Tag Archives: DSL

AT&T in the News

AT&T has not been in the headlines a lot this year, but recently I’ve seen the company’s name everywhere.

In the recently released financial results for the third quarter, AT&T noted that it now has more fiber broadband customers than non-fiber customers. At the end of the quarter, AT&T had 6.93 million fiber customers compared to 6.86 million remaining non-fiber customers. Non-fiber customers are predominantly U-Verse customers served by two pairs of telephone copper. The company still also has 340,000 DSL customers served by a single copper pair. There are also some rural fixed-wireless customers.

In the third quarter, AT&T added 338,000 fiber customers. The company lost 367,000 non-fiber customers in the second quarter – although counting them as lost is probably a misnomer since many were likely upgraded to fiber.

Upgrading to fiber is good for the company’s bottom line. For the quarter, the average revenue per user (ARPU) was $62.62 for fiber customers compared to only $54.60 for non-fiber customers. AT&T has also been saying for years that the cost of maintenance for copper is a lot higher, so the company is likely shedding costs as it sheds customers served on copper.

We also got a peek at market AT&T’s penetration. AT&T says it passes 18.5 million potential customers with fiber, meaning the company has achieved an overall 37% market penetration on fiber. In the third quarter, the company added fiber to pass 500,000 new locations.

I saw another interesting news blurb about AT&T. Bloomberg reported that AT&T is looking for an equity partner to invest in a major expansion of fiber. That would be a big departure from the past since AT&T has always funded its own capital expenditures and networks.

But it’s not hard to see from the third quarter results why AT&T might be seeking additional funding. In the third quarter, the company generated $9.87 billion of cash. It invested $4.71 billion in new infrastructure and paid $3.75 billion in dividends – leaving only $1.41 billion in free cash.

I would conjecture that AT&T wants to invest more heavily in fiber immediately since it’s clear that there is a mad rush nationwide to build fiber in cities. Fiber overbuilders hope that if they are the first to a market with fiber that it might dissuade other fiber overbuilders – so we are currently seeing a fiber land grab. In the long run, sharing fiber profits with an investor will decrease future AT&T earnings. The calculus that the company is betting on is that the market share gained by building first to markets outweighs the cost of sharing profits.

AT&T is currently debt-heavy. AT&T hasn’t had a recent track record of making good investment decisions. It’s been reported that AT&T lost as much as $50 billion from its purchase of DirecTV. In almost the same time frame, the company lost as much as $42 billion from its purchase and sale of WarnerMedia. The company might not be able to easily borrow the money, particularly at current interest rates.

The final news is that AT&T was fined $23 million to resolve a federal investigation that the company had “unlawfully influenced” the former Illinois Speaker of the House, Michael J. Madigan. AT&T admits that it paid Madigan, through an ally, to promote legislation that would eliminate carrier of last resort in the state – meaning that the company is obligated to serve people who ask for a telephone line. That obligation also comes with legacy regulatory requirements that AT&T wanted to ditch.

What always dismays me, but never surprises me, is that nobody at a big company like AT&T got in trouble for breaking the law – in this case, bribing a government official. The size of the fine might be appropriate for the magnitude of the crime, but I’ve always thought that the folk at big companies would be more likely to hesitate to be unethical if they saw others going to jail for breaking the law. The only real consequence for AT&T, in this case, is that they got caught, and the fine will just be viewed as the cost of doing business.

The Birth of the Digital Divide

A lot of the money being spent on broadband infrastructure today is trying to solve the digital divide, which I define as a technology gap where good broadband is available in some places, but not everywhere. The technology divide can be as large as an entire county that doesn’t have broadband or as small as a pocket of homes or apartment buildings in cities that got bypassed.

I can clearly remember when the digital divide came about, and at that time I remember discussing how the obvious differences between technologies were going to someday become a major problem. Today I’m going to revisit the birth of the digital divide.

Until late in the 1990s, the only way for almost most people to get onto the Internet was by the use of dial-up access through phone lines. ISPs like AOL, CompuServe, and MSN flourished and drew millions of people online. At first, dial-up technology was only available to people who lived in places where an ISP had established local dial-up telephone numbers. But the online phenomenon was so popular, that ISPs eventually offered 800 numbers that could be reached from anywhere. There was no residential digital divide, except perhaps in places where telephone quality wasn’t good enough to accommodate dial-up. Some businesses used a faster technology to connect to the Internet using a T1, which had a blazingly fast speed of 1.6 Mbps, almost 30 times faster than dial-up. To people connecting at 56 kbps, a T1 sounded like nirvana.

The digital divide came into being when the faster technologies of DSL and cable modem were offered to homes. My first DSL line had a download speed of almost 1 Mbps, an amazing 18 times increase in speed over the dial-up modem. At almost the same time, some cable companies began offering cable broadband that also had a speed of around 1 Mbps. Homes in urban areas had a choice of two nearly-identical broadband products, and the early competition between telephone and cable companies was loud and fierce.

The advent of DSL created the first digital divide – the gulf between urban areas and rural areas. While telcos theoretically offered DSL in much of rural America, the 2-mile limitation of the DSL signal meant the speed didn’t carry far outside of the towns that housed the DSL transmitters, called DSLAMs. Many telcos were willing to sell rural DSL, even if speeds were often barely faster than dial-up. Soon after the first DSL was offered to customers, the vendors came up with ISDN-DSL that could deliver a speed up to 128 kbps deeper into rural copper networks – twice the speed of dial-up. But decent DSL never made it very far into most of rural America – and still doesn’t today for much of rural America.

The DSL and cable modem technologies improved within a few years after introduction, and the technology improvements created the second digital divide. I recall versions of DSL that had a maximum speed of 3, 6, 12, 15, 24, and eventually 48 Mbps. The big telcos upgraded to later DSL technology in some neighborhoods, but not others. Sadly, even today we continue to find places where the earliest versions of DSL are still offered, meaning there are places where DSL speeds never climbed above 3, 6, or 12 Mbps. This was particularly painful in towns that didn’t have a cable competitor because they were stuck with whatever flavor of DSL the telephone company offered to them. This was noticeable in big cities where some neighborhoods never saw any DSL upgrades. There was a well-known study done a number of years ago documenting the DSL technologies available in Dallas, Texas. The study showed that poor neighborhoods still had the slowest versions of DSL while more affluent neighborhoods had DSL speeds up to 50 Mbps.

Cable modem technology improved more quickly than DSL. By 2005, the cable modem won the speed game. And that’s when the cable companies started charging more for cable broadband – something they could do because the broadband was faster. This price difference largely meant that low-income households were stuck with DSL, while folks who care about speeds migrated over the years to the cable companies.

The digital divide in rural areas deepened as older DSL was not upgraded while the DSL that had originally been deployed started to reach end-of-life. Copper networks have lasted far past the expected economic useful life and get a little worse every year. In cities, any parts of the city stuck with only DSL fell far behind the neighborhoods where speeds increased significantly from both DSL and cable modems.

Unfortunately, we are not at the end of this story. There is a huge amount of fiber being constructed today in urban areas. But there is no reason to think that most of the ISPs building fiber are going to serve every neighborhood. The big telcos that build fiber like Verizon, AT&T, Frontier, CenturyLink, and others have always cherry-picked what they think are the best neighborhoods – best in terms of either demographics or the lowest in cost of deployment.

Unless we reach a time when fiber is everywhere, the digital divide will stick around. Right now, we’re tackling the rural digital divide – I expect in 5 or 10 years we’ll have to do this all over again to tackle the urban digital divide.

AT&T to Chop Copper Networks

In a pronouncement that is news to nobody, AT&T announced at a recent investor day event that it has plans to cut its copper network footprint in half by 2025. This can’t be a surprise from a company that stopped connecting new DSL customers in October 2020. I figured we could start the countdown clock on copper from that date.

However, Jeff McElfish, the CEO of AST&T’s Communications division, said something that is surprising. He said the company isn’t planning to forcibly move customers off copper as they decommission copper. He says customers are naturally migrating off copper. I find that hard to believe.

My consulting firm administers surveys, and we are still seeing DSL penetration rates in cities between 10% and 40%. Our surveys indicate that the people who are staying with DSL are doing so because of price – they largely hate DSL performance, but it’s what they can afford. This is not hard to understand when looking at the rates for broadband from the big cable companies.

In this blog, I’ve often talked about how expensive broadband is from Comcast and Charter, but broadband rates from some of the other cable companies like Cox and Atlantic Broadband are even higher. There are a lot of homes that can’t afford the cable company prices. It’s hard for me to believe that all of these people are going to voluntarily walk away from DSL over the next two or three years. The last estimate I vaguely remember reading was that there is still something like 19 million households still using DSL.

McElfish said AT&T plans to have 75% of its footprint covered by fiber or fixed cellular wireless by 2025 – I have to assume that in terms of square miles of footprint that this will mostly be wireless. AT&T is going to have a PR problem with trying to push customers to wireless. For rural customers within reach of a tower, a switch from DSL to fixed cellular wireless will be a no-brainer. The broadband speeds will be faster, and the price still affordable. But the big problem in rural markets is that there are huge parts of rural America where fixed wireless won’t work. The rural cellular coverage maps for all three big cellular companies are a joke, and anybody who drives into rural areas can see that you don’t usually have to go far to run out of bars of service. It’s worth noting that cellular voice covers a much larger footprint than cellular data. At some point, AT&T will have to drop rural DSL customers who might have no other alternative than satellite broadband. Extrapolating from McElfish’s statement of covering 75% of the footprint means that AT&T will be abandoning folks in 25% of its footprint.

Urban areas are a bigger issue for AT&T because that’s where most of the DSL customers remain. It’s clear that AT&T has no goal of overbuilding whole cities with fiber but is building in selected neighborhoods. It’s not clear if those neighborhoods are chosen due to the most affordable construction costs or the best demographics – but AT&T will not be building fiber to cover the majority of its footprint in most cities.

With today’s 4G LTE technology that’s been branded as 5G, AT&T is not prepared to deliver fixed cellular broadband to huge numbers of people in cities. That’s what 5G is supposed to fix, and it’s not here yet. But even when AT&T finally implements real 5G (estimated to be 5 – 7 years in the future), the company would have to install a huge number of small cell sites to have enough broadband capacity to migrate DSL customers to fixed cellular broadband. And that means building more fiber deep into neighborhoods to serve the small cell sites. None of that is happening by 2025, so AT&T must be planning on turning down rural copper markets first.

Perhaps AT&T is really counting on everybody else to pick up its DSL customers. T-Mobile is already aggressively rolling out fixed cellular broadband, and Verizon plans a big push starting in late summer of this year. Dish plans to open 25 major markets with cellular data by June. Smaller wireless player like Starry might be making a dent by 2025.

AT&T is ultimately going to have to force people off DSL. The download speeds on much urban DSL are not dreadful, at 15 – 30 Mbps, although upload speeds are nonexistent. I don’t see millions of people voluntarily abandoning the product so that AT&T can tear down the copper without a public stir.

But maybe there is another motive behind this – as the technicians who understand DSL keep retiring, AT&T might not be able to keep DSL running by 2025. I know that sounds cynical, but I don’t think it’s far from the truth.

A Brief History of Rural Broadband

Last week I lectured on the topic of digital redlining in Western North Carolina for a senior class at the University of North Carolina at Asheville. These students are tackling a senior project of developing an advocacy program to help find broadband solutions for our region. I realized when describing redlining that I was also describing how the history of rural telephony had contributed to the poor condition of rural broadband today. Following is a brief history of the key events that tell the story of why rural broadband in much of the country has fallen so far behind broadband in larger towns and cities.

Bell Builds to Cities. Within a few short years after the invention of the telephone in 1876, the well-funded American Bell telephone company, founded by Alexander Graham Bell and investors, had built telephone networks in all of the major cities in the country. Within a decade, they made it out to county seats like Asheville, here in Western North Carolina. American Bell had no interest in building to rural areas, and rural America was not offered the new telephone technology.

Expiration of the Bell Patent. In 1894 the Bell patent on the telephone expired, and telephone companies sprang up across the country. Most of these new companies were started in rural areas by farmers, businessmen, or groups of citizens that came together to bring telephone service. By 1927 there were over 6,000 local telephone companies.

Bell Becomes a Monopoly. American Bell became a monopoly in an extraordinary move where Theodore Vail asked the government to grant monopoly status to the company. This was done to fight off competitive telephone companies in cities. Regulating telephone companies was gradual as states accepted the idea and awarded franchise areas to telephone companies. A consequence of regulation was that small telephone companies also became regulated. The regulation trend culminated in the Telecommunications Act of 1934 that created the FCC.

Birth of Cooperatives. The Rural Electrification Administration, a New Deal agency, had been funding rural electric cooperatives since 1935. In 1949 the agency started making loans to create rural telephone cooperatives, and these companies filled in the remaining rural areas nobody else had built. By 1960 the U.S.was the envy of the world with a 99% landline telephone penetration.

The Growth of Long-Distance Calling. AT&T was created in 1885 to construct the long-line networks to connect cities. The first coast-to-coast call was made in 1915 between New York and San Francisco. Regulation led to low local telephone rates, and telephone companies made up for lost revenues with expensive long-distance rates. The downside of long-distance for rural America was that calls between rural areas and county seats all became long-distance. Coping with long-distance calling became a major concern for many rural families, and rural residents were at a disadvantage compared to city dwellers since they needed to use long-distance to communicate with basic services.

Massive Consolidation. Companies like General Telephone, Continental Telephone, and Citizens Telephone purchased many rural telcos. These companies were the precursors of big rural telcos like Frontier, CenturyTel, and Windstream.

Divestiture of the Bell Companies. In 1984, divestiture separated AT&T, the long-distance company, from the local Bell telcos, with the goal of introducing competition into long-distance. This worked spectacularly, and long-distance prices tumbled. The unfortunate consequence of lower long-distance rates was that the larger telcos saw lower profits in rural areas. Big telcos began to cut staff, close businesses offices, and generally neglect rural properties.

Local Competition and Deregulation. The Telecommunications Act of 1996 mandated local telephone competition. Over time, this led to the big telcos seeking and winning local deregulation, and as the big telcos were deregulated, the neglect of big telco rural properties accelerated.

The Rise of DSL Broadband. The first generation DSL that delivered 1 Mbps download got installed in most telcos. Rural speeds were not as fast as cities due to distance limitations on DSL. The technology improved rapidly in a few short years, and the DSL in towns was upgraded to faster speeds while rural properties mostly were not. When cities got DSL speeds up to 50 Mbps, rural area DSL stayed slow since the big telcos didn’t want to make rural investments.

Small Companies Did Much Better. The remaining smaller telcos and cooperatives did a great job during all of these industry transitions. They maintained copper wiring, upgraded DSL as needed, and eventually started upgrading in many places to rural fiber.

The Large Cellular Companies Shunned Rural America. Big cellular companies built close to major highways to capture roaming traffic, but rural cellular coverage rarely extended to where people live and work. It’s gotten better over a few decades, but rural cellular coverage maps are still largely fictional.

Summary. The poor state of rural broadband can be traced to the ways that the big telcos reacted to industry changes. Small telcos built rural networks, but large telcos gobbled them up over time. The big rural telcos then neglected rural properties in reaction to the changing economics from the deregulation of long-distance and local telephone service. Small telcos showed that it wasn’t necessary to abandon rural properties, but the big telcos stopped making investments in rural networks and for all practical purposes walked away from rural communities.

No Home Broadband Option

We spend a lot of time arguing policy questions, such as asking if 25/3 Mbps is adequate broadband. What policymakers should really be talking about are the huge numbers of homes with dreadful broadband. The worst thing about the deceptive FCC maps is that they often give the perception that most rural areas have at least some broadband options when many rural residents will tell you they have no real broadband options.

Policymakers don’t grasp the lousy choices in many rural areas. The FCC maps might show the availability of DSL, but if it’s even available (often it’s not), the speeds can be incredibly slow. Rural households refuse to pay for DSL that might deliver only 1 or 2 Mbps download and practically no upload.

I think the FCC assumes that everybody has access to satellite broadband. But I’ve talked to countless rural residents who tried satellite broadband and rejected it. Real speeds are often much slower than advertised speeds since trees and hills can quash a satellite signal. The latency can be crippling, and in places where the speeds are impaired, the high latency means a household will struggle with simple real-time tasks like keeping a connection to a shopping site. Satellite plans also come with tiny data caps. I’d like to put a few Washington DC policymakers on a monthly data plan with a 40 GB or 60 GB cap so they can understand how quickly that is used in a month. But the real killer with satellite broadband is the cost. HughesNet told investors last year that its average revenue per customer was over $93 per month. Many rural homes refuse to pay that much for a broadband product that doesn’t work.

We hear a lot of stories about how fixed wireless technology is getting better to the point where we’re hearing preposterous conversations about bringing gigabit fixed wireless to rural areas. There are still a lot of places with woods and hills where fixed wireless is a poor technology choice. I worked with one county recently that gathered thousands of speed tests for fixed wireless that showed average download speeds under 5 Mbps and upload speeds below 1 Mbps. There are still a lot of WISPs that are cramming too many customers on towers, chaining too many towers together with wireless backhaul, and selling to customers who are too far from towers. This is not to say that there aren’t great WISPs, but in too many rural places the fixed wireless choices are bleak.

Rural residents have also suffered with cellular hotspots. These are the plans that cellular companies have had for years that basically price home broadband at the same prices and data caps as cellular broadband. During the pandemic, I’ve heard from families who were spending $500 to $1,000 per month in order to enable home-schooling during the pandemic. This product is not available in huge parts of rural America because of the poor or nonexistent cellular coverage. We complain about the FCC’s broadband maps, but those are heads and tails better than the cellular company coverage maps which massively overstate rural cellular availability.

There is some relief in sight for some rural homes. I recently talked to farmers who are thrilled with the T-Mobile fixed cellular product – but they said distance from cell sites is key and that many of their neighbors are out of range of the few cell sites found in most rural counties. There are rural folks who are happy with Starlink. But there are a lot of people now into the second year on the waiting list to get Starlink. Starlink also has reported problems with trees and hills and also comes with a steep $99 per month price tag.

When a rural household says they have no broadband connection, I’ve learned that you have to believe them. They will have already tried the DSL, fixed wireless, satellite, and cellular hotpots, and decided that none of the options work well enough to justify paying for them. The shame is that the FCC maps might give the impression that residents have two, three, or four broadband options when they really have none.

The New Speed Battle

I’ve been thinking about the implications of having a new definition of broadband at 100/20 Mbps. That’s the threshold that has been set in several giant federal grants that allow grant funding to areas that have broadband slower than 100/20 Mbps. This is also the number that has been bandied about the industry as the likely new definition of broadband when the FCC seats a fifth Commissioner.

The best thing about a higher definition of broadband is that it finally puts the DSL controversy to bed. A definition of broadband of 100/20 Mbps clearly says that DSL is no longer considered to be broadband. A 100/20  Mbps definition of broadband means we can completely ignore whatever nonsense the big telcos report to the FCC mapping process.

Unfortunately, by killing the DSL controversy we start a whole new set of speed battles with cable companies and WISPs that will be similar to the controversy we’ve had for years with DSL. Telcos have claimed 25/3 Mbps broadband coverage over huge parts of rural America in an attempt to deflect broadband grants. In reality, there is almost no such thing as a rural customer who can get 25/3 Mbps DSL unless they sit next to a DSLAM. But the telcos have been taking advantage of the theoretical capacity of DSL, and the lax rules in the FCC mapping process allowed them to claim broadband speeds that don’t exist. I hate to admit it, but overstating DSL speeds has been a spectacularly successful strategy for the big telcos.

We’re going to see the same thing all over again, but the new players will be cable companies and WISPs. The controversy this time will be more interesting because both technologies theoretically can deliver speeds greater than 100/20 Mbps. But like with DSL, the market reality is that there are a whole lot of places where cable companies and WISPs are not delivering 100/20 Mbps speeds and would not be considered as broadband with a 100/20 Mbps yardstick. You can take it to the bank that cable companies and WISPs will claim 100/20 Mbps capability if it helps to block other competitors or if it helps them win grants.

The issue for cable companies is the upload speed. One only has to look at the mountains of speed tests gathered around the country to see that cable upload speeds are rarely even close to 20 Mbps. We’ve helped cities collect speed tests where maybe 5% of customers are reporting speeds over 20 Mbps, while the vast majority of cable upload speeds are measured at between 10 Mbps and 15 Mbps. Usually, the only cable customers with upload speeds over 20 Mbps are ones who have ponied up to buy an expensive 400 Mbps or faster download product – and even many of them don’t see upload speeds over 20 Mbps.

This begs the question of what a definition of broadband means. If 95% of the customers in a market can’t achieve the defined upload speeds, is a cable company delivering broadband under a 100/20 Mbps definition? We know how the telcos answered this question in the past with DSL, and it’s not hard to guess how the cable companies are going to answer it.

It’s not a coincidence that this new controversy has materialized. The first draft of several of the big grant programs included a definition of broadband of 100/100 Mbps – a speed that would have shut the door on cable companies. But cable company lobbying began immediately, and the final rules from Congress included the slimmed-down 100/20 Mbps broadband definition.

WISPs have a more interesting challenge because the vast majority of existing WISP connections are nowhere close to meeting either the upload or download speed of 100/20 Mbps. But fixed wireless technology is capable of meeting those speeds. A WISP deploying a new state-of-the-art system can achieve those speeds today for some reasonable number of miles from a tower in an area with good lines of sight. But most existing WISPs are deploying older technology that can’t come close to a 100/20 Mbps test. Even WISPs with new technology will often serve customers who are too far from a tower to get the full speeds. Just like with cable companies, the 100/20 Mbps definition of broadband will allow WISPs to stay in the game to pursue grants even when customers are not receiving the 100/20 Mbps speeds. So brace yourself, because the fights over speeds are far from over.

Broadband Interference

Jon Brodkin of ArsTechnica published an amusing story about how the DSL went out of service in a 400-resident village in Wales each morning at 7:00 am. It turns out that one of the residents turned on an ancient television that interfered with the DSL signal to the extent that the network collapsed. The ISP finally figured this out by looking around the village in the morning with a spectrum analyzer until they found the source of the interference.

It’s easy to think that the story points out another weakness of old DSL technology, but interference can be a problem for a lot of other technologies.

This same problem is common on cable company hybrid-fiber coaxial networks. The easiest way to understand this is to think back to the old days when we all watched analog TV. Anybody who watched programming on channels 2 through 5 remembers times when the channels got fuzzy or even became unwatchable. It turns out that there are a lot of different devices that interfere with the frequencies used for these channels including things like microwave ovens, certain motors like power tools and lawnmowers, and other devices like blenders. It was a common household occurrence for one of these channels to go fuzzy when somebody in the house, or even in a neighboring home used one of these devices.

This same interference carries forward into cable TV networks. Cable companies originally used the same frequencies for TV channels inside the coaxial wires that were used over the air and the low TV channels sat between the 5 MHz and 42 MHz frequency. It turns out that long stretches of coaxial wires on poles act as a great antenna, so cable systems pick up the same kinds of interference that happens in homes. It was pretty routine for channels 2 and 3, in particular, to be fuzzy in an analog cable network.

You’d think that this interference might have gone away when cable companies converted TV signals to digital. The TV transmissions for channels 2 through 5 got crystal clear because cable companies relocated the digital version of these channels to better frequency. When broadband was added to cable systems the cable companies continue to use the low frequencies. CableLabs elected to use these frequencies for the upload portion of broadband. There is still plenty of interference in cable networks today – probably even more than years ago as coaxial networks have aged and have more points for interference to seep into the wires. Until the pandemic, we didn’t care much about upload bandwidth, but it turns out that one of the major reasons that cable companies struggle to deliver reliable upload speeds is that they are using the noisiest spectrum for the upload function.

The DSL in the village suffered from the same issue since the telephone copper wires also act as a big outdoor antenna. In this village, the frequency emanating from the old TV exactly matched the frequencies used for DSL.

Another common kind of interference is seen in fixed wireless networks in a situation where there are multiple ISPs using the same frequencies in a given rural footprint. I know of counties where there are as many as five or six different wireless ISPs, and most use the same frequencies since most WISPs rely on a handful of channels in the traditional WiFi bandwidth at 2.4 MHz and 5 MHz. I’ve heard of situations where WiFi is so crowded that the performance of all WISPs suffer.

WiFi also suffers from local interference in the home. The WiFi standard says that all devices have an equal chance of using the frequencies. This means that a home WiFi router will cycle through all the signals from all devices trying to make a WiFi connection. When a WiFi router connects with an authorized device inside the home it allows for a burst of data, but then the router disconnects that signal and tries the next signal – cycling through all of the possible sources of WiFi.

This is the same issue that is seen by people using WiFi in a high-rise apartment building or a hotel where many users are trying to connect to WiFi at the same time. Luckily this problem ought to improve. The FCC has authorized the use of 6 GHz spectrum for home broadband which opens up numerous new channels. Interference will only occur between devices trying to share a channel, but that will be far fewer cases of interference than today.

The technology that has no such interference is fiber. Nothing interferes with the light signal between a fiber hub and a customer. However, once customers connect the broadband signal to their home WiFi network, the same interference issues arise. I looked recently and can see over twenty other home WiFi networks from my office – a setup ripe for interference. Before making too much fun of the folks in the Welsh village, there is a good chance that you are subject to significant interference in your home broadband today.

AT&T Stops DSL Sales

USA Today reported last week that AT&T stopped selling new DSL to customers on October 1. This is an event that will transform the broadband landscape in a negative way across the country. There are a number of immediate consequences of this action by the company.

Probably the most dramatic impact will be that many rural customers will no longer have an option for landline broadband. While rural DSL broadband is slow, a DSL connection at speeds between 1 Mbps and 6 Mbps beats the alternatives – which is satellite broadband or cellular hotspots. Since there are a lot of rural homes where those two technologies don’t work, this means some homes will suddenly have no broadband option. Expect to soon see stories of folks who buy rural homes and then find they have no option to buy broadband.

In cities where AT&T DSL is the only alternative to a cable company broadband service, this move bestows total monopoly power to the cable company. Our firm does broadband surveys and we still find markets where AT&T DSL represents as much as a 30% market share. Many homes buy DSL because it costs less, and that option just got taken off the table in AT&T markets. And just like in rural markets, every city has customers who’s only choice is DSL. For various reasons, there are streets in most cities where the cable companies never constructed network. Any customer moving into one of these broadband deserts will find themselves with no broadband alternative.

According to an article just published by Ars Technica, only 28% of AT&T broadband customers have access to AT&T fiber – anybody living in the neighborhoods without fiber will no longer be able to buy broadband from AT&T. That has to equate to tens of millions of households that just lost a broadband option. The FCC proudly measure the number of homes with multiple broadband options, and I’ll be curious to see if they recognize this sea change in the market.

This change will stop the practice of customers who hop back and forth between DSL and cable company broadband to save money. I just talked to a customer the other day that has bounced between DSL and cable company broadband for almost twenty years. Both the cable company and the telco offer introductory prices each time for swapping, and this customer has gone back and forth between the ISPs regularly every few years. In neighborhoods where AT&T is the telco DSL provider, this might mean the end of introductory special prices from the cable company – they now have zero incentive to compete for customers.

I would have to think that Verizon will eye this announcement closely. They have openly said that they want to do away from copper network technology. This might be all of the push needed for Verizon to follow suit. This announcement might be citied in telco history as the beggining of the end of copper wires. AT&T says they won’t be tossing folks off DSL service, but will no longer connect new customers to the DSL technology. Over time this is going to mean fewer and fewer customers on copper, and I suspect AT&T already has a date in mind when they walk away from the technology completely.

Ironically, AT&T just recently announced that they were going to claim a seventh year of CAF II support in 2021 and will collect over $427 million in subsidies next year to supposedly support rural DSL. Hopefully, the FCC will view this announcement as grounds for stopping such payments. It would be absolutely insane to give millions to AT&T to support a technology that the company will no longer sell or install.

This timing of the announcement is also curious at a time when the pandemic is still raging. This means a home that needs to buy broadband to support students or adults working from home will no longer have that option if the only wired connection is AT&T DSL.

This announcement also creates an interesting dilemma for the FCC. Will the FCC pretend that the huge AT&T DSL footprint still exists? It’s impossible to pretend that areas have a broadband option when the only provider of landline service refused to connect new customers. I’m sure the FCC will act as if this announcement never happened – because recognizing it means now counting millions of homes as having no broadband option.

This day has been inevitably coming for decades. Regulators have long pretended that they could demand that the big telcos keep supporting an obsolete technology. AT&T and Verizon have been telling regulators for years that they are going to walk away from copper, and now one of the big telcos is doing so. It’s just a matter of time until AT&T begins decommissioning DSLAMs and starts tearing down copper wires for the salvage value – and I can’t see any way that regulators can stop them.

Who Has the Fastest Broadband?

Ookla recently released a report for the second quarter that summarizes its findings on speed tests conducted throughout the US. The report was generated using the results from 85.1 million speed tests taken during the quarter at the speed test site operated by Ookla. This kind of summary is always interesting, but I’m not sure how useful the results are.

The report looks at both wireless and landline speeds. Ookla says that AT&T was the fastest of the four major wireless carriers in the first quarter, with a ‘speed score’ of 41.23, with Verizon the slowest with a speed score of 30.77. The speed score is a unique metric from Ookla that weights 90% of the download speed and 10% of the upload speed. The reported speeds also toss out the slowest and fastest speeds and concentrate on the median speed.

T-Mobile had the best average latency at 31 milliseconds with Sprint the slowest at 39 milliseconds. The most interesting wireless statistic in the report is called the ‘consistency score’. This is the measure of the percentage of the traffic from each wireless carrier that was at least 5 Mbps download and 1 Mbps upload. AT&T had the highest consistency score at 79.7% with Sprint at the bottom with 66.1%. This score implies that between 20% and 35% of cellular data connections were are at speeds under 5/1 Mbps.

The landline speed results used the same criteria for summarizing the results of the many speed tests. For example, Ookla used the ‘speed score’ that uses 90% of the download speed and 10% of the upload speed – and the results also throw out the slowest and fastest speeds. Verizon had the highest speed score at 117.1, with Comcast and Cox being the only two other ISPs with speed scores over 100. Charter achieved a speed score of 95, AT&T at 82.8, and CenturyLink at 36.1. The AT&T and CenturyLink scores are lower due to customers still using DSL.

Verizon had the best latency at 9 milliseconds, which is a good indication that a large percentage of their customers are using Verizon FiOS on fiber. AT&T and Sprint had the highest latency of the big ISPs at 18 and 22 milliseconds, indicating that the two companies still have a lot of customers on DSL.

The consistency score is more of a headscratcher for the landline ISPs. For example. Spectrum and Comcast had the highest consistency ratings at over 84%, meaning that only 16% of the speed tests on these companies didn’t meet the 25/3 Mbps landline target speed. However, other than perhaps a few grandfathered customers that are still being sold slow products, these companies don’t sell products that should fail that test.

This raises the question of what speed test results mean since there are factors that likely influence the results. For example, I would guess that a lot of customers take a speed test when they are experiencing a problem. I know that’s what prompts me to take speed tests. The other issue that might make Comcast or Charter test at slower than 100 Mbps download is customer WiFi connections. It’s hard to know how many people get slow readings due to poor WiFi. I again understand this issue first-hand. I have a 3-story narrow and long house. The broadband enters on the first floor at the front of the house and my office is at the top of the rear of the house, with some thick hundred-year-old walls in between. Even with an array of WiFi repeaters, the speed in my office varies between 35 and 45 Mbps download – about one-third of the speed delivered at the router. How can Ookla understand the context of a given speed test result? Maybe it doesn’t matter since all of the ISPs have customers with WiFi issues and maybe it averages out. I would think situations like mine are what drive the consistency score. These kinds of questions make it hard to make meaningful sense out of the Ookla results in the report.

Ookla also uses the median broadband speeds to rank the 100 cities with the fastest broadband and also ranks the states. As would be expected, the states in the northeast with a lot of Verizon Fios like New Jersey, Massachusetts, and Rhode Island top the list as having the fastest average broadband speeds. More interesting to me is the bottom of the list. Ookla says that the states with the slowest median broadband are Wyoming, Montana, Idaho, and Alaska. Several other entities that rank state broadband usually put West Virginia and New Mexico at the bottom, followed by Idaho and Arkansas. Those other rankings include an assessment that there are many homes in some states with little or no broadband options at home, while a ranking using speed tests only counts home with broadband.

Overall, this is an interesting way to look at broadband. States with median download speeds under 50 Mbps (6 states) certainly have a different broadband environment than states with the median broadband speeds over 90 Mbps (11 states). But there are places in the highest-ranked states with no broadband options and places in the states with the poorest broadband that are served by fiber.

Will Congress Fund Rural Broadband?

Members of Congress seem to be competing to sponsor bills that will fund rural broadband. There are so many competing bills that it’s getting hard to keep track of them all. Hopefully, some effort will be made to consolidate the bills together into one coherent broadband funding bill.

The latest bill is the Accessible, Affordable Internet for All Act, introduced in the House of Representatives. This is part of a plan to provide $1.5 trillion of infrastructure funding that would include $100 billion for rural broadband. $80 billion of the funding would be used to directly construct rural broadband. It’s worth looking at the details of this bill since it’s similar to some of the other ideas floating around Congress.

The bill focuses on affordability. In addition to building broadband it would:

  • Require ISPs to offer an affordable service plan to every consumer
  • Provide a $50 monthly discount on internet plans for low-income households and $75 for those on tribal lands.
  • Gives a preference to networks that will offer open access to give more choice to consumers.
  • Direct the FCC to collect data on broadband prices and to make that data widely available to other Federal agencies, researchers, and public interest groups
  • Direct the Office of Internet Connectivity and Growth to conduct a biennial study to measure the extent to which cost remains a barrier to broadband adoption.
  • Provide over $1 billion to establish two new grant programs: the State Digital Equity Capacity Program, an annual grant program for states to create and implement comprehensive digital equity plans to help close gaps in broadband adoption and digital skills, and the Digital Equity Competitive Grant Program which will promote digital inclusion projects undertaken by individual organizations and local communities
  • Provide $5 billion for the rapid deployment of home internet service or mobile hotspots for students with a home Internet connection.

This bill also guarantees the right of local governments, public-private partnerships, and cooperatives to deliver broadband service – which would seemingly override the barriers in place today in 21 states that block municipal broadband and the remaining states that don’t allow electric cooperatives to be ISPs.

This and the other bills have some downsides. The biggest downside is the use of a reverse auction.  There are two big problems with reverse auctions that the FCC doesn’t seem to want to acknowledge. First, a reverse auction requires the FCC to predetermine the areas that are eligible for grants – and that means relying on their lousy data. Just this month I was working with three different rural counties where the FCC records show the entire county has good broadband because of over-reporting of speeds by a wireless ISP. In one county, a WISP claimed countywide availability of 300 Mbps broadband. In another county a WISP claimed countywide coverage of 100 Mbps symmetrical broadband coverage, when their closest transmitter was a county and several mountain ranges away. Until these kinds of mapping issues are fixed, any FCC auctions are going to leave out a lot of areas that should be eligible for grants. The people living in these areas should not suffer due to poor FCC data collection.

Second, there are not enough shovel ready projects ready to chase $80 billion in grant funding. If there is no decent ISP ready to build in a predetermined area, the funding is likely to revert to a satellite provider, like happened when Viasat was one of the largest winners in the CAF II reverse auction. The FCC also recently opened the door to allowing rural DSL into the upcoming RDOF grant – a likely giveaway to the big incumbent telcos.

This particular bill has a lot of focus on affordability, and I am a huge fan of getting broadband to everybody. But policymakers have to know that this comes at a cost. If a grant recipient is going to offer affordable prices and even lower prices for low-income households then the amount of grant funding for a given project has to be higher than what we saw with RDOF. There also has to be some kind of permanent funding in place if ISPs are to provide discounts of $50 to $75 for low-income households – that’s not sustainable out of an ISP revenue stream.

The idea of creating huge numbers of rural open-access networks is also an interesting one. The big problem with this concept is that there are many places in the country where there a few, or even no local ISPs. Is it an open-access network if only one, or even no ISPs show up to compete on a rural network?

Another problem with awarding this much money all at once is that there are not enough good construction companies to build this many broadband rural networks in a hurry. In today’s environment that kind of construction spending would superheat the market and would drive up the cost of construction labor by 30-50%. It would be just as hard to find good engineers and good construction managers in an overheated market – $80 billion is a lot of construction projects.

Don’t take my negative comments to mean I am against massive funding for rural broadband. But if we do it poorly a lot of the money might as well just be poured into a ditch. This much money used wisely could solve a giant portion of the rural broadband problem. But done poorly and many rural communities with poor broadband probably won’t get a solution. Congress has the right idea, but I hope that they don’t dictate how to disperse the money without talking first to rural industry experts, or this will be another federal program with huge amounts of wasted and poorly spent money.