As part of a $2.3 trillion infrastructure proposal, President Joe Biden is pushing Congress to spend $100 billion fixing a problem that mostly doesn't exist: widespread lack of access to broadband internet.
The Federal Communications Commission (FCC) estimates that there were about 14.5 million Americans, living in an estimated 4.3 million households, that lacked access to broadband internet at the end of 2019. That's a serious but narrow problem that's already being addressed by a combination of private and public efforts. New technologies like SpaceX's low-orbit satellites can beam broadband internet to homes even in far-flung rural places, and the FCC has already budgeted more than $9 billion over the next 10 years as part of what the agency says is the "biggest single step ever…toward closing the rural digital divide." The number of Americans without broadband access fell by 20 percent in 2019, according to an FCC report published in January, and it's likely that the total is significantly less today than at the end of 2019.
But Biden's infrastructure plan suggests a major change to what counts as "broadband" internet. As a result, as many as 64 million American households could suddenly appear to lack adequate online speed—even though nothing about their current services would change.
With a simple bureaucratic adjustment, the Biden administration could manufacture the appearance of a market failure where one plainly does not exist, opening the door for an expensive taxpayer-funded intervention to subsidize government-run internet boondoggles. Critics charge that Biden's plan will crowd out private investment in broadband infrastructure while steering money to parts of the country where residents already enjoy state-of-the-art connection speeds. Instead of targeting a small amount of funds toward the truly needy who lack access to fast internet, Biden could end up spending $100 billion only to make the digital divide larger than ever.
"It's not going to achieve the goal of bridging the digital divide in America," Deborah Collier, a vice president at Citizens Against Government Waste, a fiscally conservative nonprofit, tells Reason. "It's just going to throw more money at cities and localities that already have broadband."
To understand exactly why changing the definition of broadband matters, you have to first understand a little bit about how internet speeds are measured and what currently counts as broadband.
Since 2015, the Federal Communications Commission (FCC) has defined a broadband connection as internet access with download speeds of at least 25 megabits per second and upload speeds of at least three megabits per second. In layman's terms, that's fast enough to stream a high-definition movie in the living room while three other people check Facebook, send email, or do some online shopping simultaneously.
A so-called "25/3" connection might not be fast enough for all households, but it is a standard that's meant to reflect the needs of most Americans. The higher download speeds relative to upload speeds are a reflection of what consumers demand—because the vast majority of internet usage involves downloading, largely due to the huge demand for streaming video. Even after a surge in upload demand during 2020—thanks to all those Zoom calls and everyone working from home—about 93 percent of all internet traffic involves downloading content, according to data from the Internet and Television Association (NCTA), an industry group.
The White House's fact sheet for Biden's infrastructure plan calls for "building 'future proof' broadband infrastructure." That term has a specific meaning in the broadband world.
"'Future proof' networks often means symmetric speeds," explains Jeffrey Westling, a technology and innovation policy fellow at the R Street Institute, a free market think tank. In other words, Biden isn't just calling for faster speeds, but equal speeds between uploads and downloads. Advocates for government-mandated "future proof" networks typically aim for a 100/100 standard—that is, 100 megabits per second in both uploads and downloads.
As a practical matter, those standards are just silly. A typical Zoom call uses about 1.5 megabits per second in upload bandwidth. With a 100/100 connection, "you could have ten kids pretending to do Zoom school but actually doing TikToks while you're in the other room pretending to work but actually watching Netflix, and still not run out of bandwidth," writes Scott Wallsten, president of the Technology Policy Institute.
Politically, however, defining broadband connections as "100/100" would mean two significant things.
First, it would radically change the number of Americans who currently have a "broadband" connection to bolster calls for government intervention in the marketplace. According to data from the Technology Policy Institute, there are approximately 4.3 million American households that do not currently have access to a 25/3 internet connection. But there are more than 64 million households—about 40 percent of the country—that don't currently have access to a 100/100 connection.
"If 100/100 Mbps or asymmetrical speed similar to this threshold was adopted as the minimum standard, more Americans living in urban areas that already have reliable broadband would become underserved," writes Will Yepez, a policy associate at the National Taxpayers Union Foundation.
The Biden administration wants to argue that the $100 billion broadband effort is the 21st century equivalent of the federal government's electrification efforts during the 1930s, which helped bring power to wide swaths of the country. "Broadband is the new electricity," reads the White House's fact sheet on the proposal. That comparison looks a lot better if the Biden administration can say there are 64 million households lacking sufficient internet connectivity—even though almost all of them already have access to broadband-level speed.
Secondly, this maneuver would allow for a bit of political favoritism by prioritizing fiber optic internet services over the alternatives that have sprouted up. There are lots of different services that can offer broadband internet at 25/3 speeds: cable connections, fixed wireless, and even new low-orbit satellite systems like the Starlink service recently launched by Space X. But there's really only one way to deliver reliable 100/100 speeds, and that's via fiber optic cable.
Of course, fiber internet is also one of the more costly and difficult types of internet service from an infrastructure perspective. It requires digging trenches, laying pipes, and physically connecting each and every household. It's a practical impossibility for rural parts of the country and tends to be a more expensive option elsewhere. That's a problem if you're in the business of providing internet service to people as quickly and cost-effectively as possible, but it's an opportunity if you're looking to spend lots of government money on a high-profile infrastructure plan.
"It's actually going to harm areas of the country that do not even have basic, minimum broadband service," says Collier. "The funding is going to be redirected away from those areas and put into areas of the country that already have basic or better broadband service right now—just to upgrade those networks to 100/100 speed."
Put another way: if there are suddenly 64 million American households without access to "good" internet under the Biden administration's new definition, then broadband providers will focus their upgrade efforts on areas with dense populations. That will likely crowd out efforts to reach households that still lack even 25/3 connections and leave far-flung rural areas behind, again.
"Companies would be more likely to invest in these now more profitable areas rather than focus on those who truly lack reliable access to high-speed internet," says Yepez.
The Biden administration could end up spending $100 billion to accomplish the proverbial bridging of the digital divide—and then discover that the divide has only gotten larger.
A critic might point out that using the 25/3 standard is no less arbitrary than the 100/100 standard, and in some ways that's correct. The FCC has changed its definition of what counts as "broadband" on three occasions already. The first standard, in place from 1996 through 2010, required at least 200 kilobits per second upload and download speeds. From 2010 through 2015, that was upped to 4 megabits per second for downloads and 1 megabit per second for uploads. It's been six years since the 25/3 standard was adopted, so maybe it's time for another change?
"Any new definition should be based on evidence and take into account the tradeoff between the expected costs of achieving those speeds versus the benefits of the increase," says Wallsten.
The costs of the 100/100 switch are apparent—crowding out investment in non-fiber broadband, a huge increase in government spending to speed up already fast internet in many places—but the benefits are murky at best.
Keep in mind that close to 60 percent of American households already have access to 100/100 speeds, if they choose to pay for them. Most don't. Those that do rarely use that much bandwidth.
The Wall Street Journal and researchers at Princeton University and the University of Chicago teamed up last year to study the internet use of 53 Journal staffers—people who likely use the internet more heavily than most Americans. The eight users in the study who had connections with download speeds of at least 100 megabits per second used, on average, 7.1 megabits per second of their capacity.
"People who paid for even faster speeds still streamed video at about the same speeds as everyone else," the Journal concluded. The benefits of 100 megabit connections are "marginal at best, according to the researchers," and the evidence suggests that most Americans who are paying for internet connections that fast are being "oversold."
And that's just on the download side of the equation—the direction that 93 percent of all household internet traffic travels. It's almost impossible to imagine a scenario where an ordinary American would use 100 megabits of upload capacity.
Of course, the right internet connection speed is going to vary from household to household and user to user. For some people, paying for 100/100 connections might make sense. But there seems to be little evidence to support the idea that 100/100 should be the FCC's standard for what counts as passable internet access.
If broadband access is indeed "the new electricity," then Biden's proposal looks less like hooking up a power line to every house and more like a mandate that taxpayers fund the construction of a hydroelectric dam in every backyard.
There is one final issue here. Having access to broadband is not the same as actually using it. If there is a fiber optic cable running past your house or if you live in an area covered by fixed wireless or low-orbit satellite internet, your household is counted among those that currently have access to 25/3 broadband. That's not the same as actually paying to use it—something that many Americans either can't afford or choose not to do.
If the Biden White House was interested in ensuring that more Americans could use broadband—as opposed to creating phantom justifications for spending huge sums of money on municipal broadband—one way to do it would be to subsidize the cost of internet access for low-income households. That's what some major broadband providers have been urging the White House to do in recent weeks, which both spares non-fiber broadband companies from being declared obsolete and expands their customer pool.
It's right to be skeptical anytime an industry says that the solution to a social problem is more subsidies for itself. Still, the broadband providers are at least asking the right question. Namely: How do we get more Americans connected to the existing broadband infrastructure that's already built and available for use?
The Biden administration, meanwhile, is trying to spend lots of money to solve a different problem—one that doesn't even really exist, at least not in a way that demands $100 billion in new federal spending—and already seems to be setting itself up for failure. The White House's fact sheet about the infrastructure plan says it "prioritizes support for broadband networks owned, operated by, or affiliated with local governments, nonprofits, and co-operatives," because those providers have "less pressure to turn profits."
That seems like a clear indicator that municipal broadband operations will get to move to the front of the line when the Biden administration starts handing out piles of cash to solve the broadband connection problems that don't actually exist in most places. But municipal broadband has been a major boondoggle in many places where it has been tried—the Taxpayers Protection Alliance has a list of more than 200 taxpayer-funded internet projects that are deep in debt or have been abandoned. Even if municipal broadband didn't have an established track record of failure, it seems completely unnecessary for the federal government to prop up new competitors to existing broadband providers in places that already have fast internet service.
In some ways, this coming debate over the definition of broadband is likely to mirror the broader debate over what, exactly, should count as infrastructure. The Biden administration and its allies are pushing the idea that everything from health care programs to job training to child care is infrastructure—while only about half of Biden's $2.25 trillion spending bill is aimed at traditional infrastructure priorities like highways, bridges, rail, and pipelines.
Everything is infrastructure. Nothing is broadband.
The Biden administration should focus on the few remaining pockets of the country where high-speed internet isn't available, rather than futz with the FCC's definition of broadband in order to justify spending billions of dollars so residents of urban areas with already fast connections can stream 10 movies at the same time.