November 14, 2021

We're All Connected

”…New York Telephone.” The old jingle from the local phone company of my childhood is still stuck in my head. I’m not sure why they felt the need to advertise given their monopoly on phone service in most of the Empire State, but they did. As did the long-distance companies, with their endless promotions for their low per-minute rates, and the dueling collect call services, and those weird “10-10” numbers you could dial to use a different long distance provider for a particular call. Odd how all of that, which seemed so ubiquitous and permanent at the time, faded away so quickly, along with the traditional landline itself. We’re still all connected, and many of us use the company that used to be New York Telephone (they call themselves Verizon now) but it feels like a different world from the one with dial tones, where those noises the phone keypad makes actually did something.


In 1844 Samuel Morse and his associates sent their first message by telegraph, What hath God wrought, from Washington to Baltimore over a line the federal government had funded as a pilot project. The Morse syndicate had hoped to sell their technology to the Post Office who would build a nationwide network of telegraph lines. But the government wasn’t interested in buying, so the syndicate went about building the network itself. Thus was born the American telecommunications industry.

Considering that the Morse syndicate hadn’t originally planned to go into business for themselves, and that there was no precedent for an industrial concern on a nationwide scale like they were attempting, it’s unsurprising that they made a number of missteps in the early days. Lacking enough capital to build beyond the first line, they sublicensed their technology to outside investors who established their own businesses. The first Morse business, the Magnetic Telegraph Company, stuck to the Northeast while licensees and sublicensees built the network westward. The problem was that their contracts were sloppily drafted and granted overlapping rights, leading to a situation where multiple competing telegraph companies would build in the same territory, each claiming to have a valid right to use the technology.

This was the context for the seminal case in American patent law, O’Reilly v. Morse. Henry O’Reilly held an early license to the Morse patent but soon went rogue and started competing against the companies of the Morse syndicate. When Morse sued him for patent infringement and won, he appealed to the Supreme Court, arguing that the patent was invalid in its claim to all possible methods of sending information over a wire. This claim, I learned in the first weeks of patent law, was struck down for going beyond the scope of Morse’s invention. No inventor could claim exclusive rights to an abstract scientific principle, because no one could invent such a principle. What I wasn’t taught in law school was that although O’Reilly won on this one point, the Court upheld the rest of Morse’s patent and found that O’Reilly had infringed it. This was in 1854; the following year an O’Reilly company and a Morse company serving the crucial Midwestern market merged to form Western Union. Within 15 years Western Union had acquired every other significant telegraph company in the country.

Unlike many of the later utilities, Western Union never had a government-granted or enforced monopoly. It emerged from a chaotic period when different telegraph lines would engage in circuitous routing to avoid a high-priced competitor, who would respond by buying up the less expensive but longer lines and bringing their prices in line. In the late 1850s the major regional telegraphs (the “Six Nations”) reached an agreement on rates, interconnections, and staying out of each others’ territories; Western Union, through its control of the line to the West Coast, soon gained enough capital to buy out the other “nations.” The last holdout, incidentally, was the American Telegraph Co., the direct successor to Morse’s Magnetic Telegraph, whose main line up and down the East Coast had been severed by the Civil War. But from then until the telegraph became obsolete, there were still rival companies cobbled together from failed and discontinued lines who tried to compete with WU. Most failed quickly, but Postal Telegraph stuck around for a few decades.

Meanwhile the governments, state and federal, were only able to impose weak and ineffective regulations, in no small part due to WU’s aggressive bribery lobbying of legislators. The Post Roads Act, for instance, was initially proposed to create a rival National Telegraph Company with the right to run its cables along any public right-of-way free of state or local interference. By the time it got out of Congress, it instead allowed any telegraph company to claim that privilege in exchange for discounted rates for government business. So naturally WU took advantage of this law, used it to assert immunity from state regulation (mostly unsuccessfully), and of course quickly bought out National Telegraph.


So with one behemoth in place, we set the stage for the next behemoth: the Bell System. This story begins with Alexander Graham Bell offering to sell Western Union the rights to his research into telephony, and WU turning him down. This was not quite as short-sighted as it seemed: Elisha Gray also did pioneering work into telephony at a WU affiliate, Western Electric, and Bell and Gray would ultimately file with the patent office on the same day. This led to another landmark case in patent law in which the Supreme Court upheld Bell’s patent and declared he had priority over Gray. By that time, Bell and WU affiliates had already started telephone service in major cities. Shortly afterward, the patent case was settled with the newly organized American Bell Telephone Company taking exclusive rights to the telephone and its affiliates taking over the larger WU telephone companies.

The Bell System had a monopoly due to Bell’s patent, but only until it expired in 1894. At the time it was much more loosely organized than it later became. Local companies licensed the rights to the telephone from American Bell - and unlike the Morse Syndicate, their contracts were written well enough to prevent any unwanted competition. But the central American Bell only took minority stakes in most of the locals and they operated autonomously. There were, however, two important moves during this period that would lead to Bell’s later dominance: the acquisition of Western Electric, which became the dominant producer of telephone equipment, from WU, and the establishment of a subsidiary, American Telephone & Telegraph, to build and operate long-distance lines.

(You may be wondering - wasn’t AT&T at the top of the Bell System? That was a little while later. See, American Bell, being headquartered in Boston, was a Massachusetts corporation, and MA corporate law limited the rights of corporations to hold stock in other corporations and capped their size at a lower level than American Bell wanted to be. AT&T, a New York corporation, was under much less restrictive laws in this regard. So in 1899 American Bell and AT&T merged, with AT&T and its less restrictive charter the survivor. For similar reasons Delaware, with even more permissive laws than New York, became the dominant jurisdiction for incorporation around this time. Everyone thinks it’s for lower taxes but it’s not.)

From the expiration of the Bell patent in 1894 until the early 1920s, many cities had two telephone companies, a Bell affiliate and an independent. Often they were complementary services: Bell focused on business customers and most independents were aimed at residences, “Home Telephone” being a common name for them. But if you wanted to call businesses and residences, you needed two telephones, since the networks weren’t interconnected. Furthermore the long distance lines gave Bell an edge among business customers, and Western Electric often wouldn’t sell equipment to upstart challengers. After Theodore Vail became president of AT&T in 1907, Bell’s tactics became far more aggressive, locking out independents from equipment and interconnection entirely, as well as taking increasingly large stakes in the local Bell companies, centralizing management, and buying up independents whenever they came up for sale.

But this was the Progressive Era, antitrust enforcement was the rule of the day, and the Attorney General threatened to prosecute. Washington took a particularly harsh view of AT&T buying a large stake in Western Union, by then weaker than in the 19th century but still one of few companies with enough of a nationwide presence to challenge Bell. (What would’ve happened if WU had aligned itself with the independents and built a competing long-distance network? That’s an interesting one to speculate about.) The case was settled when AT&T agreed to the “Kingsbury Commitment”: they would sell their stake in WU, stop buying independent companies, and allow independents to interconnect with the long-distance network. This placated the government but the independents were displeased. Many of them wanted to be bought, and long-distance calling was far less important than local calling to most customers. A few years later Congress passed a law permitting AT&T to buy independents, subject to approval by the Interstate Commerce Commission, which by then had been given regulatory authority over wire communication. The “wild west” was over and the age of regulated utilities had begun.


For most of the 20th century, it worked like this: AT&T controlled the long distance lines and every local Bell company except two. Western Electric made the phones, which AT&T would rent (but not sell) to customers. Rates were fixed by government-approved tariff. Relationships between the Bell system and the independent companies (which were mostly in rural and suburban markets, but also the Gulf Coast of Florida and Las Vegas) were dictated by various regulations and antitrust settlements that accumulated over the years. For most services there was no competition.

Many at AT&T chafed at the restrictions of being a regulated utility, and when another antitrust case rolled around in the 1970s, they decided to sell off the “natural monopolies” of the local Bells and open themselves up to more competition in the long-distance space, in exchange for which the restrictions on doing non-telephone business would be lifted. The Unix operating system, which AT&T initially had to distribute under permissive terms and had become wildly popular, could become a profit center. And with the resources and expertise of Bell Labs behind it, the new Western Electric 3B computer would surely make AT&T the next DEC or Sun, right?

Well, setting aside how it ended for DEC and Sun, AT&T never reached anything near their heights in the computer business. They ended up selling Unix to Novell and spinning off Western Electric under the name of Lucent, which had a short undistinguished existence selling telecom equipment before merging into Alcatel and then Nokia. The long-distance business, meanwhile, saw AT&T in fierce competition with a pair of upstarts. MCI had originated as “Microwave Communications, Inc.” and supplemented its microwave radio network with a number of shrewd asset purchases, including most of what used to be Western Union’s communications network. (WU was on the verge of bankruptcy and realized their only profitable business was money tranfers, so they sold off everything else.) Sprint originated as the *S*outhern *P*acific *R*ailroad’s *int*ernal communication network, but by the time the long-distance price wars kicked into gear, had been taken over by one of the larger independent telcos (United Telecom) which rebranded itself under the Sprint name. The three rivals could be seen on every TV commercial break with Candice Bergen and Whoopi Goldberg and Paul Reiser selling their respective discount calling plans. Prices dropped, call volume soared, capitalism did its thing.

Why not, thought the government during this idealistic neoliberal time, let competition do the same thing for local telephone services? This was the main thinking behind the deregulatory Telecommunications Act of 1996, which established the modern FCC legal regime. It was all about forcing the Baby Bells, sold off by AT&T in the 1984 break-up, to open up their local phone lines to competitors, who had to be allowed to interconnect at the source. The Bells themselves would no longer be limited to their regulated businesses. State and local laws standing in the way of new entrants into the landline business were explicitly preempted. Soon, thought the regulators, we’d have nationwide competition and ever improving service on landlines.

Of course it didn’t happen that way, but the regulations are still written as if it would. All kinds of alphabet soup about ILECs and CLECs and LATAs, which proved irrelevant as nobody wants landline service. Meanwhile the Internet, and anything going over the Internet, is considered an “information service” and exempt from FCC regulation. This made sense in 1996 when the dominant mode of Internet access was dial-up and anyone could dial into any ISP over their landline. Nowadays, if you’re lucky your former landline company has made fiber available and you have a choice of them or your cable company. If you’re unlucky, you’re stuck with cable or a much slower DSL connection which your telco is intentionally making worse so they can get rid of copper and the 1996 regulatory requirements once and for all.


Should ISPs be regulated as utilities, made to publish detailed price lists and make service available on the same terms to all? The Obama FCC tried to put “common carrier” status on Internet service providers, which the Trump FCC promptly reversed. The effects of these swerves haven’t been obvious to me and are hotly debated among others using “net neutrality” and other buzzwords. As it stands, there are no legal requirements for interconnection or standards for service, and the various ISPs and backbone operators have opaque deals with each other about who gets paid to carry what traffic. Details of these agreements are kept quiet absent a carriage dispute, and outside observers have so little light into how things work that there’s no definitive list of the “Tier 1” backbones who peer with each other on a cash-free basis. Despite the illegibility of the system, it all works remarkably well.

The public switched telephone network, on the other hand, is a mess. As I mentioned above, all the regulations assume things still work like ‘90s landlines, there are local and long-distance calls and you have separate carriers for each, and it’s all over circuit-switched SS7. The reality is that virtually all “landlines” are VoIP, and since 4G all mobile phones are VoIP too. This is a good thing: call quality is much better and it takes up so little bandwidth it’s become too cheap to meter. But it’s still metered. Sure I’ve got unlimited domestic minutes on my cell phone, but every international minute will cost me. Why can’t I call any phone in the world? For that matter, why can’t I call any computer in the world? Or set up Asterisk on my Raspberry Pi and get an adapter and plug in a classic Model 2500 phone and chat old-school?

Part of it is that PSTN-interconnected VoIP, which was initially considered an “information service” exempt from FCC regulation, has started to be treated as a “telecommunications service” for some purposes. Notably a VoIP company has to provide the same 911 emergency service as traditional phone companies and it has to contribute to the Universal Service Fund. These rules have noble intentions, sure, but they do create a barrier to entry.

Not enough of a barrier to stop the plague of robocallers with their warranty scams from tormenting us. Most internet-facing protocols have developed defenses and security mechanisms to protect from attackers. SS7, which still controls enough of the PSTN to be essential, comes from an earlier time and a much more restricted environment, and blindly trusted that any incoming call from another carrier would be legitimate. If there’s even one sleazy VoIP carrier to look the other way it’s easy as pie for scammers to flood the network with garbage. And thanks to the FCC’s deregulation of access, VoIP interconnection agreements are the same kind of secretive, opaque arrangement as peering between internet providers, meaning it’s not easy to tell how many VoIP carriers there are or how they get their access to the PSTN. The implementation of security protocols and fraud blockers on cell phones has lessened the problem, but given how much energy has gone towards regulating traditional phone service, should it really have gotten that bad?

The real mess, though, is video calling. Oh it’s nice that a basic video call is free, but how should I call? There’s Zoom and FaceTime and Skype (which may or may not be the same as Microsoft Teams) and a half-dozen half-abandoned Google products, I need to either download all these apps or use a less functional in-browser app (and I’m old enough to remember how crummy web apps used to be, I still go with the native app whenever I can)… and apparently they’re all internally almost standard SIP but with enough modifications to be incompatible. Maybe this is silly to complain about given how much better it is than the old days, but… a phone call is a phone call, no matter what phone or software you use. A video call should just be a video call.

This will probably all be worked out someday. Or not. Who knows.


This started with telegraphs and Western Union. When WU left the communications business, they vacated their headquarters at 60 Hudson Street in lower Manhattan, which also served as a hub for their communications network. The various companies that bought pieces of WU got parts of the 60 Hudson facility, and it has since grown into one of the largest data centers in the world.

It’s all connected.

© 2020-21 Bronx River Software

Powered by Hugo & Kiss.