The Need for Speed – Gauging Required Speeds for Broadband Buildout

In our previous post on broadband, we provided a high-level overview of the current state of broadband in the United States, given the context of the $65 billion that the Bipartisan Infrastructure Law has provided to close the digital divide. But a major debate in this buildout concerns the precise speeds that US citizens need to work, play, and connect in modern cyberspace. This is not an easy question because the answer is always changing as new applications, infrastructure availability, and societal demands for online activity develop. Activities like telehealth and remote work are likely permanent fixtures in American life, so broadband demand is more pressing than ever. The Federal Communications Commission’s (FCC) formal definition of “broadband” serves as the federal standard for the minimally acceptable speeds for which an average household connection can meet modern-day demands. The goal of the National Telecommunication and Information Administration’s (NTIA) $65 billion buildout is to ensure that every American can fully participate in online life well into the future, so it is important that we try to get these numbers right as best we can.

Background

We have outlined the definitions of download (receiving) and upload (sending) speeds (usually expressed as a ratio of download/upload). We also hosted a number of stakeholders to discuss what is required to narrow the digital divide. The FCC’s current 25/3 definition was set in 2015, but just five years earlier, in 2010, the definition was a mere fraction of that at 4/1. At the turn of the century, the term “high-speed Internet” referenced broadband download speeds of just 200 kilobits per second (see figure 1). Our production and consumption of data have increased by orders of magnitude, not gradually, over time, and there is no reason to believe that this trend will abate. To illustrate, the average US household used 38 times the amount of data in pre-pandemic 2020 —approximately 350 Gigabytes (GB) a month — than it did a decade earlier, and in 2021 that number was approaching 650 GB/month. This is important to remember when attempting to forecast future speed requirements.

Let’s Talk Numbers

To put the current broadband definition in context, a household with a 25/3 connection can download a 5-minute web video file (about 30 megabytes of data) to a local computer in about ten seconds. This might be surprising because, based on the stated numbers, you would assume it would only take around 1.2 seconds. However, there are often other connections sharing the line, interruptions to service, or competing factors that will slow down the connection speed. ISP Allconnect found the actual capacity of a fiber or cable Internet connection is only about 80% of its advertised speed, while the FCC’s most recent Consumer Fixed Broadband Performance report found that measured download speeds were 100% or better than advertised speeds during peak hours. These results are inconsistent because broadband speeds themselves are subject to circumstantial factors. They depend in part upon the number of simultaneous users, the type of connection, the investment in infrastructure, and the speed tiers available and paid for. Therefore, measurements of broadband speeds will usually report a mean or median speed over a period of time or a broad location.

Not every household or business requires the same high internet speed. Needs differ based on a host of variables, including location, household applications, educational needs, and employment situations. A household of one will have a different set of data requirements than a family of six, and a household utilizing the Internet for virtual doctor visits, remote work, or remote education will have different needs than one that does not. Each online activity has its own speed requirements, and the FCC has published its speed recommendations for the most common activities, but these are very much open to debate.

FCC Actions

In light of the increasing demands for speed from both users and various applications, FCC Chairwoman Jessica Rosenworcel has proposed to change the FCC’s definition of broadband to 100/20 as part of the agency’s annual evaluation of the state of broadband throughout the US. Her concern is that the 25/3 definition is outdated and hides just how disadvantaged rural and low-income communities have become since the pandemic (She also states that a longer-term goal is 1 Gbps/500 Mbps – true Gigabit Internet on par with South Korea). However, according to the latest FCC data, only 30% of the reported speeds from various ISPs across different states and block codes meet the 100/20 threshold, in contrast to the 60% that meet the 25/3 threshold. Because of this differential, broadband speed definitions have significant implications for which states and neighborhoods may be eligible for NTIA’s money. However, NTIA’s Assistant Secretary Alan Davidson is on board with the 100/20 definition, having pledged to deliver those speeds across the country.

Some experts do not believe that a change in speed definition is worthwhile at this juncture, fearing that doing so will reallocate funds away from areas that lack even 25/3 to those with otherwise acceptable bandwidth. In 2019, the Wall Street Journal determined that over 50 of its journalists rarely used all the bandwidth they were paying for, even when streaming videos on several devices simultaneously. Former FCC Commissioner Michael O’Reilly described download speeds greater than 25 Mbps as a “novelty” and explained why redefined broadband speeds “means longer waits for those who have no access or still rely on dialup service.” In a Senate Commerce Committee hearing, O’Reilly spoke again against changing the speed threshold. The difficulty ultimately comes down to the fact as mentioned earlier that the needs of individuals and households will be different. Should the definition of broadband speeds be minimal or aspirational?

A major challenge for this debate is the inaccuracy of our broadband availability maps. Data is collected at the level of blocks, not individual households, so it can be challenging to know precisely who has adequate access and who does not. The maps also fail to consider the costs of Internet service, which often act as a barrier to access even when the infrastructure is available. The FCC has informed ISPs that they need to provide updated data on their broadband availability and speeds, but the process of submitting the data to the FCC has faced numerous problems. Nor is the FCC the only entity that undertakes its mapmaking: Commissioner Brendan Carr expressed frustration that at least 133 federal funding programs are administered by 15 different agencies using different maps. With so many eyes on this issue right now, policymakers need better measurements and data.

New Demand for Faster Speeds

The pandemic has spotlighted many new demands for improved broadband speeds at different levels. To use Zoom to its full capacity, the company recommends a minimum speed of 3/3.8, demonstrating that even some households with FCC-defined broadband may miss out on teleworking opportunities. Telehealth visits, work video conferences, and online learning are just a few necessities pushing for enhanced access and the potential to upgrade the FCC’s definition. High-speed internet was one of the barriers keeping students from learning and patients from accessing virtual health care in 2020. Even before the pandemic, Pew Research Center found that 40% of schools and 60% of healthcare facilities outside metropolitan areas lacked adequate 25/3 access.

Virtual telehealth care requires sufficient bandwidth to function correctly. HealthIT.gov suggests a minimum of 10 Mbps download for small providers and 100 Mbps for hospitals and larger facilities. However, the speed at the level of individual users and institutions is not the only issue. Capacity (i.e., how much data the connection can carry simultaneously) is another major contributing factor. Targeted infrastructure deployment must meet the needs of both small and large health providers, especially in rural areas without access to sufficient speeds. In conjunction with SSRS, the Bipartisan Policy Center’s health team conducted a survey of nearly 2000 adults on their use of telehealth. Almost half of the respondents cited a lack of access to high-speed Internet as a significant obstacle to their telehealth experience.

Agriculture is another sector that is increasingly affected by broadband speed and availability. BPC’s energy team explored how rural broadband connectivity can help improve precision farming and create lasting economic benefits. A 2018 benefit-cost analysis of deploying broadband in Indiana estimated a ratio of 3-4, meaning that every $1 spent on broadband could generate between $3-4 in economic activity. The United States Department of Agriculture (USDA) published a report on rural broadband, in which it recognized the importance of broadband for precision agriculture, allowing farmers more direct control over the type and number of crops they grow. The Infrastructure Bill (covered in our last broadband post) specifically provides $2 billion for USDA broadband through the ReConnect program, and the Rural Development Title of the Farm Bill also addressed broadband extensively. As technology improves, the data input and output volumes will grow, requiring greater download and upload speeds.

Wired or Wireless?

For those who want or require the fastest available speeds, a direct, wired connection to fiber optic cable is one of the primary ways to go, with speeds ranging from 250-1000 Mbps. However, mobile broadband technology is becoming increasingly popular for both availability and convenience, especially among rural or remote households lacking wired broadband service. Mobile connections can help close the digital divide, but 3G is quickly becoming outdated, only reaching maximum download and upload speeds of 7/2 Mbps. Recently, telecommunications companies have faced pushback against upgrading their 2G and 3G networks to 4G and 5G. While sunsetting 2G and 3G frees up more spectrum for current-generation wireless standards, many Americans rely on 3G, either as a cheaper alternative or because there is no access to 4G/LTE/5G services. With the shutdown of 2G and 3G networks imminent, those users will need to find alternative options.

As our previous work discusses, there are several fixed wireless options that consumers may move to, but one in particular that has gained traction is Low Earth Orbit (LEO) satellites. Satellite companies are launching thousands of satellites into Earth’s lower atmosphere, which can achieve consistent coverage by ensuring that as one satellite moves out of an area, a new one moves in to replace it. While this is an exciting development and has the potential to bring Internet access to the entire globe, satellites face unique challenges compared to other telecommunications services. LEO satellites will need further policy considerations as they are deployed to facilitate their potential to connect underserved areas.

The Internet of Tomorrow

While there is significant disagreement around whether broadband should be considered a public utility in a post-covid world, nearly everyone agrees that it is essential for modern living and working. NTIA’s $65 billion will hopefully ensure that every American will have access to broadband that can meet their typical demands. Still, since technology constantly advances, consumer expectations often advance alongside it. Some interpret this watershed moment for broadband as the chance to future-proof our internet infrastructure, while others are more focused on the economic opportunities it will provide. Either way, many people throughout the US can expect internet infrastructure of significantly greater capacity in the years ahead.