Congress is about to make critical decisions about the future of internet access and speed in the United States. It has a potentially once-in-a-lifetime amount of funding to spend on broadband infrastructure, and at the heart of this debate is the minimum speed requirement for taxpayer-funded internet. It’s easy to get overwhelmed by the granularity of this debate, but ultimately it boils down to this: cable companies want a definition that requires them to do and give less. One that will not meet our needs in the future. And if Congress goes ahead with their definition—100 Mbps of download and 20 of upload (100/20 Mbps)—instead of what we need—100 Mbps of download and 100 Mbps of upload (100/100 Mbps)—we will be left behind.

In order to explain exactly why these two definitions mean so much, and how truly different they are, we’ll evaluate each using five basic questions below. But the too long, didn’t read version is this: in essence, building a 100/20 Mbps infrastructure can be done with existing cable infrastructure, the kind already operated by companies such as Comcast and Charter, as well as with wireless. But raising the upload requirement to 100 Mbps—and requiring 100/100 Mbps symmetrical services—can only be done with the deployment of fiber infrastructure. And that number, while requiring fiber, doesn’t represent the fiber’s full capacity, which makes it better suited to a future of internet demand. With that said, let’s get into specifics.

All of the following questions are based in what the United States, as a country, is going to need moving forward. It is not just about giving us faster speeds now, but preventing us from having to spend this money again in the future when the 100/20Mbps infrastructure eventually fails to serve us. It’s about making sure that high-quality internet service is available to all Americans, in all places, at prices they can afford. High-speed internet access is no longer a luxury, but a necessity.

Which Definition Will Meet Our Projected Needs in 2026 and Beyond?

Since the 1980s, consumer usage of the internet has grown by 21% on average every single year. Policymakers should bake into their assumption that 2026 internet usage will be greater than 2021 usage. Fiber has capacity decades ahead of projected growth, which is why it is future-proof. Moreover, high-speed wireless internet will likewise end up depending on fiber, because high-bandwidth wireless towers must have equally high-bandwidth wired connections to the internet backbone.

In terms of predicted needs in 2026, OpenVault finds that today’s average use is 207 Mbps/16 Mbps. If we apply 21% annual growth, that will mean 2026 usage will be over 500Mbps down and 40Mbps up. But another crucial detail is that the upload and download needs aren’t growing at the same speeds. Upload, which the average consumer used much less than download, is growing much faster. This is because we are all growing to use and depend on services that upload data much more. The pandemic underscored this, as people moved to remote socializing, remote learning, remote work, telehealth, and many other services that require high upload speeds and capacity. And even as we emerge from the pandemic, those models are not going to go away.

Essentially, the pandemic jumped our upload needs ahead of schedule, but it does not represent an aberration. If anything, it proved the viability of remote services. And our internet infrastructure must reflect that need, not the needs of the past.

The numbers bear this out, with services reporting upstream traffic increasing 56% in 2020. And if anything close to that rate of growth in upload demand persists, then the average upload demand will exceed 100Mbps by 2026. Those speeds will be completely unobtainable with infrastructure designed around 100/20 Mbps, but perfectly within reach of fiber-based networks.

Notably, all the applications and services driving the increased demand on upstream usage (telehealth, remote work, distance learning) are based on symmetric usage of broadband—that is 100/100 Mbps and not 100/20 Mbps. And future cloud-based computing services are predicted to actually need higher upload speeds than download speeds to function.

Which Definition Will Increase Upload Speeds Most Cost-Effectively?

With upload demand skyrocketing, networks will have to improve their capacity. However, the cable infrastructure that will be maintained by a 100/20 Mbps definition is already reaching its capacity. That means that, in order to upgrade, companies will eventually have to start replacing the old infrastructure with fiber anyway. Or, they will be stuck delivering below what Americans need. The same is true for wireless internet.

In other words, the only way to upgrade a non-fiber, 100/20 Mbps network is to connect it with fiber. There is just nowhere for the current infrastructure to go. Updating with fiber now saves everyone the cost of doing minor upgrades now and having to do fiber in a few years. Slow networks ultimately cost more than just going straight to fiber because they ultimately have to be replaced by fiber anyways and become wasted investments.

Furthermore, once on fiber, increasing your speed comes much more cheaply, since the hardware at the ends of the fiber connections can be upgraded without digging and laying new cables. You can see this with the financial data from Chattanooga’s municipal fiber entity in 2015 when they upgraded from 1 gigabit to 10 gigabits. They did not experience a substantial increase in costs to upgrade at all.

Which Definition Will Deliver Gigabit Speeds?

For the same reason 100/20 cable and wireless systems can’t easily improve their upload speed, they can’t also turn around and deliver gigabit speeds. Meanwhile, the same fiber network able to deliver 100/100 Mbps is actually also capable of also delivering 1000/1000 Mbps and 10,000/10,000 Mbps with affordable upgrades to its hardware. 80,000/80,000 Mbps is already possible now over the same fiber wire, though the price of the hardware remains high. As the price comes down, 80 gigabit symmetrical could become the next standard for fiber networks. Wireless connected with fiber benefits from these gains with the only limitation being the amount of available spectrum they have for wireless transmission.

Which Definition Will Give Americans an Affordable Option That Meets Their Needs Over Time?

There is zero chance a network built to deliver 100/20 Mbps that isn’t premised on fiber can offer a scalable, low-cost solution in the future, for all the reasons listed above. Capacity constraints on cable and non-fiber-based wireless drastically limit the extent to which they can add new users. Their solution is to offer significantly lower speeds than 100/20 Mbps to minimize the burden on their capacity-constrained network. But a fiber network can share the gains it makes from advancements in hardware because it does not experience a new cost burden to deliver a scalable solution. This is why Chattanooga was able to give its low-income students free 100/100 Mbps internet access during the pandemic at very little cost to the network.

Which Definition Makes the U.S. Globally Competitive?

Advanced markets in Asia, led by China, will connect total of 1 billion people to symmetrical gigabit lines. China years ago committed to deploying universal fiber, and it is rapidly approaching that goal. The U.S. could choose to do the same. However, if it instead chooses to upgrade some cable networks and push some slow wireless connectivity out to communities at 100/20 Mbps, our capacity to innovate and grow the internet technology sector will be severely hindered. After all, if the U.S. market is not capable of offering a communications infrastructure capable of running the next generation of applications and services due to slow obsolete speeds, then those applications and services will find their home elsewhere. Not only will this impact our ability to attract a technology sector, but all related industries dependent on connectivity will be relying on speeds vastly inferior to gigabit fiber-connected businesses.

In each one of these questions, it is clear that the government needs to invest in fiber infrastructure, which means defining what technology gets taxpayer dollars at 100/100 Mbps. While the existing monopolies would like to get that money for infrastructure they don’t actually have to build—old cable lines that can meet the 100/20 Mbps definition—that is doing a grave disservice to Americans.