The United State(s) of Broadband

Written by Christine Parker

See the interactive resource, United State(s) of Broadband Map, hereUpdated on June 1, 2022.

View and download an HTML version of the map hereTo view, open it in any web browser.

Updates can be shared with Christine Parker at christine@ilsr.org.

*If at any point the HTML file stops working, it's because the map has been updated. Just return to this story or that dropbox folder and redownload the file at the link above.

Tens of billions of dollars in federal funding are poised for new broadband infrastructure deployment over the next five years. But a crucial step in allocating funds from the Broadband Equity, Access, and Deployment (BEAD) Program - for states and local governments - lies in knowing where fast, affordable, reliable broadband access currently is, so that they know where to drive new investment. The FCC’s historical and repeated failure to put together an accurate national broadband map threatens to significantly hold up the process.

Localities and states have learned that they cannot trust big monopolies or the federal government to get this right.  For years, it has ignored the problem or claimed it doesn’t have the funds to solve it. Its data updates (we’re still waiting on the December 2021 drop, and it’s April) are slow, and there’s no doubt among industry experts that even with a new process in place - initiated in the spring 2020 - it too is fraught with complications. Even under the best-case scenario, we’re not likely to see better maps for at least a year to come.

Unfortunately, we don’t know how the process will shake out quite yet. The new FCC maps represent a significant departure, but many states aren't waiting around and have begun to develop their own broadband maps. In classifying the various state-led efforts, we've developed a new resource we're releasing today to serve as an easy reference guide. It shows how states are going about mapping Internet access, and which ones we think are doing it better than others. We’re calling it our United State(s) of Broadband Maps.

Why Are the Maps So Important?

At their most basic function, a map is a handy resource that tells us where something is (or is not) located. Similarly, the current broadband deployment map hosted by the FCC is intended to illustrate where Internet access is available and where it is not. It serves as one of the crucial pillars upon which sits the commission’s charge to carry out the mission of the 1996 Telecommunications Act: “to promote competition and reduce regulation in order to secure  lower prices and higher quality services for American telecommunications consumers and encourage the rapid deployment of new telecommunications technologies.”

Since the year 2000, Internet Service Providers (ISPs) have supplied data (e.g., where they offer service, the technology used to offer service, and the access speeds advertised to customers) to the FCC via Form 477 and is then displayed in an interactive, nationwide map. This map has long had little correlation to the reality on the ground, and given the refusal of governments to collect pricing data, the challenges of predicting wireless propagation, and the lack of reliability data, there is little reason to believe maps will anytime soon accurately represent what service residents and local businesses have available.

For years we’ve struggled with the policy implications of bad data, but in 2020 the Broadband Deployment Accuracy and Technological Availability Act (Broadband DATA Act) aimed to change that. It tasked the Federal Communications Commission (FCC) with creating a “location Fabric” of "broadband serviceable locations" across the whole of the United States. More formally known as the Broadband Serviceable Location Fabric, the Fabric is a geographically-based data product derived from multiple data sources. When mapped, the Fabric is intended to display all locations across the country that could be served with broadband Internet access, and the highest connection speed available. The Fabric will provide the foundation for the newly developed Broadband Data Collection (BDC) that will be used to create the new broadband maps.

Theoretically, the FCC is endeavoring to update the broadband map to provide a more accurate representation of Internet access across the country. At present, the data informing the current map indicates that a census block is “served” if at least one location is offered service within the block, which of course introduces a whole lot of uncertainty into the picture. The result is that this block-level data renders many estimates of Internet access inaccurate. The big monopolies have a history of abusing this ambiguity - suggesting that census blocks that may be partially served should be declared ineligible for funding for a rival, often technologically-superior network. Ideally, the updated map will more accurately highlight the areas that are insufficiently or not at all served by ISPs.

What is The Fabric?

The Fabric is a geographic data product that represents multiple data sources combined to illustrate detailed locations of where broadband could be provided. According to an in-depth report by the Government Accountability Office (GAO), locations that make up the Fabric will be derived from (at least) four key data types: country parcel data (to identify property boundaries); county tax data (to determine property type); building footprints (to determine building location and verify its use type); and addresses with geo coordinates (broadband serviceable locations). The combination of these data is intended to dramatically increase the accuracy of the resulting Fabric. Each serviceable location in the Fabric will be assigned an FCC-issued Location ID. When compiling data for their BDC submission, ISPs will have the option to submit a list of the Location ID’s derived from the Fabric that indicate the locations they currently serve or could serve with a “standard broadband installation,” or they can submit polygons illustrating only the locations that they currently serve or could serve.

The deadline for ISPs and states to submit their broadband data is September 1, 2022.

For more on the new Fabric, read Brian Whitacre's and Christina Biedny's newest piece out in Telecommunications Policy.

Broadband Data Collection - The Good

There are a few unmistakably good things in the new framework. In addition to more accurate location, upload, and download data, ISPs will be required to submit a measure of the network latency, which provides an indication of the speed or efficiency with which the network handles Internet traffic. These data will, at the very least, highlight the distinct difference in user experience.

Another new feature of the BDC is the ability of third-party entities (e.g., state, local, and tribal) to submit verified broadband data that details coverage by ISPs within their jurisdiction. The FCC requires all filers to submit a description of the methodology used to collect and verify data submitted to the BDC. A sound methodology will be necessary for any entity planning to submit data to the BDC, particularly if it is in an effort to challenge ISP data. Local governments would do well to begin collecting broadband access data immediately, since the ability to challenge the monopoly providers’ inevitably overblown claims of service will have a direct bearing on their efforts to connect the unconnected with upcoming federal funds.

Broadband Data Collection - The Problematic

Unfortunately, there’s a host of potentially problematic things baked into the new framework as well. 

Apparently carried over to the BDC from Form 477, BDC filers must also include locations where they could perform a standard installation, which is defined in the Broadband DATA Act as “[t]he initiation by a provider of fixed broadband Internet access service [within 10 business days of a request] in an area in which the provider has not previously offered that service, with no charges or delays attributable to the extension of the network of the provider.” Seemingly saying the same thing as in 477 rules, but in simpler language, this reporting requirement seems to be yet another way of allowing filers to overestimate their service area with little added enforcement power to help. 

Sure, by identifying where ISPs provide service and where they could reasonably extend service to, the general public will have better information about the possible service offerings at their home or business. The problem is, ISPs have often abused this part of the reporting requirements to keep competitors out of their service territories. 

In general, we think that some sort of understanding of where ISPs could reasonably bring service is good, but funding programs and challenge processes should bake in strict requirements and penalties for abusing it. A simple modification to the filing process could include labeling these “could serve” areas in a way that carries over into one of the layers in the new fabric, distinct from the "currently serving" layer. That distinction would provide clarity in determining where service is still needed. Theoretically, this would also mean that all homes and businesses would be accounted for in the BDC, whether or not they have service.  

Another issue replicated from Form 477 is that Internet access speeds reported by ISPs will continue to be the advertised speeds rather than actual speeds. Doug Dawson describes why advertised speeds are problematic, and it boils down to competition: monopoly ISPs exaggerate their speeds to rope in customers, and to stake their claim in areas that would otherwise benefit from grants to support new, competing broadband projects.

When is All of This Change Happening?

In February, the FCC announced that it would begin accepting submissions to the Broadband Data Collection beginning on June 30, 2022. On March 4th, the FCC published detailed descriptions of the data types and formats filers will need to submit to the Broadband Data Collection. Filers must submit data by September 1, 2022, although the FCC noted that the timeline could change if the Fabric and portal become available sooner than expected. Filers are still expected to submit Form 477 data bi-annually until the FCC decides to phase out that data collection process. On April 14, the FCC announced ISPs may obtain access to the preliminary version of the Fabric to become familiar with the format and elements included in it. With the BDC portal expected to open for submissions in June, ISPs would be smart to begin designing their workflow now using the preliminary Fabric, and make adjustments later when the production version does become available - hopefully before June. Unfortunately, we still don’t know when to expect new broadband maps, but at least six months after the first round of data gets submitted.

Some States and Communities are Done Waiting for the FCC

As the nation awaits the release of the Fabric and BDC maps, some states and communities are wasting no time in developing their own maps, some even developing their own Fabric. The move to create independent state or community maps is in response to issues with the current dataset (477) and the unclear timeline for updated maps. In the face of requirements for deployment projects using BEAD funds, states must prioritize projects providing service to unserved before addressing “underserved” locations (i.e., Internet access speeds of less than 100/20 Mbps). Communities and states can avoid letting inaccurate maps dictate the release of tens of billions in broadband infrastructure dollars intended to close a major portion of the broadband gap once and for all. We think it’s a great idea for states and communities to prepare with additional data, especially given the long history of incumbent cable and telephone monopolies in weaponizing challenge processes to bring necessary new infrastructure to unserved and underserved areas. 

Introducing: A Map of Maps

We’ve attempted to summarize where those efforts are today in a new resource: our United State(s) of Broadband Maps. It shows the variety of ways states are pitching in to gather data to drive policy and investment. 

Dark green represents those states we think are doing the best job. Lighter shades of green represent less precise, complete, or useful data in deploying new broadband infrastructure efficiently. Gray represents those states that have no up-to-date map of broadband availability shared publicly. 

Click on the link below to view an interactive version of the map. Each entry contains a description of the data that state used to create its broadband maps, as well as links to the maps themselves.

See a full-screen version of The United State(s) of Broadband map here. Updated on June 1, 2022.

Or, view and download an HTML version of the map here. To view, open it in any web browser.

*If at any point the HTML file stops working, it's because the map has been updated. Just return to this story or that dropbox folder and redownload the file at the link above.

The methodology and data used to create this map can be found on ILSR's GitHub page here. 

How Each State Is Doing

Thus far, only twelve states have purchased or constructed their own Fabric of serviceable locations that are publicly available. Vermont, for example, collected data from ISPs and also developed their own map of serviceable locations based on state e-911 data (geo-located addresses). The private ISP data combined with e-911 locations in this map displays Internet access metrics at individual locations, and makes the state better-equipped to plan for distributing BEAD funds for broadband deployment projects. Other states doing something like or very close to this include Colorado, Georgia, Montana, New Mexico, New York, North Carolina, Pennsylvania, South Carolina, Utah, Virginia, and West Virginia.

Another strategy, used by at least 15 states, is crowd-sourcing Internet access speed-test data. The states collecting speed-test data are often collecting more than just download/upload speeds. For example, Washington state residents are asked to provide the address (or partial address) at which they reside, and the monthly cost of their Internet service plan, or the address where they reside that has no available Internet access. The Washington state speed test results are then mapped and compiled in a dashboard that displays the map of speed test results, locations where no access is available, and summary statistics about the results.

One flaw we foresee in relying solely on speed-test data is a low response rate from locations where Internet access is completely lacking. Those people may have to go elsewhere (with a connection) to report the lack of connection, and understandably this is not ideal. Other states doing a similar thing include Alabama, Arkansas, California, Delaware, Florida, Hawaii, Louisiana, Maine, Maryland, Missouri, Nebraska, Ohio, Oklahoma, and Wyoming. Speed test data is not perfect – it can reflect problems in the home unrelated to the ISP, for instance – but with enough completed tests, they provide evidence of where to focus further investigation for discrepancy between ISP-claimed speeds and those experiences by residents and businesses.

Many states have coordinated directly with ISPs to access data that details (with more precision than a census block) where they offer service. Six states include only private ISP data as a supplement to FCC 477 data. These data will probably still be rife with issues. However, it demonstrates that the state has begun communicating with these ISPs, and that is an important step to take. Based on the state maps where ISPs have privately shared the location of network infrastructure (i.e., copper, fixed wireless, cable, or fiber), the data are an improvement from Form 477 data in that the data often provide detail at the street level. Looking forward, we encourage these states, and communities too, to seek out additional sources of data that can provide more insight about where connectivity is a problem. In addition to private ISP details, e-911, and speed-test data, broadband offices and advocates might also consider using data from the American Community Survey, Measurement Lab, or Ookla to enhance the quality of any maps they create. States going this route include Illinois, Michigan, Minnesota, Tennessee, Texas, and  Wisconsin.

Eagle-eyed readers with a penchant for numbers will recognize that so far we’re talking about 33 states taking steps out of 50. That means there are seventeen states that don’t seem to be doing much (in fact, thirteen of them don’t even display the current Form 477 data), either because they haven’t allocated enough resources to get the job done, they are hoping the broken marketplace will magically solve the problem for them, or they are willfully ignoring cries for help from residents who have endured two years of a pandemic which has forced working, socializing, and learning from home.

Ultimately, the places (states, counties, cities, etc…) that make the effort to evaluate connectivity issues in their jurisdiction, and take an active stance in supporting local ISPs will be the places that are most likely to see connections bridge the digital divide. 

See the United State(s) of Broadband Maps here. This resource will be updated regularly (Updated on June 1, 2022), and we welcome any updates from states as they begin or change their mapping process. Updates can be shared with Christine Parker at christine@ilsr.org.

Header image via iStock by user DeskCube