Why regulate cable internet access
The cable guys have their way of saying it: “What do you want to do, nationalize our businesses?”
Another way of seeing this issue is: We have a very few very large providers of highspeed internet access in this country. They have sufficient market power to decide how and when to prioritize internet communications. And all of these providers are competing with the internet in some way - they are all (or are becoming) old media and old telecom companies that want to maintain control over their distribution channels. The internet disrupts this control, and so they are competing with it.
This market power over access, and prioritization power, is good for these few providers but not good for the rest of us. It threatens overall economic growth and the development of new kinds of businesses and new ways of making a living.
Yes, we limit the powers of our government, and we’re proud of that, but those limits do not forbid us from meeting modern conditions.
Even ten years ago, the thought of requiring nondiscrimination of cable providers would have been rejected as arbitrary and oppressive. Well, the thought of zoning regulations would have been unthinkable in 1910. The thought of regulating the upbringing of cows would have been unthinkable in 1920. But we did these things because it was better for the rest of us.
I understand that Comcast saves a dime or two a month per subscriber by doing the kind of peak-time traffic shaping that is the focus of the current uproar. That’s not insignificant - it may add up to between $5 and $20 million dollars a year. But the service they’re selling brings in over $5 billion. And the savings to Comcast is an incalculable cost to Vuze and other new video-distribution businesses that can’t rely on a stable transport platform. It’s also an incalculable cost to the rest of us, even if we’ve never heard of BitTorrent, because we can’t rely on a fast, non-discriminatory internet connection for our future.
This wouldn’t matter if we had enough choices of internet access providers in the U.S. We don’t, and so the uproar continues.
Comcast isn’t acting as just some old private business when it provides constrained internet access. It’s providing communications infrastructure, and we don’t (and shouldn’t) expect our communications sidewalks to rise up and choose winners and losers.
Yes, uncontrolled government regulation (the “nationalize our business?” horrible) is wrong. But we do regulate, when we need to, when it’s for the common good.
Three developments
1. More passive content from network providers. Comcast announced that it’s going to be providing 3,000 high-definition video-on-demand programs for subscribers to its highspeed Internet access services.
“Comcast is the largest purchaser of TV content and now we are bringing that content over to the Internet” [Comcast CEO Brian Roberts, at CES today]
Comcast is also confidently predicting that the PC will become “a full cable TV client” in the future. Dirk van der Woude pointed me to a Wall Street analyst, Douglas McIntyre, who doesn’t think Comcast’s plan makes any sense.
Why not buy some more content? According to The New York Times “Comcast is already the world’s largest buyer of content, and it is spending about $4.5 billion a year to assemble content from around the world to offer on demand.” But, all of that content, even delivered via cable and the internet, may not get Comcast any new customers and may not be the magic bullet that kills new products from the telephone companies.
A wide range of video is already available on the internet. YouTube to Hulu, AOL to MSN. The largest content firms are already pushing TV and films across dozens of delivery portals. The new Comcast online product would not seem to have much advantage here.
On the TV, VOD has always been an attractive product, but the reason most companies do not offer huge film libraries is that no one wants them. Consumers watch the most popular movies. Having an extra 5,000 films that 1% of the subscriber base wants to see is hardly a solution.
Well, we’ll see. Comcast’s plan is to create a platform of its own, a one-stop shop that you can drive from your PC - checking mail, making phone calls, and watching hi-def video. That could be quite attractive to users, particularly if they don’t have any easily-available alternatives.
2. The Chairman promises to look into Comcast’s blocking practices. FCC Chairman Martin says he wouldn’t mind investigating what Comcast was up to.
“Sure, we’re going to investigate and make sure that no consumer is going to be blocked,” Martin told an audience at the International Consumer Electronics Show.
The insouciance, the blitheness of that “Sure” tells us that “blocking” is going to be looked into. But so much happens short of blocking - so much can get in the way that doesn’t amount to an outright ban of particular applications or activities or devices. And Martin will make, and the network providers will make, huge exceptions for “reasonable network management.” Undefined. Surely we can do better. Maybe Obama would do better.
3. All Quiet at Frontline. I understand that Frontline didn’t make a required downpayment to bid in the 700 MHz auction. That means it’s even less likely that we’ll see any experimentation with internet access business models - even more likely that Verizon or AT&T will win what they want in the auction. What they really want, perhaps even more than the spectrum itself, is freedom from the specter of wholesale open access requirements. Like Comcast, they want to continue to be in the business of managing a vertically-integrated network of content.
All three of these developments point in the same direction: putting the broadcast model of the internet on a firmer footing. Yes, the Commission may look out for adequate disclosure of what that management consists of, but because there isn’t any competitive pressure to provide an unmanaged internet environment, we won’t expect one. And we won’t get one, unless different leadership emerges in this country.
Boundaries
It’s the ad hoc nature of U.S. communications law these days that gets depressing. It seems only federal courts can help - except when they refuse to get involved.
Four very quick snippets of stories to watch:
1. Warshak. The Sixth Circuit said back in June that people have a reasonable expectation of privacy in email sent through ISPs, and so the loose-and-low statutory procedures for government access to these emails (found in the Stored Communications Act) have to be measured against Fourth Amendment standards. The government had used those loose-and-low procedures to obtain a court order giving it general access to Warshak’s emails stored by ISPs - and the government didn’t tell Warshak it was doing this for a year. That failed the Fourth Amendment test, according to the Sixth Circuit.
Although the particular loose-and-low procedure (a one-sided appearance before a judge stating “reasonable grounds” rather than probable cause, the Fourth Amendment requirement) has been declared unconstitutional, there are still plenty of other ways for the government to get access to these materials if they need it. They can get a warrant based on probable cause, or go in front of a judge and let the search-ee know that they want access and are asking for a subpoena. (Meanwhile, judicial steps can be taken to avoid having the emails disappear.)
But that’s not enough for our government, which wants to avoid any additional oversight or notice requirements in connection with access to email.
Everything’s in email - a complete dossier of our communicative lives is in email. What could be more personal? Years ago, we agreed that people had a reasonable expectation of privacy (triggering Fourth Amendment protections) in phone calls - why would we treat emails differently? But our government has successfully sought rehearing of the Sixth Circuit case by the full panel of judges, and it’s likely we’ll see some clever fight over the “ripeness” (liveliness, concreteness) of the claim that avoids the merits of the dispute but keeps the status quo of easy access in place.
Ripeness is a sophisticated way of saying to the courts: “Move along here, now, nothing for you to work on.”
2. Hepting. That’s the name of the lawsuit addressing the NSA surveillance scandal. At the direction of our government, telecommunications companies copied all online transmissions crossing their network — wholesale — and sent that copy on to the NSA for further processing. This likely violated the Fourth Amendment (clearly the telecom companies were acting on behalf of the government), the Telecommunications Act, the Wiretap Act, the Foreign Sovereign Immunities Act, the Stored Communications Act, and state unfair competition/deceptive practices laws. At the least.
Right now, Congress is considering whether to grant retroactive immunity to the telecommunications companies that participated in this warrantless vacuuming up of all possible communications. A Senate panel has voted against immunity today — a welcome development. Sen. Specter (R-PA) has argued that Hepting (and cases like it) need to proceed. “[C]ourt cases may be the only way Congress can learn how far outside the law the administration has gone in eavesdropping.” Specter wants to substitute in the government for the telcos, though, which seems problematic. Why should citizens have to pay for this illegality? A House bill on the same subject already rejected immunity.
Retrospective immunity is another way of saying to the courts: “Move along here, now, nothing for you to work on.”
3. Title I and Chevron Deference. At the same time all communications have become IP-based, the FCC has manuvered regulation of those communications out of the scope of its own delegated power from Congress. Instead of keeping things within Title II of the Communications Act (the home of some heavy-handed regulation, to be sure, but at least there were guideposts for the FCC’s action), the Commission has declared that just about everything having to do with the internet and access to the internet is within its power under Title I of the Act.
Title I says nothing. So the Commission has enormous discretion to do whatever it wants - it’s a swamp, a murky, bottomless realm of unaccountable action, that Title I.
Here’s the place where (unlike Warshak and, potentially, Hepting) the courts have deferred to the broad exercise of communications discretion. Somewhat enigmatically, Justice Thomas in 2005’s Brand X decision said:
Information-service providers . . are not subject to mandatory common-carrier regulation under Title II, though the Commission has jurisdiction to impose additional regulatory obligations under its Title I ancillary jurisdiction to regulate interstate and foreign communications.
Justice Scalia thought that was weak, and said so:
This is a wonderful illustration of how an experienced agency can (with some assistance from credulous courts) turn statutory constraints into bureaucratic discretions.
I’d like to see another case that makes a court address the scope of the Commission’s power over the internet — surely the Commission can’t act without some kind of delegated authority. Surely there needs to be a guidepost in the swamp somewhere.
Right now, a combination of judicial deference and Commission brashness is keeping the courts - and Congress - from getting involved.
4. 70/70. Title VI of the Cable Act says:
[A]t such time as cable systems with 36 or more activated channels are available to 70 percent of households within the United States and are subscribed to by 70 percent of the households to which such systems are available, the Commission may promulgate any additional rules necessary to provide diversity of information sources.
We know that the Commission plans to say this “70/70″ trigger has been met. It’s true that cable has enormous market power and the empirical step makes sense.
But “any rules necessary”? I’d be worried, if I were the cable industry. Again, the issue here is a complete lack of boundaries. Not even a hint of limitations.
And, again, the argument from the Commission will be: “Move along here, courts, nothing for you to work on.”
==
These are all separate stories, each with its own history and set of acronyms. They all share, though, a certain open-endedness and ad hocery that is distressing. Not that every detail needs to be written down in legislation - but some checks, some examination has to happen at some point, provided by some institution that isn’t pressing for action.
Comcast Is Pretending to be You
This AP story makes clear that Comcast is pretending to be part of online conversations in order to frustrate users who want to use particular online applications. This happens all the time in the name of “traffic shaping” — it’s the kind of thing that China does to interfere with internet use. What’s different and important about today’s story is that people have carefully experimented. We can now understand exactly what Comcast is doing.
When you go online and click a link, what you’re doing is sending packets (think individual pages taken from a long, handwritten letter) to a machine connected to the internet. What we call “the internet” is a very simple agreement: machines agree to chunk things into packets and label those packets with unique numbers (think return address and sender’s address). Then those packets travel the best available route to the machine they’re addressed to, and that machine reassembles them.
This agreement to chunk things into packets that self-describe their destination (at a unique global address) is known as TCP/IP. IP, or Internet Protocol, is the addressing scheme — the numbers.
IP doesn’t do anything about accuracy - it doesn’t provide any way to check that all the packets have gotten where they’re supposed to go or that they’re in the right order. That job (roughly speaking) is carried out by the TCP part of this — the Transmission Control Protocol. TCP receives a stream of information from an application (say, your web browser) and divides it into packets. It gives each packet a sequence number. TCP then hands packets to the Internet Protocol for delivery through the network. TCP also opens a “window” for the number of packets that will be sent out - you wouldn’t want to send a zillion packets without acknowledgement that they had been received.
The TCP module at the receiving end of the communication does this acknowledgement job, noting that a particular number of packets have been successfully received. All of this is done very politely, quickly, and electronically — the conversation between the home TCP and the remote TCP is established, an acknowledgement is received, the conversation begins, and sequences of packets are sent. If packets are lost along the way, they’re retransmitted. When an endpoint wants to stop, it lets the other endpoint know that it’s done.
Each header in a TCP-labeled packet (think front of an envelope) has a number of fields. One of these fields includes “flags” that are applied to the packet. One of these flags is called RST, for “reset the connection.”
The Comcast system (probably provided by Sandvine, according to the Times) was setting the RST flag for both sides of any communication that it believed (probably through traffic analysis) was using Bittorrent.
So when “you” the Comcast subscriber were clicking links that were part of a Bittorrent transaction, Comcast was slapping an RST flag on your packets. And any packets crossing the Comcast network that were coming from the “outside” but were part of this conversation were also having the RST flag slapped on them as they crossed into Comcast territory. Neither user had any idea this was happening. They could just tell that things were moving really slowly and then stopping, as both machines politely agreed to reset themselves - thus cutting the conversation off.
It’s as if as soon as you entered a room an enormous “Loser” tag was stuck on your forehead unbeknownst to you. Sure, you could continue to circulate, but no one would talk to you. Or, if that’s too awkward, try this: it’s as if someone else that sounded like you got on the phone as you were talking to your mother and said “We need to hang up right now.”
Like the Verizon/NARAL flap and the Pearl Jam escapade, here’s another story about currently-legal action, permitted under someone’s elaborately-walled Terms of Service, that interferes with basic communications. Comcast will say “we’re not blocking.” But they’re degrading, prioritizing, and filtering, without telling users. And they’re planning to do much more of this.
What’s the solution? Structural separation. You’re either a plain-vanilla transport company serving all comers, or you’re something else competing for our attention. But this mixture, this hybrid of apparent-communication plus editorial control, is unacceptable.
The big picture: Why the Verizon/NARAL flap matters
I arrived in DC in the middle of last night without a phone charger, having left mine in Manhattan last weekend. So this morning my first stop was a Verizon Wireless store downtown. Right in the store, in a corner, I plugged in my phone and called back someone who had a lot of questions about yesterday’s kerfuffle. Sitting on the floor, I tried to explain why this story matters.
[The convenience of the VZ store downtown (and in fact the ubiquitousness of those stores in most cities in the US) is a telling sign, so although I certainly recognize that it might seem like abusing their hospitality and electricity to rail about them from their own sales floor I decided to use it as an element in this story. Thanks, Verizon Wireless.]
Did VZ act illegally? No. Although the regulatory story here is tricky, the bottom line is that although the FCC could require that short codes be offered on a nondiscriminatory basis, they don’t. The Communications Act says that commercial cellular providers have to act in a nondiscriminatory fashion to the extent they are providing “commercial mobile services.” But “commercial mobile telephone services” are defined as services that are interconnected with the traditional phone network - reachable via dialing a phone number. Arguably, the short codes that the carriers allow people to subscribe to (”send me weather reports daily for $2.99 a month!”) are not “phone number” services. They’re private numbers controlled by the carrier. So they’re not covered by this nondiscrimination mandate. And VZ and all the rest of the carriers have many rules about who gets allocated a short code - I linked to those rules yesterday.
If VZ didn’t act illegally, why should we care? The shock of the public in learning that wireless carriers don’t act like common carriers should tell us something. From the consumer’s perspective, a communication (for data, to a person, using any device, whatever) is a communication. We assume that communications that feel just like phone calls are supposed to be provided on a common carriage, nondiscriminatory basis. We think of communications as a utility, like electricity or water, that is regulated by the government. It’s alarming to find out that these gigantic wireless carriers think of themselves as providing a private service within which they can discriminate for any reason or no reason at all.
But isn’t the wireless market competitive? Why would we need government intervention? Here’s where the long view is important. Wireless prices have been steadily rising since 1999, at the same time that the industry has been concentrating: twelve wireless carriers that were independent as of 1999 have combined (through merger, spinoff, or joint venture) into these four large wireless carriers - Verizon, AT&T, Sprint, and T-Mobile. Verizon and AT&T already control 51% of the wireless market in the US and are growing very quickly. Verizon ($22.6 billion operating cash flow) and AT&T ($17.8 billion operating cash flow) combined have 64% of the net additions to wireless subscriptions and 120 million subscribers. These are rich networks that are getting richer.
But the salient marketplace isn’t necessarily the market for wireless phone calls. These same two companies, Verizon and AT&T, control DSL internet access through regional monopolies across the country, and offer these wireless services as part of packages that tie together traditional phone services, “IPTV” access, and internet access. Everything is bundled. The elements of these bundles don’t compete against each other, really. And, as readers of this blog are tired of hearing, we don’t have real competition for highspeed internet access.
But this issue is all about wireless phone calls. So what? This issue matters because Verizon and AT&T see the wireless future - and we should too. These phone companies are making their key revenue in the wireless area, and they like the control that they have there. They understand that we’ll all be accessing all data/interaction/people using mobile handsets, and they’d like to wrest control of these communications from the PC arena - which is traditionally more open. They’d like to have mobile phone standards, which allow network providers to charge for every transaction we do using a handset, apply to all of our interactions. They’d like all applications to have to ask permission before they can be accessed - because that will allow the carrier to charge. So they’re hoping to move away from the traditional phone-common-carriage model to the “future” wireless-carrier-private-actor model as both a regulatory matter and as a business imperative, and they have the market power to do this.
That’s why this story matters. More than a billion handsets will be bought around the world next year. We’ll all be going online using mobile devices. If the wireless “we get to discriminate” model of regulation applies to handsets in the U.S., the mobile carriers will be able to act as gatekeepers in the marketplace of ideas, choosing winners and losers. The fact that a Verizon lawyer was empowered to say NARAL’s texts were “unsavory” as a matter of company policy
is hugely important - Verizon may have withdrawn this particular decision (smart move), but that’s just the tip of the iceberg.
We have to decide what model for communications regulation we want. Do we want the carriers to decide what we do online, or do we want the carriers just to be carriers? “Online,” “cell phone,” “telephone,” and “cable” all feel the same to the consumer. It’s all just data traveling fast. The regulatory reluctance of the FCC and the weirdly out-of-date structure of the Communications Act are allowing for differential treatment of the same kinds of transactions — the FCC knows this, and so it is strongly leaning in the deregulatory direction. “We’ll treat everything the same way and everything will be a private network. The market will be victorious!” But as a matter of social policy and our own future, we should sharply question that direction. We don’t have enough competition or enough enlightened, nondiscriminatory communications companies to allow us to be confident that the market will be able to do anything. We should move back towards common carriage for telecommunications - which will require that control over transport be separated from content.
So Verizon Wireless stores are everywhere, and incredibly convenient, because we’re addicted to these devices. They’ll be even more important in the years to come. But we shouldn’t be addicted to the regulatory model that currently governs our use of these devices. This will take leadership to change.
Why the digital transition
One of my students asked whether his television set, connected to a cable system but with no set-top box, would be able to receive digital television after February 17, 2009.
So I decided to try the experiment of being a consumer with this question. I was happy to see the NCTA has this site with information about how this will work. And this:
The good news for cable customers is
that the digital transition should be easy. Thanks to a compromise
adopted by the FCC in September 2007, cable companies will carry the
main digital signal of “must carry” commercial broadcast TV stations
and will duplicate that signal into analog format so that all channels
can be viewed on any older analog TV sets connected to cable.
That’s putting a bright spin on a mandate that cable systems do the work of transforming digital broadcast signals into analog so that a majority of
the people with cable subscriptions will be able to continue to watch all of their local
broadcast stations on analog TVs until at least 2012. (I’m not sure whether cable systems will be allowed to charge for offering the analog version of the digital signal.) Also, see the words “main digital signal”? The FCC agreed to allow cable operators to remove sub-channels from the
broadcasters’ digital signal so as to allow as much compression of the signal as possible.
But if you’re not connected to cable this could be tricky. The first converter box has been approved [site requires free registration], but it costs $69.99. Lots of hearings are scheduled to examine how on earth consumers will hear about the digital transition and be allocated vouchers for these boxes.
Why are we going through all of these conniptions with broadcast, when the broadcasters themselves realize that they won’t survive unless the cable systems carry their signals? When their sub-channels won’t be carried? And when broadcasting is becoming just a subset of online content anyway?
The answer is that over-the-air television is free (in that you don’t have to pay a subscription fee, even you do have to spend time watching the ads), and no one wants to be the politician who strands people without a television signal.
Soon we’ll all be in virtual Google-worlds…