Broadband's evil twin, just a fat wasteband
Fred Goldstein, July 2005
... What we now think of as broadband Internet service might not be available much longer in the United States! A combination of greed and politics is leading to a situation where the Internet is being replaced by a different vision of "broadband" ...
.
...If a text message can be sent over IP through a server that the carrier can't bill for, SMS revenues are imperiled ...
... Totally missing from the analysis is the notion of cost. Not price, but the cost that the provider incurs ...
... the Bells have petitioned for forbearance from their common carrier obligations. They don't want to remain common carriers selling wholesale raw DSL to ISPs. Their huge market share of the retail DSL market is not enough for them...
... a content-regulated ISP service, something that is not common carriage but is presumed to follow certain rules, is again a poor substitute for true competition ...
... It's clearly the carriers who want to be enriched, as well as the vendors who will sell them equipment that is far costlier than ordinary routers...
Nowadays the hot word is "broadband", used as if it were a noun. Politicians love it. When the public hears it, they're supposed to think of high-speed Internet access, with the ability to load web pages quickly and download files in a jiffy. It sounds good, and indeed it is good. Fast Internet access is certainly nicer than the slow kind. And because broadband's "always on", there aren't those pesky dialing delays. Nobody wants to go back.
The trouble is this: What we now think of as broadband Internet service might not be widely available much longer in the United States! A combination of greed and politics is leading to a situation where the Internet is being replaced by a different vision of "broadband", one in which a service provider -- the owner of the physical wire -- makes the decisions about what kind of content one can see, and what applications one is allowed to run. It's a bit more like cable TV, but on your computer. There are channels to choose from, but users are merely consumers of content, which is metered out for a price. And what should we call it? It's a fat pipe but the bandwidth is wasted, so "fat wasteband" seems perhaps more appropriate than just plain "broadband".
This evil twin to broadband Internet service would never outsell the real thing in a truly competitive market. But that's no problem for the proponents of this Brave New World; they plan to exert their property rights over the wire to take away choice. This is sometimes called "deregulation". But unlike deregulation's poster child, the airlines, where a competitive industry was freed up to allow prices to be competitive, telecom is still subject to monopolies. Regulation's role should be to rein in the monopolies so they do not abuse their power, not to prevent competitors from offering better deals.
Who got Skyped?
In March, 2005, a venture capitalist, Roderick Randall of Vesbridge Partners, gave a conference keynote speech called "Don't Get Skyped". The conference described itself after the fact like this: "The 21st Century Communications World Forum delivered on its name by offering attendees a bold vision of the technologies and customer needs driving change in telecom carrier networks." What a bold vision he offered! He laid out this fat wasteband message as only a true believer can.Skype, in his world -- and Randall was addressing an audience of telephone carriers and their supporters -- is seen as the archetypal enemy. It is a free Internet application that competes with paid telephone services. Such chutzpah! How dare they! And if Skype can do it to long distance telephony, so can other applications attack profitable carrier services and content. It's all because of that awful Internet Protocol, which allows almost anything to be encapsulated and sent anywhere. It takes away control from the network operators, and that takes away their revenue, to which they are entitled. Networks, in Randall's world, should provide the applications, and prevent competitors from riding their wires.
Now there is some argument, however weak, to be made that Internet telephony applications such as Skype pose such a threat to the elaborate cross-subsidy system behind the telecommunications industry that it should not be supported. Of course we've heard it before, going back to the Carterfone era when Ma Bell argued that allowing customers to own their own telephone sets and answering machines would make telephones unaffordable to grandma. The Telecom Act of 1996 was supposed to get rid of implicit subsidies, but today's incumbent carriers have learned to ignore every part of that Act that they find the least bit inconvenient.
Charging for value instead of cost
But Randall's business model, which seems to be a model that many major carriers have in mind, goes well beyond Skype. Take for instance email. Randall described how it competes with carrier-provided services such as SMS. The price per bit of an SMS message is very high. The most common such message is "OK". Most wireless carriers sell a text message like that for two to ten cents. SMS is capped at a few dozen characters, so that dime can't send too much. Now contrast that with the price per bit of a 2 1/2-G wireless data service like 1xRTT. An awful lot of bytes can get sent for the same price as one SMS. By focusing on lost potential revenues rather than profit margins, this lower-cost service is perceived as costing the carrier money! And 3G's relatively cheap broadband is thus seen as a profit killer, if treated as an ISP. If a text message can be sent over IP through a server that the carrier can't bill for, SMS revenues are imperiled. Even MMS, the multimedia high-bandwidth descendant of SMS, is, in this netherworld of premium bits, a step towards perdition, as it has a far lower price per bit.Totally missing from the analysis is the notion of cost. Not price, but the cost that the provider incurs when the ratepayer (it's hardly fair to call the victim of such schemes a "customer") sends a message. SMS, for instance, is extremely costly to provide. The message rides on the Signaling System 7 network, a specialized, secure, ultra-reliable realtime network designed for call setup messages. It isn't IP, and it isn't designed for cheap bits. MMS uses the lower-cost GPRS network, and email rides the public Internet, where bits are dirt cheap.
Randall was not original in making this argument. Charging for "value" is an old telco tradition, still present, for instance, in the access charge system that classifies calls as local or long distance. Broadband carriers, in such a traditional Bellheaded view, shouldn't charge based on their costs, rather, they should realize "value", just as many telephone companies still charge less for lines in high-cost rural areas than in low-cost urban areas, where the line is putatively "worth more" because there are more lines in its local calling area. So at one extreme, the price of SMS is very high, which might be discouraging its use. At the opposite extreme, video downloads use so much bandwidth that the price per bit must be extremely low, probably below cost, or they won't happen. Therefore the price of video bandwidth needs to be lowered to generate business. The fact that the carrier would lose money on each video sale is irrelevant; they could make it up by charging monopoly rents on every "message" sent, be it SMS, MMS, or for that matter a page hit on MSN Hotmail or Gmail. Monitor and cross subsidize everything, yeah, that's the ticket. It kept Ma Bell fat for decades, right?
Telco broadband operators are naturally tempted to protect old revenues, such as telephony or even home alarms, when fishing for new ones. For a few years now, companies have been trying to sell Layer 7 filters to wireless network operators, precisely to prevent this type of revenue attrition by either blocking or metering application usage. Wireless service providers are only loosely regulated, at least in the United States, and could get away with it, especially if there weren't much competition. Even Randall's SMS cost slides were unoriginal, taken from a company presentation that I happened to see a few years ago. But not all that much filtering goes on yet (though countries like China and Iran have embraced it), and that company's bones were recently acquired, not all that surprisingly, by a Vesbridge portfolio company. At least we know why Randall's so enthusiastically flogging the concept.
The "Don't get Skyped" argument extends this across the wireline world, to "broadband" services. Subscribers wouldn't be allowed to watch Internet videos, for instance, without a pay-per-view fee to the network operator, even if the network were merely a bit pipe. After all, it can be suggested, one could and should have only watched what was on the provider's network! And why, it follows, should someone be allowed to shop at an online store if it doesn't pay a cut of sales to the wire owner? That's how cable shopping channels work, after all.
Randall's view of the public Internet is that it should only be made available on a low-priority basis, if at all. Carrier-provided pay-to-use services should get broadband speeds. The rest of the world rides the network as "cargo class" at best, or is perhaps relegated to "hobo class", where non-carrier-provided applications are treated as freeloaders. You may have fiber to the home, but it's not the carrier's application you're running, then you'll be lucky to get dialup speeds.
Now in a normal business environment, where there's competition, the price of a service or good tends to align with the cost or producing it. This is the basic principle outlined by Adam Smith in The Wealth of Nations back when George Washington was still sleeping around. Progress results in lower costs, which enables more consumption of more goods and services. Markets take care of things. This is sort of how the Internet works too. To be sure, below-cost pricing has led to huge losses on the part of some Internet companies, but that's what often happens in a new industry. ISPs don't claim to be entitled to profits. If video streaming is too costly to be profitable, then perhaps it shouldn't be the norm yet, and normal economics don't permit a competitive business to charge its subscribers for every spam received.
Innovation comes from the bottom up
This view of how to make "broadband" profitable clashes with the way technology actually works. Unrestricted bandwidth has been an enabler of innovation. Randall said, "Innovation will only happen when there is demand that is being completely unmet". Imagine, seeing supply only following clearly-delineated demand! That is, I think, how Gosplan (the Soviet Union's central planning agency) operated. It certainly is not how the innovation-driven economy works. Not even China's. Where is the notion of demand meeting supply, once supply of something new comes along? That has driven the semiconductor industry, and indeed the components industry in general, for decades. Again, it was a V.C. talking, and perhaps he felt burned by the wacky investments of the 1990s, for which there really was no demand. (The fact that supply doesn't always create demand doesn't mean that supply never does.) But he used to be with Lucent. It figures. When is the last time Lucent innovated on its own, rather than buying a company that already took the risk? Those "Bell Labs Innovations" have been pretty scarce lately.IMS and IPsphere will be enabling technologies
For several years now, the standards body for "3G" mobile telephony, 3GPP, has been working on a set of standards called IMS,
for IP Multimedia Subsystem. This is an "operator-friendly
environment" intended to "generate new revenue via deep packet
inspection" . But the idea's catching on in the wireline
world too. A bevy of wireline carriers and equipment vendors have
created the IPsphere, a
new set of standards for network intercession in IP application
flows. It's couched in frou-frou language about an "enriched
experience for consumers", but it's clearly the carriers who want to be
enriched, as well as the vendors who will sell them equipment that is
far costlier than ordinary routers. IPsphere proponents, like
CIMI's Tom Nolle, describe
the Internet we know and love as a "utility" and thus a "disaster", and
promise that "IPspheres make the carrier a player in the higher-margin
services and not just a conduit for their fulfillment." In case
IPsphere's position isn't clear enough, its founding press release
described one of its goals as "Ensuring that users of the IP
infrastructure can access only the presented services, and not the
underlying infrastructure itself".How would a fat wasteband scenario like Nolle's or Randall's -- and I don't mean to harp on Randall personally, since he's just the one vulture in the flock who stood on stage and proclaimed the message to an assembled multitude -- sell in world of competitive ISPs? Not at all. The Internet is a much bigger business than pay per view. But ISPs don't make high profits, so "broadband" providers shouldn't be competitive ISPs. They need a monopoly. Real ISPs thus need to be gone. Dead, every last one. That's the only way it could fly.
FCC forbearance can make this nightmare come true
And that's why what the FCC does this year matters so much. The Supreme Court has upheld, in the BrandXdecision, the FCC's earlier distinction between cable and telephone companies. Cable modems were never built as common carriers; they were built as competitive self-provisioned broadband ISP services, when ISPs had clear access to the ILEC networks. The BrandX decision, if unpopular among ISPs; met the Supreme Court majority's Chevron analysis (upholding a standard that a regulatory agency has broad leeway to interpret ambiguous law, even if it did not make the "best" interpretation) of the notoriously ambiguous Telecom Act. It didn't take anything away from ISPs; it just left cable alone.In contrast, telephone companies, as common carriers of long standing, have been expected to sell (for a profit; nobody says for free, and they even get to set their own prices) DSL access to any and all ISPs. But the Bells have petitioned for forbearance from these common carrier obligations. They don't want to remain common carriers selling wholesale raw DSL (telecommunications service) to ISPs. Their dominant market share of the retail DSL market is not enough for them; even the spectre of an ISP competitor on their wire prevents them from adopting Randall's evil broadband model. Chairman Martin has expressed sympathy with the Bell's demand for favored treatment, and seems predisposed to grant their wishes. If cable companies don't have to be a carrier of last resort to ISPs, their arguement goes, why should anyone?
The Bells want to be like cable companies, allowed to monopolize the content, the ISP layer. (Not that cable companies have broadly degraded their own ISP services along Randall's lines. For one thing, they currently compete with all of the ISPs who can still ride DSL. For another, cable companies are usually run as real businesses, less concerned with "value" than profit margins, and less disposed to speculative investments like IMS/IPsphere filters. But deliver us from temptation: Duopolies can get very cozy, and some cable companies are financially desperate enough to be willing to try such stunts.) So the consumer would end up with only two choices of ISP, the cable company and the telephone company. And with such limited competition, collusion, at least in some markets, is quite imaginable. If both decide that fat wasteband with Layer 7 filters, value pricing and pay-per-view content is more profitable than high-speed Internet access, then that's what the public will be stuck with. Sure, it's theoretically possible for additional carriers to spring up, pulling new wire on poles or trenching the streets. And there are narrow, geographically-limited opportunities for wireless ISPs to continue to operate. But one obvious option, for municipalities to pull fiber to the home themselves, is under attack from Bell lobbyists who have gotten several state legislatures to ban it. And there's even a bill in Congress, brought by a former SBC executive now in the Texas delegation, to prohibit local governments from providing any telecom or information service that a private company might be interested in providing. Even to themselves.
A few years ago, pundit David Isenberg, a disgruntled former AT&T employee, wrote an essay, "The Rise of the Stupid Network", essentially the Internet model of the world, where the content and applications were all outside of the network and anyone could contribute. I think he got carried away in attacking the "Intelligent Network" (IN), a straw horse view of the telephone network, where that term refers to a set of features used to help set up calls, provide number portability and caller ID, and other simple connection-management services. The "IN" model of telephony doesn't let the carrier touch the content, ever, at all. Nor does the Stupid Network; the latter just does it differently. Randall, though, called for "The Fall of the Stupid Network"! He wasn't praising IN or ISDN or ATM or other non-IP data technologies; he was calling for an end to transparent common carrier pipes.
An alternative view of the world is being socialized now by a number of parties, who accept the notion of a monopoly or duopoly of ISP service, but who want to protect the parasitic VoIP operators such as Vonage and Skype who are the most obvious targets of the garden walls. Various organizations are drafting or suggesting legislation that includes such terms. Creating a content-regulated ISP-like service, something that is not common carriage but is presumed to follow certain rules, is again a poor substitute for true competition. Who, in 1993, could have forecasted the way the ISP industry would turn out? Who even today can forecast the impact of, say, streaming video service on ISPs? A fully competitive marketplace irons this out, allowing, for instance, some ISPs to block high-rate video streams in exchange for a lower price. Telecom providers, though, should pass bits along without knowing or caring what they're for.
To be sure, an IPsphere or IMS world would indeed have plenty of "broadband". But it would be like voting for a government that promised to make the trains run on time. Even if they fulfilled the promise, the cost would be devastating. It should never be allowed to happen.
