In 2012, the Federal Communications Commission (FCC) released its eighth Broadband Deployment Report, its annual account of the state of broadband deployment that Section 706 of the Telecommunication Act requires. The so-called 706 report found that approximately 19 million Americans at the end of 2011 lacked access to high-speed internet access. This figure, the FCC concluded, meant that “broadband is not yet being deployed in a reasonable and timely fashion.”
Fast forward to May 2019: the FCC distributed a press release summarizing findings from its revised 2019 706 Report and stated that at the end of 2017 21.3 million Americans lacked access to broadband networks. However, the FCC now concludes that this gap in broadband deployment meant that broadband is being deployed in a reasonable and timely fashion.
This is not an apples-to-apples comparison (for instance, the speed threshold for broadband was 3 Mbps download in 2012 and 25 Mbps in 2017 data analyzed in the FCC’s latest report), but it raises the question: How can 19 million people without broadband be a problem in 2012 but some 21 million without broadband not be a problem in 2019?
Politics is an obvious answer. The 706 report can prompt calls for policy intervention and the 2012 FCC was more disposed to action to address issues in the broadband market than today’s FCC is. Even if faster speeds were used today as the threshold for the definition for broadband (suggesting a greater number if Americans would be unserved) one gets the sense that this FCC would nonetheless describe progress as adequate. It is implicit in the press release, which notes, from 2016 to 2017, an 36% increase in the number of Americans with access to networks with 250 Mbps download speeds.
But FCC Chairman Ajit Pai’s press release points to two other aspects of the 706 report that deserve attention:
- The shaky metrics it rests on, and;
- Its constricted view of the digital divide.
The Range of Estimates on Who Lacks Broadband is Huge
Last year, Microsoft released a report that found that 162.8 million Americans do not use broadband, 19 million of them in rural areas. That is about 8 times the figure the FCC cites; in fact, the rural broadband gap in Microsoft’s analysis is not far short of the figure the FCC cites for the broadband shortfall for the entire nation. It is not just Microsoft that questions the FCC’s numbers. Analysis of speeds in Kansas and Maine, using the M-Lab speed measurement tool, showed that only 2 of 24 cities had network speeds that hit the FCC’s 25 Mbps speed threshold, even though FCC data showed extensive 25 Mbps coverage. Most cities had a median speed below 10 Mbps. A similar exercise in Iowa had comparable results. FCC data shows that all of Iowa has access to 25 Mbps networks, but speed tests found that internet users experience that speed just 22% of the time.
The discrepancy in findings is due to how data are collected. Microsoft and M-Lab (in different ways) examine actual network speeds that consumers use. The FCC’s 706 report relies on carriers reporting advertised speeds (which can and do differ from actual speeds) by Census block using something called Form 477. That approach overstates broadband coverage (as the independent, nonpartisan Government Accountability Office has told the FCC and Congress), since an entire Census block will show coverage even if a carrier can provide service in a small portion of it (even just one address!). This is a widely acknowledged problem and the National Telecommunications and Information Administration (NTIA) is trying to address it through partnering with states to develop broadband maps with more granular geographic data.
The Form 477 method is also vulnerable to errors in the data carriers report. Figures in Chairman Pai’s February press release were questioned because one carrierapparently over-reported its network’s coverage. The FCC’s revised 706 report has corrected that error, which, as Commissioner Geoffrey Starks notes in his dissent on the broadband deployment report, changed the figure for the number of Americans not having access to broadband to 21.3 million from 19.4 million. This is not to say that other methods, such as consumer speed-tests, are flawless. Nonetheless, as NTIA examines ways to improve broadband mapping, speed tests that measure network performance have a role to play in helping policymakers better understand network coverage.
The Digital Divide is About More than Nationwide Network Deployment
In finding progress in broadband network deployment, Chairman Pai’s May press release states that “the nation’s digital divide is narrowing” and he traces that outcome to FCC policies under his leadership. The implication is that the digital divide is a national problem for federal officials to address and is primarily about network deployment.
But the digital divide has multiple dimensions. It encompasses home broadband adoption, smartphones, digital readiness, and tech support. Certain population segments – such as low-income Americans – significantly trail well-off Americans. The American Community Survey’s 2017 1-year estimate shows that 83.5% of households subscribe to broadband at home. Yet 95% of households with annual incomes over $75,000 subscribe to broadband compared with 59% for poor households (those with annual incomes below $20,000).
Smartphones are another tool for online access. Some 77% of all Americans had one by early 2018, but variation around income is significant. The Pew Research Centerfinds that 67% of those in households with annual incomes below $30,000 had smartphones compared with 93% for those in homes with annual incomes over $75,000. Although smartphones are obviously pathways to the internet, it is not at all clear that smartphones as a person’s sole access point to the internet is an adequate substitutefor a home wireline subscription. In fact, most of the growth in “smartphone-only” Americans in recent years has been among low-income Americans, suggesting the substitution phenomenon is more about affordability and less about functionality.
Although these aggregate gaps in adoption by income warrant attention, their severity tracks household income differences that unfold at the local level. This means solutions to digital inequality should emerge locally – often neighborhood by neighborhood. For the most part, though, federal policymakers are not well attuned to this.
The current buzz over 5G wireless networks is an example, as 5G deployment is both expensive and radically local. Estimates for nationwide buildout reach $275 billion, which means, like any other capital intensive network infrastructure, 5G buildout cannot happen all at once everywhere. Buildout may not just be block-by-block, but structure-by-structure; 5G will require 800,000 small cell sites by 2026, more than twice the number (323,448) deployed at the end of 2017.
For sound reasons, this means 5G deployment will be inherently unequal, which is why 5G portends a very local “coming digital divide.” Yet the FCC is framing deployment as an international “race to 5G,” but a race that is likely to exacerbate inequality of deployment within and across cities. A “race to 5G” means an investment sprint to wealthy neighborhoods in wealthy cities, followed by a slow-walk to lower-income places. Poor and even middle-income communities are likely to see nifty new 5G apps in movies and ads long before they see them in their neighborhoods.
Back to the original question: What does the latest 706 report tell us about the digital divide? Not very much. It uses carrier-reported data that vary wildly from other methods and focuses only on a single part of the digital divide. Here are suggestions for action:
- Continue to improve metrics: As the NTIA addresses broadband mapping, it should consider bringing network speed measurement into the picture to capture network performance at the local level.
- Bolster capacity at the local level: With the digital divide becoming more local, the federal government should equip localities with the resources they need to address it. The FCC recently gutted the ability of local governments to raise resources to address digital inclusion by limiting their ability to charge fees for rights-of-way for 5G deployment. Prior to federal preemption, some places used such fees for digital inclusion planning and funding. The federal government should develop a digital inclusion planning and grant program to help defray, in part at least, lost local revenues from the FCC’s action. Senator Patty Murray recently introduced the Digital Equity Act of 2019 which offers a vehicle for such a program.
Policymakers’ focus on the digital divide will not go away anytime soon. As dialogue on it continues, better data and a broad understanding of the problem are crucial to helping decision-makers make progress on the digital divide.
John B. Horrigan is Senior Fellow at the Technology Policy Institute, with a focus on technology adoption, digital inclusion, and evaluating the outcomes and impacts of programs designed to promote communications technology adoption and use. Horrigan is also currently a consultant to the Urban Libraries Council. Horrigan served at the Federal Communications Commission as a member of the leadership team for the development of the National Broadband Plan. Additionally, he has served as an Associate Director for Research at the Pew Research Center, where he focused on libraries and their impact on communities, as well as technology adoption patterns and open government data.