On July 25, Chairwoman Jessica Rosenworcel shared with the other FCC Commissioners a draft Notice of Inquiry that would begin the process of raising the federal definition of broadband from 25/3 Mbps to 100/20 Mbps. In order for that to become the new definition, the FCC must work through the NOI process and eventually vote to adopt the higher speed definition.
This raises a question of the purpose of having a definition of broadband. That requirement comes from Section 706 of the Telecommunications Act of 1996 that requires that the FCC make sure that broadband is deployed on a reasonable and timely basis to everybody in the country. The FCC interpreted that requirement to mean that it couldn’t measure broadband deployment unless it created a definition of broadband. The FCC uses its definition of broadband to count the number of homes that have or don’t have broadband.
The FCC is required by the Act to report the status of broadband deployment to Congress every year. During the last week of Ajit Pai’s time as FCC Chairman, he issued both the 2020 and 2021 broadband reports to Congress. Those reports painted a rosy picture of U.S. broadband, partially because progress was measured using 25/3 Mbps definition of broadband and partially because the FCC broadband maps were rife with overstated speeds. The FCC has not issued a report since then, and I can only suppose there aren’t the votes in an evenly split FCC to approve a new report.
To give credit, Chairwoman Rosenworcel tried to get the FCC to increase the definition of broadband to 100/20 Mbps four years ago, but the idea went nowhere in the Ajit Pai FCC. At that time, 100/20 Mbps seemed like a reasonable increase in the definition of broadband. Most cable companies were delivering 100 Mbps download as the basic product, and a definition set at 100/20 Mbps would have made the federal statement that the speeds that most folks buy in cities is a reasonable definition of broadband for everybody else.
Chairwoman Rosenworcel is now ready to try again to raise the definition. Perhaps the possible addition of a fifth Commissioner means this has a chance of passing.
But this is now too little too late. 100/20 Mbps is no longer a reasonable definition of broadband. In the four years since Chairwoman Rosenworcel introduced that idea, the big cable companies have almost universally increased the starting speed for broadband to 300 Mbps download. According to OpenVault, almost 90% of all broadband customers now subscribe to broadband packages of 100 Mbps or faster. 75% of all broadband customers subscribe to speeds of at least 200 Mbps. 38% of households now subscribe to speeds of 500 Mbps or faster.
I have to think that the definition of broadband needs to reflect the broadband that most people in the country are really using. One of the secondary uses of the FCC broadband definition is that it establishes a goal for bringing rural areas into parity with urban broadband. If 75% of all broadband subscribers in the country have already moved to something faster than 200 Mbps, then 100 Mbps feels like a speed that is already in the rearview mirror and is rapidly receding.
When the 25/3 definition of broadband was adopted in 2015, I thought it was a reasonable definition at the time. Interestingly, when I first read that FCC order, I happened to be sitting in a restaurant that was lucky enough to be able to buy gigabit speeds and was sharing it with customers. I knew from that experience that the 25/3 Mbps definition was going to become quickly obsolete because it was obvious that we were on the verge of seeing technology increases that were going to bring much faster speed.
I think the FCC should issue two broadband definitions – one for measuring broadband adoption today and a second definition as a target speed for a decade from now. That future broadband target speed should be the minimum speed required for projects funded by federal grants. It seems incredibly shortsighted to be funding any technology that only meets today’s speed definition instead of the speeds that will be needed when the new network will be fully subscribed. Otherwise, we are building networks that are too slow before they are even finished construction.
Another idea for the FCC to consider could take politics out of the speed definition. Let’s index the definition of broadband using something like the OpenVault speed statistics, or perhaps the composite statistics of several firms that gather such data. Indexing speeds would mean automatic periodic increases to the definition of broadband. If we stick to the current way of defining broadband, we might see the increase in the federal definition of broadband to 100/20 at the end of this year and won’t see another increase for another eight years.
This is all just a thinly veiled money grab. Everyone talks big and fast about “finally” fixing the digital divide, between city/urban and rural broadband availability. But as long as they keep raising the broadband minimum speed then the moneys will continue to go to the denser urban areas. The big ISP’s can promise all kinds of things, and make headlines about getting broadband to rural folks but it’s all smoke and mirrors. If the minimum broadband speeds would be kept at 100×20, then massive areas, all of the denser populated areas, would not be eligible to receive grant money. Therefore it would pour into the undeserved areas. If they raise it, all that does is allow the ISP giants to pull massive money grabs to upgrade their dense, high paying areas yet again and default, again, on any real rural expansion. But who am I to complain, this way we don’t have to compete with the giant players in our market. And our people are happy.
Maybe I am just oversimplifying, There are two questions here. The first is the minimum “useful” speed. What speed do you need to give that is truly useful. I would argue that the upload speed is the thing to worry about. 10 mbits of upload, while not ideal, does at least allow for single stream video conferencing pretty effectively. 3mbits, not so much. Giving a bunch of people a service that is of little value is a waste of money.
The second is cost / benefit related. You have a pile of money. What is the biggest bang for the buck? The 100/20 definition will cost more per home than the 25/3 definition. The cost of the tech is not linear with the speed. It has pretty steep jumps along the curve. You want to set the definition to maximize speed per public dollar spent. How many homes can I serve (with the same pile of money) at 100/20 vs 100/10 vs 50/20 vs 100/25.
What is clear to me is that 3mbits up is not broadband because it wont reliably support a video call. Less focus on numbers and more on use cases, which will drive numbers, is probably a good place to start.
I don’t think usage increases have a linear growth. For many years, growth was primarily just the increase of people using the internet day to day. Then streaming media arrived on the scene ie Netflix and that dramatically increased the download requirements. Average usage then went back to it’s slow creep upward until the covid+work/school-from-home surge which increased upload needs (but not really download needs) and again we’re at a slow and steady increase that I presume is based on an increased number of connected devices and backgroud updays.
25/3 is bad, primarily because the *3* doesn’t handle the most recent usage change. if this were 25/10 then it would be much more funcitonal but still occasionally limiting.
I think the 100/20 number is quite good and appopriate as a definition. I don’t think it’s appropriate to try to cover the next usage change because we don’t know what that will be. Rational budgeting would have us pick an appropriate number that should last through the next cycle and then up it again when necessary.
This is similar to the ‘fiber first’ policy we are doing now. We do not have the funds, manpower, or even product to bring fiber to everyone even if it was the singular superior product and by pursing this we are not able to pull nearly as many people out of the 3/1 DSL desert they are in.
Trying to build the 10 years+ requirements into today’s service is incredibly costly.
Today’s newly-launched tech across the board from XGSPON, DOCSIS4, WIFI6 based FWA, and some other more exotic radios, can all do 100/20 economically to many many people and leave room for the mid and high speed plans on top and a 5-10 year lifespan on most of this gear seems to fit a 100/20 for 10 years definition really well, and by well I mean economically and quick to deploy.