Through the Broadband Equity, Access and Development (BEAD) Program, each state will receive a minimum of $100 million. Additional funds will be provided based on the number of locations that are “unserved” (do not have access to reliable service of at least 25 Mbps download and 3 Mbps upload) or “underserved” (do not have reliable 100/20 Mbps service).
It has been an open secret for some time that mapping who does and who doesn’t have service at these levels is far from an exact science. The Federal Communications Commission (FCC) counts an entire census block as “covered” if one household in it has broadband service.
In 2018, Congress provided funding to the NTIA to create a National Broadband Availability Map (NBAM) and to work with FCC as well as state and local governments, nonprofits, network owners and operators and other stakeholders to achieve this goal.
So far, 40states and territories have provided data to NTIA. But much of this data is from resident surveys or requests for service collected via web portals, says William Rinehart, a senior research fellow at Utah State University’s Center for Growth and Opportunity. It might be more than was previously known, but it’s still missing important details; moreover, the data is not available to researchers.
One state took a more comprehensive approach to tracking coverage and shared its data set. Rinehart used this work as the basis for a national projection.
FCC data consistently overstated coverage ranges, according to data collected by University of Georgia researchers. Hover over a county to see comparisons.
Household by Household
In June 2021, the state of Georgia published a broadband availability map created by overlapping the location of every home and business in the state with broadband service available to those locations. The work of creating the map was overseen by the Institute of Government at the University of Georgia.
The map, the first of its kind, revealed coverage gaps that were not apparent in FCC data. For example, in Fulton County, designated 100 percent covered by the FCC, the enhanced map included 16,000 households without broadband.
Rinehart took a deep dive into the Georgia data set, analyzing coverage rates and their relationships to a variety of other demographic data points it contained, including such things as education, median income, housing units and FCC metrics. He spent months developing a mathematical model that he could use to estimate coverage across the country.
The projections that resulted suggested that actual coverage could be in the range of about 90 to 93 percent of U.S. homes, significantly lower than the 97.5 percent estimated by the FCC. (The lower estimate came from a model that controlled for age, income, education as well as geographic size of a place and the number of homes.)
“I’m trying to add more data into these estimates,” says Rinehart. “We don’t have a good sense of where the gaps in broadband actually exist — it seems that money was allocated before there was understanding of the actual need.”
Modeling by the Center for Growth and Opportunity based on data collected by the University of Georgia researchers suggest that FCC estimates overstate coverage by 40 percent or more in some counties. Scroll over a county for details.
Rinehart created another map emphasizing counties that are priorities for broadband investments based on their low estimated coverage rates. In some cases, his estimates are consistent with FCC data, but in many counties the need appears to be much greater than FCC’s estimates would suggest.
Rinehart has problems with the process behind federal decisions. “Obviously, there are gaps and people need service,” he says. “Broadband support is important and it can be transformative in lives — but at the same time, I’ve been calling for better research methods and a better sense of the agenda that we should be working through."
This map from Utah State University’s Center for Growth and Opportunity, which includes expected connections supported by the Rural Digital Opportunity Fund, highlights counties that should be priorities for broadband investment. Hover over a county for details regarding coverage gaps.
A Multidimensional Problem
A growing number of states are making it a priority to collect their own data, says Francella Ochillo, executive director of Next Century Cities. “I think everyone is rightfully concerned about sending tons of money out the door based on inaccurate data.”
Federal mapping relies too much on data from service providers that is not independently verified, including the actual speeds obtained by users, according to Ochillo. She cites a recent survey in Ottawa County, Mich., that found 11 percent of its residents don’t have broadband Internet — and that 15 percent of those with service have speeds below the FCC’s broadband threshold.
What this means is that one in four residents lack broadband service, not one in 10. When the federal benchmark changes — BEAM already characterizes service slower than 100/20 Mbps as “underserved” — the problems encountered thus far in data collection could begin anew, points out Ochillo.
Some providers are already reluctant to have their data regarding baseline speed verified externally, she says. “These types of shenanigans will only increase as the baseline increases.”
Ochillo believes that FCC Chair Jessica Rosenworcel and her colleagues are making a genuine effort to resolve mapping discrepancies. Still, it’s her perception that the process continues to be “slow walked," impeded by private-sector concerns about the economics of inclusive high-speed access.
With accurate data generally scarce, it will be even harder for disadvantaged communities to cobble together the data they need to support requests for federal grants, she says. Some will not even know where to get in line for them.
“Obstacles related to affordability and adoption are just as much of a problem as those related to infrastructure deployment,” says Ochillo. “If we know something is multidimensional and our response is one dimensional, talking only about access, then we haven't made the type of progress that we think we have.”
Toward Consensus and Standards
Last fall, a group of public- and private-sector stakeholders joined together to launch the National Broadband Mapping Coalition, hoping to draw on technology, research and policy expertise to raise the standard for mapping. The coalition is a project of the Marconi Society, with partners including Google, the Internet Society, mLab, XLAB and the Institute for Local Self-Reliance.
At present, the coalition isn’t focused on making yet another map but to make sense of a complicated field, says Samantha Schartman-Cycyk, executive director of the Marconi Society. “There is no lack of approaches to measurement — what is lacking is understanding, standardization and consensus.”
Schartman-Cycyk believes it will take years to develop an authoritative approach to mapping that would be open source, transparent and auditable by researchers — all while ensuring privacy and safety. This work will require building agreements regarding matters such as metrics, measurement tools, data analysis and visualization of aggregated data.
As long as map-making and access revolve around expensive, proprietary approaches and interpretation by data scientists, affluent communities will continue to have an advantage when competing for federal funds, says Schartman-Cycyk. “Until we create transparent, accessible standards of measurement that can be collected equally and understood in all communities, we are going to continue to deal with issues of inequality.”
Even though better maps can help prevent underserved communities from falling through the cracks, she isn’t counting on maps alone to ensure equity. “Each geography is unique and brings with it a unique blend of access to data, public-private partnerships and types of infrastructure that need to be measured,” she says. “One map will never rule them all.”
While the coalition hasn’t yet done extensive outreach to state and local governments, some officials have joined the conversation it was created to foster. It hosts a listserve and monthly calls to highlight mapping efforts by states, consultants, researchers and technologists.
“For the moment, the best we can do is bring some organization to a very chaotic environment and gently guide the field toward productive, collaborative and transparent progress,” says Schartman-Cycyk. “We are uniquely situated to be able to convene these disparate parties and guide this conversation.”
Related Articles
