Measuring Broadband Should No Longer Be Binary

The Federal Communications Commission (“FCC”) is responsible for measuring broadband access. It publishes an annual broadband deployment report to detail where broadband is and is not available across the United States.  

There’s no need to argue about whether the FCC’s data collection is inaccurate or overstates broadband availability. That would belabor a point that the agency concedes, Congress recognizes, and even the Government Accountability Office views as a problem.   

However, it is essential to highlight a foundational issue that should color how the agency revamps its data collection processes. The way that broadband access is currently measured is based on a telephony model that was originally used to determine whether a household had or did not have a landline phone.  

In a moment when bringing high-speed broadband within reach for every American is a national priority, using the same binary measurement to determine whether a household does or does not have broadband access ignores considerations that simply did not exist in the universal telephone service movement.  

Whether a household is or is not on the wrong side of the digital divide reflects a snapshot of dynamic conditions at one point in time that reveals a household's proximity to high-speed digital infrastructure, subscription status, access to a computing device, and digital literacy. Additionally, as benchmarks for connectivity continue to evolve, so will the data points that need to be collected.  

Long before the FCC existed, between 1887 and 1934, the availability of landline telephones grew exponentially in affluent households and businesses. The 1934 Communications Act empowered the FCC with regulatory authority over the telephone networks, disrupting a monopoly that benefited one company from coast to coast.   

Meanwhile, public sentiment shifted to view telephone access as a utility that could be used to create jobs, build community, promote democracy, support emergency services, and more. The FCC’s had an imperative to (a) collect information on which households did or did not have service and (b) identify the most disconnected areas in order to accelerate deployment.  

Policies that governed how telephone connectivity was measured and regulated left an indelible mark on the broadband landscape. Considering that expanding telephony and expanding Internet access were both viewed as being transformational for society, the challenges associated with telephone deployment should have been used as a cautionary tale for using a light-touch regulatory approach to broadband.  

When telephone service began to penetrate markets, it was most readily available in densely populated areas that promised high returns on investment. Service in hard-to-reach places was expensive and unreliable, if available at all. The most disparately impacted populations were in poor, rural, and Tribal areas. By the early 2000s, similar inequities plagued communities that could not get reliable or affordable access to broadband.  

Strategies that originally governed who would get Internet access were based on supply and demand market principles that paved the way for telephone monopolies. In effect, the same populations that struggled with access to landline phones would also be among the last populations to get access to high-speed broadband Internet.  

Beyond the parallels, there is a notable distinction. Not having access to a landline phone would not necessarily prevent a household from being able to apply for a home loan, complete a work training program, read a newspaper, or participate in a town hall meeting. Today, in a digital society, not having broadband access and the tools to adopt ensures that the most disconnected residents have vastly different outcomes than their connected counterparts. The impact of being on the wrong side of the digital divide is more costly, complex, and could have generational consequences. 

Currently, the FCC determines who is and is not connected to broadband using a system in which a provider self-reports whether it is able to serve one household on a census tract. If a provider is able to service only one location without extraordinary measure, that census tract is marked as served. In turn, all the households on that census tract are recorded as having broadband access regardless of whether a single household in the designated area has a subscription. 

As broadband access expands, landline telephone service has been in rapid decline and cannot always support high-speed Internet. Even though wireless technologies are increasingly a replacement for landline telephone service, which has been eliminated in over 62 percent of households nationwide, federal broadband data collection protocols continue to employ a 20th-century telephony-inspired method to measure 21st-century high-speed broadband connectivity.   

Using a binary approach to determining where broadband is and is not available memorializes a crippling deficiency in the FCC’s data sets. The current methods and metrics fail to capture the variations of connectedness in between broadband access and being able to benefit from technology.  

Federal broadband data does not only serve as the starting point for state and local broadband data collections, it determines which areas get desperately needed resources. Ultimately, inaccurate data means that unserved and underserved communities are deprived of essential grant funding opportunities.  

The FCC is in the process of updating its current broadband maps which, in years to come, may eventually include crowdsourced data. Until then, policies and funding programs will continuously fall short of expectations when priorities are designed with poor data.  

This is an ideal time to consider new metrics that reflect the contours of broadband connectedness – including speeds, bandwidth, devices, digital skills – which will continue to increase over time. Otherwise, deficient broadband data will sabotage good policy intentions and create new digital inequities that are even more expensive to remedy.