data for what it actually says before deciding how to proceed.

For example, the most-frequently cited reason given by Americans not to use the Internet, according to the Pew survey, is that they are "just not interested." Is that because the Web really holds no relevance to them, or because the case for connectivity just hasn't been made? Given the increased importance of Internet technologies and protocols for basic communications, entertainment, education, employment, health care, and public safety, going online is becoming a crucial part of daily life, especially for the groups who today have the lowest levels of adoption. They just may not fully appreciate the value to them of doing so.

Governments could be much more effective in using their bully pulpit to make that case. The FCC's exceptional 2010 National Broadband Plan, for example, included detailed descriptions of the promise of the Internet to improve the lives of all Americans in each of these core areas. Yet the agency did almost nothing to promote the hard work of the plan's authors once the document was published.

It would cost little for the FCC and the rest of the administration to dust off the plan and highlight the vision it painted for a fully connected America. This could take the form of speeches, government Web sites, and public demonstrations coordinated with industry leaders. Much of that activity already exists; it just needs to be focused.

The FCC could also accelerate its ongoing efforts to remove regulatory roadblocks, especially at the state and local level, that cause unnecessary delay and cost in the deployment of additional and critical broadband infrastructure. Despite a modest "shot clock" established by the FCC in 2009, for example, permit requests for new cell towers and equipment changes on existing utility poles and buildings continue to drag on for months or even years, often held back or even denied based on little or no zoning-related justification. Laying new fiber-optic cable, likewise, often requires navigating a maze of local authorities.

The agency should also restart stalled efforts to begin trials for a final switchover from the aging, obsolete public switched telephone network to all-IP networks.

A majority of Americans have already cut the cord to the old networks. Maintenance costs for providers who have to support both the old and new networks are inhibiting even more investments in broadband. Landline carriers have asked the FCC for permission to conduct trials on switching over the remaining customers to better and cheaper IP networks, but the agency has stalled, with special-interest groups gumming up the works.

Ironically, households still using analog telephone switches are also those most likely not to be Internet users. Accelerating the transition to IP networks would get those customers onto the Internet, at least for voice services.

These are the kind of government interventions that will actually help those few Americans who still don't see the value of Internet adoption. And they require no new laws -- just a careful analysis of data that is already clear and convincing for those who are willing to read it.