What is the state of broadband deployment in the United States? The simple answer is that despite years of substantial funding and programs focused on broadband expansion, there is no single federal data source for an accurate answer to that question. The range of estimations varies wildly, depending on what you consider “rural”; what you consider “broadband” speeds; whether you distinguish among transmission technologies (fiber, fixed wireless, mobile/cellular, satellite) or not; and what it actually means for an area to be “unserved,” “underserved,” or “served” – the last of which may not actually mean that broadband is available to everyone living within the census block. Performance metrics such as speed vary across programs and according to NTIA’s most recent milestone report on the federal American Broadband initiative, “too few of the programs measure the direct service outcomes such as the number of miles of fiber constructed, number of access points installed, build-out speeds, household/businesses passed, community anchors connected, implementation timeline, and take-rate. Better accountability is needed to ensure that agency owners, partners, and the public experience the results of these investments.”
There are at least three federal-level efforts underway to develop better broadband data information and insights, but they are still in early stages. However, there is existing data that has historically been used and at least comes somewhat close, plus analysis of that data by third parties which tries to capture what it does not. Together, those approaches can paint an overall picture of how pervasive rural broadband is, what the rural user experience is like, and a sense of year-to-year trends. With that caveat, here are five different data-based perspectives on the extent of rural broadband access.
–The FCC issued its annual Broadband Deployment Report in late April, which emphasized that while broadband deployment isn’t done, there is progress. The report said that for the past three years, its “top priority has been closing the digital divide” but acknowledged comments that “even though ‘remarkable progress has been made[,]’ it remains the case that ‘many people, particularly rural and Tribal areas, do not enjoy the fastest possible broadband speeds or even access to advanced telecommunications services.’” The agency concluded that some 18 million Americans, including 14.5 million people living in rural areas, still lack broadband access. The FCC’s report, looking over a five-year retrospective, found that as of the end of 2018, 94.4% of the U.S. population had wired broadband access at 25/3 Mbps, up from 93.5% in 2017. While the gap – aka, the digital divide — between rural and urban areas has been narrowing, it is still substantial: 22.3% of Americans in rural areas, and 27.7% of Americans living on tribal lands, lack fixed 25/3 broadband service, compared to only 1.5% of Americans in urban areas. Meanwhile, LTE mobile coverage of at least 10/3 Mbps has been extended to 99.4% of Americans as a whole, the FCC found – though there’s still a rural/urban gap. In rural areas, LTE coverage increased from 69.3% in 2017 to 83.3% in 2018, the FCC found.
However, rural and tribal speeds still lag behind urban capabilities. The FCC reported that across five tiers of speed up to 250/25, “deployment in rural areas and on Tribal lands lags behind deployment in urban areas at all five speed tiers, but the data show year-over-year improvements for all speeds in these areas.” For example, the FCC noted, deployment of 250/25 Mbps increased from 28.2% to 51.6% of the rural population.
The FCC’s report is based on information provided by network operators in what is called Form 477, and while it provides what the agency calls a “consistent yardstick” in assessing change from year-to-year, the FCC also acknowledges that the data is “imperfect.” Its most glaring flaw is that a census block is classified as “served” if the Form 477 data indicates that service is available anywhere in the census block – even just to one household out of many. The FCC says that “this analysis likely overstates the coverage experienced by some consumers, especially in large or irregularly-shaped census blocks” – which are more likely to be located in rural areas. The FCC voted last fall to begin a new, more granular data collection effort for fixed broadband availability.
-In August of last year, broadband-focused telecom industry group USTelecom put together a pilot broadband mapping program that had the goal of identifying the precise number of, and location of, every structure in a state that required broadband access, to identify the “unseen” unserved locations located within areas that were considered to have broadband access. The pilot looked at two states, Virginia and Missouri, and included collaboration from USTelecom members AT&T, Verizon, CenturyLink, Frontier, Windstream and others. The pilot found that because of the limitations of the Form 477 data, 38% of rural locations that the FCC would consider “served” were actually unserved. In addition, location data of unserved premises was inaccurate as well, USTelecom found. In the majority of cases (61%), the geocoded locations for structures were off by at least 25 feet and in 25% of cases, were off by more than 325 feet – meaning that would-be providers didn’t even have good information about the distance to an unserved location in order to figure out the cost to extend service, and locations could even be incorrectly listed in terms of which census block they were located in. Nationwide, USTelecom estimates that there are 5-6 million “hidden” unserved locations in the U.S. within census blocks that are considered “served.”
Based on its pilot, USTelecom estimated that a proprietary national dataset based on third-party data of “all broadband serviceable locations” was achievable within 12-15 months for between $8.5-$11 million, plus $3-4 million in annual update costs. A similar open-source dataset, relying on labor-intensive visual verification of the locations (via satellite data), had estimated costs of $22-$24.5 million upfront and annual costs of $7-8 million.
-In addition to the data from USTelecom, consumer broadband advocacy group Broadband Now has conducted its own research that tries to identify the gaps that Form 477 does not. In February (before this year’s Broadband Deployment Report was published), it released an estimate that the real number of Americans who don’t have access to either wired or fixed wireless internet service is around 42 million – more than double the FCC’s figure. Broadband Now arrived at its estimate by manually checked the availability of broadband service to 11,000 randomly selected addresses with nine major ISPs who offered coverage check tools and which, according to Form 477, were served by one of those ISPs. Broadband Now also found that there were variations by state, and that “FCC overreporting [of deployments] disproportionately impacts rural communities.” It found that in Arizona, where 90% of people live in urban areas, unserved addresses were only 11% higher than the FCC’s estimates. But in South Carolina, where only half of residents live in an urban area, unserved addresses were 30% higher than what the FCC reported as unserved.
-Last fall, network benchmarking and analysis company OpenSignal put together data analysis on the user experience of rural mobile subscribers. While 4G availability was generally high, OpenSignal found that the quality of a rural user’s mobile experience was directly related to how far that user was from an urban center. The further away you were, the slower the network speeds. In general, even in the most rural parts of the U.S., OpenSignal found that rural 4G users “could on average spend the vast majority of their time connected to 4G across all rural regions, even in those locations far away from densely populated areas.” OpenSignal, which leverages crowd-sourced, mobile-device-based data for its analysis, classified rural areas as ranging from Fringe to Distant to Remote, using a framework developed by the National Center for Education Statistics for U.S. locales. Speeds decreased as OpenSignal looked further and further outside metro areas.
“Most of our users experienced average download speeds close to 20 Mbps in the Fringe locales, decreasing on average by 4.8 Mbps in Distant locations, and [decreasing]an additional 1.3 Mbps in Remote rural areas,” OpenSignal said. The company found regional variations as well: rural users could connect to an LTE signal more than 90% in 14 states mostly on the East Coast, but that percentage dropped in Western states and large states. Rural users in Alaska and Wyoming had an LTE signal less than 70% of the time. The analysis included some insights that were not so obvious, however, such as that being located in a geographically small state doesn’t necessarily mean a smaller gap in rural vs. urban mobile speeds. Large states like Alaska and Wyoming typically had a smaller gap between the experiences of urban and rural mobile users, “mainly because users, on average, had a slower — and therefore more similar — download speed experience in both urban and rural areas,” OpenSignal said. Upload speeds varied from 4.5 Mbps to 10 Mbps across U.S. urban areas, while ranging from 2.4 Mbps to 7 Mbps in rural areas. Urban users in 46 states had, on average, upload speeds of at least 5 Mbps, while rural users averaged 5 Mbps or more in only six states.
-While the affordability of broadband service usually is considered a separate issue than whether there is local access to being with, it’s worth noting that there is apparently a significant difference between the number of Americans covered by broadband and those who actually utilize the service. The Pew Research Center tracks internet and mobile use via tools including consumer surveys, and as of early 2019, the organization found that rural Americans have “consistently lower” levels of broadband adoption, and “generally remain less likely than urban or suburban adults to have home broadband or own a smartphone.” Pew found that 63% of rural Americans reported having a broadband internet connection at home. That was a significant increase from the 35% who reported having broadband internet in 2007. Smartphone ownership had risen substantially as well, but rural Americans are still less likely than their urban or suburban counterparts to own multiple connected devices. In a separate Pew survey conducted in 2018, 24% of rural American adults said that access to high-speed internet service was a “major problem” in their community, compared to 9% of urban adults and 13% of suburban adults. Another 34% of rural residents categorized broadband access as a “small problem,” meaning that 68% saw it as a barrier to a greater or lesser extent. Meanwhile, a majority of suburban and urban residents said that broadband access was not an issue in their communities.
Want to know more about rural broadband? RCR Wireless News will be publishing a free editorial special report on Wednesday. Stay tuned!
