Monthly Archives: March 2024

Carrier Hotels

Today’s blog talks about a critical part of the broadband network that most people don’t know about – carrier hotels. These are locations that have been created for the specific purpose of allowing carriers to connect to each other.

 

The need for carrier hotels became apparent in the year after the passage of the Telecommunications Act of 1996. That new law allowed local competition for telephone service. New competitive exchange carriers (CLECs) had to tie their network into the public switched telephone network (PSTN), which was largely controlled by the big Bell Telephone companies and a few others.

 

The big telcos purposefully made it ponderous, expensive, and time-consuming to collocate inside a telco central office. The interconnection agreements between a CLEC and a telco were several inches thick, and the rules related to collocation and interconnection were over fifty pages long – and the telcos meticulously adhered to every rule. The telco got to approve the use of every piece of equipment that went into their space – and that meant only using the brands of gear that the telcos used. CLECs had to pay an expensive telco employee to accompany them every time they visited their equipment. It could easily take nine months to a year to establish a single new collocation.

 

A few smart entrepreneurs came up with the idea of creating a collocation space very near to major telco offices. The language in the 1996 Act allowed CLECs to request collocation at any technically feasible location. It took a few fights at state commissions to allow collocation outside of Bell offices, but regulators universally agreed that was the intention of the 1996 Act. Once the use of external collocation sites was blessed, carrier hotels sprang to life and thrived.

 

Carrier hotels make money by selling collocation to carriers. All types of carriers come to the carrier hotels – long-distance carriers, CLECs, ISPs, wireless carriers, cable companies, major content providers, and specialty carriers of all types. A big carrier hotel in a major city houses hundreds of domestic and international carriers.

 

There are a number of benefits for carriers to locate in a carrier hotel. The most obvious is to make it easy to connect to other carriers and services. In today’s environment, an added benefit is the high degree of security at the typical carrier hotel. Carrier hotels are convenient places to exchange vast amounts of data.

 

The typical carrier hotel is also connected to multiple major fiber routes, so it’s an easy place for carriers to jump onto transport routes provided by fiber owners or companies that have leased or bought fibers on major fiber routes.

 

Another industry function provided by the carrier hotel is to create competition that holds down prices. There is fierce competition in the biggest carrier hotels for bandwidth, transport, and other services, and prices are kept competitive by having multiple carriers willing to provide the same service to carriers.

 

A lot of the traffic that we think of as the “Internet’ changes hands in carrier hotels. The direct fiber connections within the hotel allows for keeping bits moving in the most efficient manner. A given carrier might interconnect with a dozen or more other carriers to hand-off specific subsets of data.

 

It’s rare for a carrier hotel to hit the news. The most famous event at a carrier hotel occurred at the carrier hotel at 60 Hudson Street in Manhattan. That location got knocked out of service by the attack on the neighboring Twin Towers on September 11. The carrier hotel went dark, which contributed to widespread broadband and cellular outages, particularly in New York City. The good news was that the carrier hotel was up and running by 9:00 AM the following morning.

Massive African Internet Outage

Eight countries in West Africa reported Internet outages after damage was reported to four different undersea fiber cables. The most affected countries are Ivory Coast, Liberia, and Benin, with additional problems caused in Ghana, Nigeria, and Cameroon. Lesser impact was felt in South Africa, Senegal, and Portugal. There has been no official word on what caused the problem, but it’s likely due to the shifting of the seabed due to seismic activity.

It’s going to take between five weeks and several months to fix the cables. The four affected cables are owned by West Africa Cable System, MainOne, South Atlantic 3, and ACE. Several repair ships have already been dispatched to find and fix the problems.

The process of fixing an undersea cable is similar to that of fixing any damaged fiber – just complicated by the fact that the fiber is located deep in the ocean. Once the location of the cut or damage has been determined, ships use grapnels to snag the cable and pull it to the surface. If the fiber has been cut, then each end must be located and brought to the surface. If a fiber has only been damaged, it’s brought carefully to the surface in order to not cause additional damage. Once at the surface, any bad sections are cut out, and new fiber spliced.

Undersea fibers have a rough existence and there is an underseas fiber cut somewhere in the world every three days. It’s unusual to have multiple fibers in the same region at the same time. There is a fleet of ships that travel the world continuously to fix underseas fiber cuts or bends.

Undersea fibers are not large. Near to shore the fibers are encased in thick shielding to protect against cuts by boats and anchors. But the largest undersea fibers only include 16 fibers and are slightly smaller than a garden hose.

This map shows the various undersea fibers in place today. In much of the world, almost all of the broadband in and out goes through the undersea fibers. This is true for Africa and for large islands like Australia and Ireland. There are not many landline fiber routes that traverse the entire African continent, although there are regional middle-mile fiber routes. The undersea fibers also carry almost all data that moves from continent to continent. If you pull up a web article from Europe or India, that connection almost certainly came through an undersea fiber.

These cuts show how fragile the world’s Internet connections are. Threats are often made by bad actors to cut undersea fibers. Most recently, the Houthi rebels in Yemen threatened to cut the fibers that pass through the Red Sea. After an attack on oil and gas pipelines in the Baltic Sea in 2022, Russia threatened to retaliate by cutting undersea fibers. The protection of undersea fibers was a prominent issue at a NATO summit later that year. The fibers feeding broadband to Taiwan have been cut several times, with a suspicion that China was behind the cuts.

Folks who live where there are large landline networks tend to take Internet access for granted. But there are many places around the world where an outage can be crippling to an entire country or region and last for weeks or months – similar to the horror stories we hear from time to time about rural telcos!

Telcos Shedding Jobs

I heard a chilling story recently. AT&T apparently notified a bunch of employees in Los Angeles that their jobs are being eliminated and that they need to report to other cities like Dallas or lose their job. Many of these employees were relocated to Los Angeles in the last five or six years, and the company paid for that past relocation. Employees now must move at their own expense. I was told the same thing was happening in other AT&T markets across the country. This is a particularly callous way to eliminate employees, and AT&T is clearly trying to induce employees to resign to avoid paying severance. This is not the kind of behavior that would normally be expected from a large corporation. It certainly tells the remaining employees of the company that they are not valued.

There is rarely a month that doesn’t go by without hearing that one of the big telcos is laying off a group of employees somewhere. The story piqued my interest, and it took only a little research to see that telcos have steadily been eliminating staff while the biggest cable companies have not.

Consider the following chart that shows employment at the biggest ISPs and carriers since 2018.

2018 2023 Change
AT&T 268,220 150,500 -44%
Verizon 144,500 105,400 -27%
Lumen 45,000 28,000 -38%
T-Mobile 80,500 67,000 -17%
Comcast 184,000 186,000    1%
Charter 98,000 101,100    3%

It’s not easy to make sense of the staffing changes at the various carriers. Consider some of the big trends at each company since 2018.

Some of the staff reductions at AT&T can be justified since the company suffered from several disastrous investments. The biggest was buying Time Warner Media and spinning it off just three years later to Discovery with a huge loss. The company had another big failure from its purchase of DirectTV. While AT&T flourished from 2018 to 2023 in adding cellular customers, competition dropped the average revenue per customer over that time period. AT&T lost only 3% of its net broadband customers over that period while it has been transitioning from copper to fiber.

Verizon has a similar story of making bad investments in AOL and Yahoo. Due to the big surge of FWA cellular broadband and good sales in FiOS, Verizon has 54% more broadband customers today than it had in 2018. Verizon also thrived and grew cellular customers during this period.

Everybody has likely heard Lumen’s story. The company has struggled since it was spun off from AT&T as US West. The company divested it’s copper assets in twenty states and recently announced more layoffs.

T-Mobile is an interesting case. Cellular customer additions have been sluggish since it merged with Sprint. It recently added 4.8 million FWA broadband customers. The layoffs at T-Mobile seem to be clearly aimed at improving the bottom line – even though one of the big promises made to employees with the Sprint merger was that it would create new jobs, not lose jobs.

Comcast has thrived in everything except cable TV. Since 2018, the company added 5 million broadband customers and 6.5 million cellular customers.

Charter also did well except with cable TV. Charter added 5.3 million broadband customers since 2018 and 7.8 million cellular customers.

The bottom line of my quick analysis is that telcos have been reducing staff at a much greater pace than can be justified by looking at the overall trends of each business. I have to wonder how Comcast and Charter are going to react to the sudden slump in broadband growth? Will they now start shedding employees like the telcos have done?

The BEAD Subsidy of Utilities

When ISPs are asked about the impediments they encounter for building new fiber networks, they almost always list pole issues at or near to the top of the list. Why are poles of such big concern?

Building aerial fiber means putting the fiber on poles. Most poles are owned by electric utilities, although some belong to telephone companies or municipalities. Invariably, some poles have to be replaced in order to add a new fiber line. This mostly occurs when there is not enough room on an existing pole to provide for the required safe distance between wires that are required by national safety codes. The common fix for this problem is to install a taller pole and move existing wires from the old pole to the new one – the process can be agonizingly slow since the pole owner has to coordinate with existing attachers. Worse is the expense, since the new attacher has to pay the full cost of replacing the pole and moving the old wires.

But poles often must be replaced because they are obsolete or in bad condition. The FCC recently issued new pole rules which designate poles that are out of safety compliance or that are already scheduled to be replaced as ‘red-flagged poles’.  Under the new rules, a new fiber attacher will not be responsible for the full cost of replacing a red-flagged pole. We’ll have to wait to see how the new rules play out in actual practice since pole owners are likely to argue vehemently about when to assign the red-flag designation.

Regardless of why poles have to be replaced, the BEAD grants are going to be used to replace huge numbers of poles. I don’t have any easy way to estimate the number of poles that will be replaced, but it wouldn’t be surprising if it is in the millions. It’s fairly normal for aerial construction to require the replacement of 5% to 10% of poles. In places where the poles are in bad shape, this can be a lot higher.

The cost of replacing poles is built into the construction costs of adding new aerial fiber. Since BEAD will pay up to 75% of the cost of fiber construction, that means BEAD will pay up to 75% of the cost of replacing poles.

That is a huge windfall for electric utilities. In far too many cases, the poles that will replaced by BEAD should have already been upgraded and replaced by the pole owners. Depending on the local conditions in different part of the country, poles typically are expected to last from 30 to 50 years. Unfortunately, it’s not unusual for utilities to keep poles far past the expected economic life.

Fiber providers have been yelling for years that the process is unfair. They are routinely being asked to replace poles that have held other wires for decades. During that time, the pole owner collected pole attachment fees – which should have more than covered the periodic replacement cost of the poles.

Regulators blew it a long time when they didn’t require pole owners to put pole attachment fees into a sinking fund that could only be used to repair poles after storm damage or eventually replace poles. But pole owners count pole attachment fees like any other revenue stream, and it’s quickly used for something else – and in many electric companies might even use the fees to pay dividends to shareholders.

Some states have recognized the pole issue as a big problem and used a solution that has me scratching my head. These states have taken some funding from CAREs or ARPA and created a pole replacement fund to replace the worst poles in the state. This makes some sense if bad poles are a major impediment to building fiber.

But you don’t have to think about this very hard to realize that a pole replacement fund is exactly the wrong economic incentive to give to pole owners. If I’m a pole owner and my State will fund pole replacement, I’m going to cut way back on the poles that I’ll voluntarily replace. I’d let the inventory of bad poles build up and then ask for more pole replacement money. To use common political vernacular – a pole replacement fund fosters welfare for electric companies.

The FCC had an opportunity in its recent pole order to take an important step, which was to require all pole owners to create a public inventory showing the age and condition of every pole. Such a database would make it easy to spot the pole owners who are not replacing poles as needed. Without this basic data, it’s incredibly hard for regulators to force pole owners to do the right thing.

The BEAD grants will bring a lot of broadband to places that need it. It’s just a shame that a lot of the funding will be used to replace the pole inventory of utilities that failed to invest in their own networks.

Competition in MDUs

FCC Chairwoman Jessica Rosenworcel announced plans to introduce a Notice of Proposed Rulemaking that would expand customer choice in apartments, condos, public housing, and other multi-tenant buildings.

The NPRM will cover two new areas of regulation. The first is that tenants will be able to opt out of bulk billing arrangements where landlords build broadband or cable TV into the price of rent. It’s an interesting idea and will benefit tenants who don’t want to buy the forced services. Some landlords have padded profits for years with practices like distributing a bulk satellite TV feed to tenants at a markup. In a country where less than half of homes still want to watch linear cable programming, it seems likely that a whole lot of tenants will opt out of that service, given the option.

Allowing people to opt out of broadband is trickier. There are certainly landlords who buy bulk broadband and mark it up as a moneymaker. But there are many landlords today that recognize that having gigabit fiber broadband is a great marketing tool for their apartments. These landlords buy a bulk broadband connection and provide broadband to tenants at a price lower than what the tenants could find on the market. Nobody should be forced to buy something they don’t want, but if enough tenants opt out of this arrangement, the landlord will likely have to raise the rates for everybody else. This is a dilemma for landlords, and it could result in landlords ceasing to offer cheap broadband. That might disadvantage more households than it benefits.

It gets even trickier to think about the many arrangements that have been made to bring bulk broadband to public or low-income housing. If tenants can opt out of these arrangements to save a little money or lower rent, it could kill the model of bringing broadband to public housing.

The NPRM also proposes to “increase competition for communications service in these buildings by making it more profitable for competitive providers to deploy service in buildings where it is currently too expensive to serve consumers because tenants are required to take a certain provider’s service.”

This proposal really piques my interest. I’ve been thinking about it, and I can’t think of anything that regulators can do to make it cheaper to physically bring broadband into MDUs. Landlords often have restrictions on the way that ISPs can build infrastructure. For example, they might require outdoor drops to be buried. They might not allow the installation of communications devices like fiber ONTs on the outside of a building. They might have restrictions on not allowing open wiring in public spaces. They might require an ISP to get its own electric feed and meter.

It’s hard to think that the FCC could eliminate these kinds of restrictions. The FCC has run afoul in the past when creating rules that conflict with the rights of property owners for issues like inside wiring and antenna placement. It seems unlikely that the FCC can find a way to reduce the cost of physical and aesthetic restrictions, and I’m looking forward to see what the FCC has in mind.

There is nothing that the FCC can do to fix the market behavior of ISPs. New rules that improve the chance of ISPs to compete in large MDUs will be welcome news to some ISPs – but new rules might drive other ISPs away from the MDU market. It’s hard to think that these new rules will have a significant impact on the majority of MDUs in the country, which are eight units or less. It’s hard to think of any rules that will boost competition in smaller MDUs.

However, the FCC is right to keep trying. The National Multifamily Housing Council says there are currently 23 million apartment units in the U.S. A lot of current broadband rules and subsidies have been clearly aimed at improving broadband in single family homes. The FCC has tried for several decades to improve competition in MDUs trying rules that did things like making it harder for landlords to have exclusive arrangements with an ISP. But landlords and ISPs have always quickly found loopholes in every set of new rules.

Barriers to Grant Funding in Minnesota

There is a lot of broadband legislation introduced every year in State legislatures, and most States only pass a few of the dozens of bills proposed each session. Every once in a while, a particularly egregious or curious bill gets introduced.

The Blandin Foundation broadband blog recently posted the text of proposed legislation in Minnesota for House Bill HF4659 that makes it much harder to build projects with grant funding in the state. The stated purpose of the bill is to provide protection for labor and to promote safety standards. But the practical purposes of the bill is to make it incredibly hard for ISPs to use broadband grant funds.

The proposed legislation would apply to both the State’s Border-to-Border grants as well as for federal BEAD grants.

For Minnesota Border-to-Border grants, all employees working to build grants, including employees of ISPs, contractors, or subcontractors, must:

  • Be paid a prevailing wage.
  • Be provided at least 80 hours of skill training annually at no cost to the employee. At least 40 of the hours must be hands-on instruction.
  • Any employee working more than 500 hours per year must be provided with employer-paid family health insurance.
  • Any employee working more than 500 hours per year must be provided with a post-retirement benefit equal to at least 15 percent of total taxable wages.
  • If the grant office doesn’t get applications from ISPs that meet these rules, it is required to solicit applicants that will meet the new rules.

These new rules would apply to 50% of state grants awarded this year, 60% awarded in 2025, and 70% awarded in 2026 and beyond.

There are even more stringent rules for anybody applying for a BEAD grant:

  • Applicants must use a directly employed workforce instead of a subcontracted workforce to perform placing, splicing, and maintenance work on networks. Public applicants for grants can meet this requirement by partnering with an ISP that uses a directly employed workforce.
  • The criteria established by the legislation must represent 25% of the scoring criteria for winning the BEAD grant.
  • Anybody winning a BEAD grant must publicly disclose a lot of information about wages twice per year.
  • An ISP violating any of the rules would be barred from future grants.
  • Locating existing buried facilities must be done by a safety-qualified underground telecommunications installer. All construction within 10 feet of existing telecommunications infrastructure must be done by a safety qualified telecommunications installer. No less than two safety-qualified underground telecommunications installers must be present at all times during directional drilling.

This bill is such a major departure from the way that networks are built that it looks to be intended to drive ISPs away from seeking grants. Throughout the industry, network construction is largely performed by small crews of sub-contractors. These companies have never been required to carry employer-paid health insurance or post-retirement benefits. Even those that do would likely not give these benefits to employees who work as few as 500 hours in a year. Even large telcos and cable companies don’t give health insurance and retirement benefits to part-time employees.

The requirement that all technicians must be employed by the grant recipient would kill most applicants from pursuing BEAD grants. Even giant ISPs use contractors to build networks. While there are likely a few ISPs who could use 100% employees to fulfill a grant, my guess is that this requirement would drive away most potential BEAD applicants.

It’s hard to understand the motivation for the bill since it doesn’t seem to benefit any particular class of ISP. Instead, the legislation just adds more costs to ISPs willing to accept grants and makes it more expensive to build networks. The practical result would be that very few ISPS would be willing to pursue grants in the state. The motivation behind the legislation baffles me.

Remote Learning and Preschoolers

A recent article in the MIT Technology Review described the benefits that remote learning can bring to preschoolers. The articles describes a study by the MacArthur Foundation that has not yet been peer-reviewed. The research describes the results of bringing preschool to Syrian refugees.

Syrians are the largest displaced population in the world. Seventy percent of Syrians fled the country over the last decade as violence in the country grew out of control. Most of the displaced population is living in nearby countries like Lebanon. The life of the refugees remains unsettled because they have not been accepted in the new countries and have no path to citizenship. For example, Lebanon is itself in strife. The country currently has no president. Events like the Israel-Palestine crisis have brought new violence to the country.

The vast majority of children on the planet have had disruptions in education due to the pandemic, climate disasters, and war. According to UNICEF, there have been 43.3 million children driven from their homes worldwide since 2020.

The study concerns an effort coordinated by the International Rescue Committee, the Sesame Workshop, and educators to create a preschool program that can be delivered to students in tough situations. One of the programs created in the package of materials is Ahlan Simsim (Welcome Sesame) that is created in Arabic using many of the familiar Sesame Street characters. The educational programming is being broadcast into refugee centers using cellphones and is being made available to many other children in more traditional ways to deliver video content. Over 2 million preschoolers have watched the content and participated in the program.

Lebanon is a challenging environment for anything digital because only 78% of residents have any access to broadband, and only about two-thirds of adults have a cellphone. The Syrian refugees typically have one cellphone per family.

There was already a movement underway in Lebanon before the pandemic to create this content. Sesame street had created programming using Ma’zooza the goat, Jad a yellow monster, and Basma a purple friend. But the pandemic put the project on the shelf.

The study that been conducted with the release of the content seems promising. It appears that remote learning is nearly as effective as live preschool. Children viewing the programming have seen noted improvements in literacy, number skills, motor skills, and social-emotional skills. Best of all, kids viewing the programming have incorporated what they have learned into their play.

One thing that probably enhanced the experience for kids is that Syrian mothers have embraced the programming and watch it with their children. That provides strong reinforcement for tasks like learning the alphabet or how to count.

The effort has been so successful that it’s now being copied and modified for other refugee populations around the world. Educators understand the risk to children who grow up in stressful environments without learning the basic skills that lead to life-long learning. Children who don’t get this kind of mental stimulation in the early years lose the ability to ever catch up with their peers.

Broadband Subscribers 4Q 2023

Leichtman Research Group recently released broadband customer statistics for the end of 2023 for the largest cable and telephone companies. Leichtman compiles most of these numbers from the statistics provided to stockholders other than for Cox and Mediacom, which are estimated and now reported together. Leichtman says this group of companies represents 96% of all US landline broadband customers.

Broadband growth for the fourth quarter is still coming almost entirely from the growth of FWA cellular broadband provided by T-Mobile and Verizon. This is particularly true since AT&T’s smaller FWA growth is buried in its overall numbers. For the quarter, FWA fixed cellular customers grew by 929,000 thousand while telco grew by 28,000 subscribers, and cable companies collectively lost 144,000. The fourth quarter of 2023 might be remembered as the quarter when cable companies started to collectively lose customers while telcos collectively gained customers.

Annual
4Q 2023 4Q Change Change
Comcast 32,253,000 (34,000) (66,000)
Charter 30,588,000 (61,000) 155,000
AT&T 15,288,000 (8,000) (98,000)
Verizon 7,650,000 38,000 166,000
Cox & Mediacom 7,020,000 (15,000) (8,000)
T-Mobile FWA 4,776,000 541,000 2,130,000
Altice 4,517,900 (27,500) (114,100)
Verizon FWA 3,067,000 388,000 1,536,000
Frontier 2,943,000 62,000 75,000
Lumen 2,758,000 (78,000) (279,000)
Windstream 1,175,000 0 0
Cable ONE 1,059,300 1,900 (1,100)
Breezeline 663,286 (8,476) (29,184)
TDS 539,800 7,200 29,800
Consolidated 393,219 6,998 25,761
Total 114,691,505 813,122 3,522,177
Cable 76,101,486 (144,076) (63,384)
Telco 30,747,019 28,198 (80,439)
FWA 7,843,000 929,000 3,666,000

In the telco sector, Lumen continued to shrink and lost 2.8% of its broadband customers in the quarter. AT&T had a tiny loss, and all other telcos saw growth. Surprisingly, the biggest telco gainer in the fourth quarter was Frontier, which is finally seeing its fiber strategy working.

The only cable company with a tiny growth is Cable One, and every other large cable company lost broadband customers in the fourth quarter.

For the year, the big broadband companies grew by 3.5 million customers, and all of the annual growth came from FWA cellular. As expected, T-Mobile FWA passed Altice in size, and Verizon FWA surpassed Frontier in the fourth quarter. Lumen is the biggest loser, having lost 9.2% of its broadband customers in 2023.

BEAD Pressure on Broadband Rates

State Broadband Offices and the BEAD grant process have designed grant rules that put pressure on ISPs to provide inexpensive rural broadband. But in doing so, I’m not sure that they understand the high prices that rural folks are paying for broadband today.

To provide an example, my consulting firm just finished a statistically valid broadband survey in a rural county. Most of the rural folks in this county are already spending more than $100 per month on broadband. The incumbent telephone company has largely walked away from the county, and there is no DSL available in most of the rural parts of the county. There is only one cell tower that has been upgraded to offer FWA broadband. This means that most rural residents only have access to expensive broadband.

There is one WISP that covers only a small portion of the county. The WISP’s rates are high, with a 10/1 Mbps product for $76 and faster speeds cost more than $100.

Cellular hotspots are expensive. For example, T-Mobile has hotspot plans that vary in price from 5 gigabytes of data for $20 to 50 gigabytes of data for $50. AT&T and Verizon have similar prices for hotspots. Thos rates sound low until you realize the cost of buying broadband over the data caps. For example, Verizon offers 150 gigabytes of data for $110.

The data caps on hotspots put horrendous pressure on a household. OpenVault recently said that the average U.S. home uses 641 gigabytes of data per month. It’s hard to imagine trying to restrict a home to using less than 100 gigabytes per month to keep the bill under $100 per month. During the pandemic, I talked to several parents of students working from home using a hotspot who were paying more than $500 per month through extra bandwidth fees to get the same kind of bandwidth that homes with unlimited bandwidth take for granted.

Satellite broadband is mostly over $100. Starlink prices start at $110. HughesNet advertises unlimited data, but their plans have data caps for all practical purposes. For example, a plan with 200 gigabytes of priority data costs $109.99. They label the plan as unlimited usage, but data usage above 200 gigabytes is choked to extremely slow speeds. All plans require a 2-year contract.

Viasat is even more expensive. For $119.99 per month a customers gets 100-150 gigabytes of high-speed data and then unlimited choked and slow broadband. 500 gigabytes of usage is $249.99.

In this county, most rural residents are already paying over $100 per month for broadband. Since many of the plans include data caps, 4% of residents report spending more than $150 per month.

There are state BEAD rules that are trying to force rates down to rates between $50 and $75 per month for gigabit speeds. I find several faults with these rate-setting efforts:

  • The rates the Broadband Offices are seeking are typically far cheaper than what the large cable companies charge – the companies that have the most customers in almost every state. It’s hard to think of a reason why grant offices want rural rates to be lower than urban rates, particularly when operating costs are higher in the rural areas.
  • The low rates don’t acknowledge that rural residents are already paying a lot more for inferior broadband. Do we really need to cut what folks are paying in half if they are going to upgrade from crappy broadband?
  • The high rates are counterproductive. I talked to an electric cooperative this week that has zero interest in BEAD grants due entirely to the insistence on low rates. They understand the rates necessary to win BEAD would create a losing business plan.
  • Probably most important is that the IIJA legislation strictly forbid NTIA and the States from using the grants for setting rates – and yet States are openly doing it anyway. Why isn’t the NTIA rejecting these rate plans?

This is one more issue that I find ironic about the BEAD process. The Department of Commerce and NTIA have repeatedly said that BEAD grant rules are conservative to protect taxpayers. However, putting tremendous pressure on ISPs to have low rates does the exact opposite and will put ISPs at risk for failure after taking the grant funding. There seems to be an underlying assumption that ISPs have unlimited and hidden profit reserves and can absorb whatever penalty grant folks want to invent. That may be somewhat true for a few giant ISPs, but it sure isn’t true about everybody else.

A New Definition of Broadband

The FCC finally increased the definition of broadband from 25/3 Mbps to 100/20 Mbps. The change was too long in coming. This should have been done when Ajit Pai headed the FCC, but politics got in the way. It should have happened when Congress set the definition to 100/20 Mbps in the BEAD grant rules over two years ago – but again, politics interfered.

Coming four or five years too late, the 100/20 Mbps definition of broadband is not only not forward-looking, the new definition of broadband is out of sync with the market. Consider the following table that comes from OpenVault that shows how broadband subscriptions have changed in the country since before the pandemic.

Dec 2019 Dec 2024
Under 50 Mbps 22%  6%
50 – 99 Mbps 24%  4%
100 – 199 Mbps 37% 16%
200 – 499 Mbps 11% 34%
500 – 999 Mbps  4%  7%
1 Gbps  3% 33%

Just before the pandemic, 46% of households were subscribing to speeds under 100 Mbps. Today, only 10% of households buy slower speeds – and many of them are likely in rural areas where they have no other option. The numbers have flipped since the pandemic and 40% of households are now subscribing to speeds of 500 Mbps or faster.

Every time I write a blog about broadband speeds, a few ISPs will respond, saying that people don’t need faster broadband. The first time I heard that sentiment was a unified response from the CEO of every big cable company, who said the same thing when Google upset the market by introducing gigabit broadband. The reality is that it doesn’t matter what people need – what matters is what they are willing to buy. The table above shows that people want to buy faster broadband when given the option. I look at that table, and it’s hard to conclude anything other than the public broadband is something faster than 200 Mbps.

What’s missing in the above chart is any recognition of the importance of upload broadband speeds. I think many consumers who upgrade to faster speeds do so because of issues with upload speeds. Serious gamers and folks who work from home with large data files will tell you that the new 20 Mbps definition of upload broadband is massively obsolete.

Unfortunately, the definition of broadband has political and financial overtones. It determines where grant money can be spent. Upping the definition of broadband also has market consequences. Any ISP that is offering speeds less than 100/20 Mbps today is no longer selling broadband. They might as well be selling dial-up, because even the slow-to-change FCC says that what they are selling is obsolete and is something other than broadband.

It’s hard to say if changing the definition of broadband has any practical impact. It’s possible that this kind of announcement will filter down through the public and cause folks buying slower connections to search for something faster.

The natural question is, what’s the next step in defining broadband? My bet is that the FCC is going to rest on its laurels for a while after finally getting through the politics and making this change. It took nine years to move the definition from 25/3 to 100/20 Mbps. Hopefully, the FCC won’t wait another nine years. Congress already said that 100/20 Mbps is obsolete when it said that federal broadband grants ought to be spent to build gigabit networks. The OpenVault table above says that the public’s desire for gigabit broadband is already here today.