Happy updates (July 25)
1) Ms Marsha. One of the best illustrations of the political clout wielded by the incumbent US broadband providers lies in their ability to kill any attempt at the creation of muni broadband networks. Twenty states have passed laws banning public-sector broadband alternatives, encouraged by the industry lobbies and those who might be harmed by competition, like poor, struggling Comcast. As I note in the 2nd para below, Marsha Blackburn of Tennessee is one of the great congressional champions of this free-market exercise. She is more determined than ever to ensure nobody in her own state or any state ever gets better, more affordable service. Such is her reputation that in a comment posted to a story in Ars yesterday, a reader added this apology:
A bit off topic: As a Tennessee state resident, I’d like to personally apologize to the rest of the country, hell the world maybe for that matter, for Marsha Blackburn. Another fine example of what you can buy of [sic] you are a corporation with enough money to line the right pockets.
On the upside, Brodkin’s story is entitled “FCC gets its chance to overturn state limits on broadband competition” – reporting on a petition to the FCC from a community-owned electric utility to overturn the state law barring it from providing fiber-based Internet access – in Chattanooga (Tennessee!).
2) Transparency enforcement. And in other good news related to policing the ISPs, the FCC on Wednesday issued an Enforcement Advisory that holds two surprises. The FCC will fine any broadband provider that intentionally misleads its customers; and the rule behind this notice is one of the few survivors of the DC Circuit appeal by Verizon that vacated most of the 2010 Open Internet Order. The advisory begins thusly:
Providers of broadband Internet access services must disclose accurate information about their service offerings and make this information accessible to the public. This requirement, known as the Open Internet Transparency Rule, has been in full force and effect since 2011. The Transparency Rule ensures that consumers have access to information that helps them make informed choices about the broadband Internet access services they buy, so that consumers are not misled or surprised by the quality or cost of the services they actually receive.
I’ve uploaded the pdf here.
How about we blame the Republicans instead? (cont’d)
As I was saying in the previous post, FCC Chairman Tom Wheeler should not be getting the rap for the open Internet NPRM issued in May. Especially not for his alleged determination to push through paid prioritization, since a) Wheeler has plainly disavowed it, and b) the Notice spends far more time analyzing other issues I see as a greater threat, like the utter lack of transparency or accountability in what broadband providers sell to the public. And picking this fight with paid prioritization is going to do nothing for the pro-Internet movement in the US or elsewhere (NPRM pdf here).
Instead, say I, opponents of the big-business, anti-consumer school of thought should pay more attention to what the black hats are saying – who, for convenience, we’ll call “Republicans.” These guys include a broad swath of personalities, from the FCC’s two Republican Commissioners, to the incumbents like Verizon that want no regulation and lots of “flexibility,” to my favorite right-wing curmudgeon, Marsha Blackburn, the Tennessee congresswoman who has succeded in getting a bill passed to call a halt to all that outrageous muni broadband that competes with Comcast, TWC et al. They have main three arguments, all of them pure sophistry, but great headline-grabbers.
One, we love the open Internet and consumer protection as much as the next guy. Two, the Internet has achieved all that can be hoped for, thanks to free enterprise and no government meddling. Three, all regulation is bad, especially regulation in aid of net neutrality, and even more especially Title II, which is for trains. (Title II is that part of the US Communications Act containing provisions to regulate common carriers. It has become the net neutrality battleground, since if applied to ISPs it would require them to behave much more like phone companies and much less like unregulated “information services.”)
We’re all right, Jack
Some of the Republican claims amount to empty sloganeering, like the clever idea of always adding “heavy-handed” or “job-killing” to any mention of regulation. A more contentious variation on this theme is the notion that net neutrality is a “solution in search of a problem.” Some have taken the trash-talk farther – e.g. pundits writing under the American Enterprise Institute (AEI) seal of approval. Richard Bennett guest-posted on GigaOM last week and made a couple of assertions that give a tired brain pause.
In one, Bennett accuses Tim Wu of preferring “monopoly-style regulation for increasingly competitive broadband networks.” In this disingenuous double-whammy, Bennett manages to imply that regulation intended for broadband would be fashioned after regulation intended to discipline a monopoly like the old AT&T; he then suggests without offering any evidence that retail broadband in the US is getting more competitive.
Leaving aside whether he’s referring to Title II as being fit for monopolies but not other highly concentrated markets, three factors are making the US market less not more competitive: increasing consolidation of ownership, increasing vertical integration, and the decreasing ability of DSL and hydrid DSL to compete in the ultra-fast retail market segment against DOCSIS 3.x.
“If eBay had been subject to
telephone-style regulation phone company decision-making, we’d still be waiting for the first online auction.” — to paraphrase Bennett
Bennett’s other sleight-of-hand is to blame regulation for what the incumbent phone companies did enthusiastically all by themselves: make sure no one innovated on their networks:
The internet has succeeded because it’s flexible and open to new applications, new users, and new networking technologies, not because it’s rigidly confined to a regulatory box with an established history of stifling innovation.
Seriously? It was AT&T that tried to keep any device, even as benign as the Hush A Phone and Carterfone, off their precious network in a blow to both innovation and competition. It was the phone companies that fought TCP/IP tooth and nail, and whose engineers treated it with contempt – and who promoted development of the OSI networking platform for 20 years, until it collapsed under the weight of phone company bureaucracy. It was the US phone companies that tried for years to keep the Internet from disrupting their business models or voice-optimized engineering.
And as they’ve all become vertically integrated, with conflicts of interest between their platform and content holdings, they’re still out to keep the Internet from being disruptive – entirely despite the regulators, not because of them.
Two examples of how US broadband is not working
Item #1: American end-users can be deprived of reasonable service for weeks or months by their ISP, without any recourse – including switching providers, a recourse that would be available in a competitive broadband market.
Take Verizon. Like Comcast, Verizon has been embroiled for months in a fight with Netflix over transmission quality on subscriber devices. Both have negotiated a paid interconnection agreement with Netflix – with one crucial difference. Comcast got its money and has apparently fixed the congestion problem; Verizon has not. In fact, Verizon has not only got stuck at the bottom of the Netflix ISP speed rankings. It has also been caught in what appears to be a stupid, stubborn refusal to upgrade the port cards at its points of interconnection with Netflix transit provider Level 3. Our first exhibit is the June Netflix ISP speed rankings:
Netflix: June ISP speed ratings for the USA
There is some dispute as to whether the Netflix ratings measure the capacity of local access networks, or customer demand. But that hardly matters here, since even if the top performers are given a boost by running on the Netflix Open Connect platform, Verizon is currently 16th – dead last in the USA, months after they got their money (and after their average speed dropped 2 months in a row).
In a widely reported post published a few days ago on the Level 3 blog, Mark Taylor (VP, Content and Media) wrote a scathing indictment of Verizon’s role in the interconnection imbroglio, which happens to be entirely consistent with the measures Netflix is reporting on its speed ratings. Taylor says Verizon blew its own cover when its attempt at deception backfired – blaming Netflix for the end-user quality problem:
[Verizon VP David Young] has clearly admitted that Verizon is deliberately constraining capacity from network providers like Level 3 who were chosen by Netflix to deliver video content requested by Verizon’s own paying broadband consumers.
The crux of the Level 3 claim is that Verizon is using ports at certain POPs – like in Los Angeles – that need to be upgraded to handle all the traffic Level 3 is handing off on behalf of its customer Netflix. To keep traffic flowing smoothly, ISPs like Verizon and transit providers like Level 3 have to engineer their networks so that their average utilization level stays very low – way below 100%.
The big revelation here is not just that Verizon is using congested ports, ensuring millions of their customers get shitty service while blaming Netflix for the problem. It also turns out, according to Taylor, that upgrading would be a trivial and inexpensive proposition. Taylor adapts the very diagram Verizon posted to show the existing ports congested (at 100% of capacity), but with plenty of headroom left over to install several new 10 Gbps Ethernet ports:
Level 3 network diagram: interconnect topology with Verizon in L.A.
Those X’s on the Verizon border router are exactly where the ISP should be adding capacity:
So in fact, we could fix this congestion in about five minutes simply by connecting up more 10Gbps ports on those routers. Simple. Something we’ve been asking Verizon to do for many, many months, and something other providers regularly do in similar circumstances. But Verizon has refused. So Verizon, not Level 3 or Netflix, causes the congestion. Why is that? Maybe they can’t afford a new port card because they’ve run out – even though these cards are very cheap, just a few thousand dollars for each 10 Gbps card which could support 5,000 streams or more.
It gets worse. Congestion on the Verizon network takes place only at interconnection points where Netflix streams are handed off – nobody else’s traffic is affected in the same manner. Could this be a conflict of interest or anti-competitive behavior or a breach of promise on Verizon’s part?
Item #2: Hot off the press… The updated OECD broadband data released on Tuesday (July 22) show the US stuck in a mediocre, middle-of-the-pack position on the ranking for overall penetration per 100 inhabitants: i.e. 16th out of the 34 member countries. Even Canada is doing better, at #11.
The new penetration data (worksheets 1a through 1m) are all valid up to December 2013. As the table below shows, the US is just ahead of the OECD average for all platforms – 29.8 per 100 vs 27.0. Another figure that stands out is the number of lines running over fiberoptic: the US stands at 2.4 while the OECD average is 4.5, nearly twice as high.
The OECD data show clearly the American incumbents have nothing to crow about when compared to the other OECD members. It’s not enough for innovation to be alive and well online if 30% of the population aren’t on broadband at home (the Pew survey from August 2013 finds 70% of US adults have broadband at home: pdf).
And if fiber is the future, or at least a big part of it, then the US has a long way to go to become an even middling jurisdiction. Not to mention the awkward fact that Verizon is slowing down its FiOS growth, which may be peaking at about 6.2 million subscribers – while letting many of its paying subs suffer congestion no matter what platform they’re on.
Paid prioritization: standing with the Republican FCCers
Needless to say, the Republican FCC Commissioners, Michael O’Rielly and Ajit Pai, dissented vigorously from the majority in the May 15 Open Internet vote. Commissioner Pai’s concern is jurisdictional. His opinion is that banning paid prioritization has not a legal leg to stand on, a microcosm of the FCC’s alleged and larger problem of the statutory authority to make rules on the open Internet in the first place:
I see no legal path for the FCC to prohibit paid prioritization or the development of a two-sided market—which appears to be the sine qua non objection by many to the Chairman’s proposal. As the NPRM frankly acknowledges, section 706 of the Telecommunications Act “could not be used” for such a ban. And while the NPRM resists saying it outright, neither could Title II. After all, Title II only authorizes the FCC to prohibit “unjust or unreasonable discrimination” and both the Commission and the courts have consistently interpreted that provision to allow carriers to charge different prices for different services.
Commissioner O’Rielly is no less brash than Pai in denouncing their three Democratic colleagues for a set of proposals he describes variously as flawed, ill-advised, absurd, a regulatory boondoggle, and hopelessly vague and unclear (pp.98-99). But O’Reilly takes a more radical view of prioritization than Pai. Rather than disputing whether the FCC is empowered to bar prioritization, O’Rielly argues it has been happening for years and is a necessary part of managing IP networks, as “even ardent supporters of net neutrality recognize” (p.99). He notes the engineering truisms that voice must be prioritized over email and video over plain data. He’s even more vehement when he moves to the issue of paid prioritization:
The Notice is particularly skeptical of paid prioritization and contemplates banning some or all such arrangements outright. Yet companies that do business over the Internet, including some of the strongest supporters of net neutrality, routinely pay for a variety of services to ensure the best possible experience for their consumers. […] In short, fears that paid prioritization will automatically degrade service for other users, relegating them to a so-called “slow lane,” have been disproven by years of experience (p.99).
I was surprised to find myself agreeing – in part at least – with O’Rielly’s comments. I’m not convinced paid prioritization is likely to be the cause of slow-lane experiences for end-users, if only because lots of other, more plausible culprits are lying in wait – like Verizon’s screw-the-customer attitude, which didn’t need a fast lane to generate congestion, just dishonesty and incompetence.
Dan Rayburn blog: Direct interconnects between content owners and ISPs
Moreover, O’Rielly seems to be right about the degree to which the US networking market already depends on direct interconnections between large content owners and the large ISPs. The chart above, from a May 21 post on Dan Rayburn’s blog, shows just how pervasive direct interconnects have become in the US. Here’s what Rayburn, no fan of the Netflix camp, has to say about these connections:
What you can see on the chart is that all of the major content owners have negotiated direct business relationships with multiple ISPs for their CDN. There are various business and technical reasons why they pick one ISP over another, but there are a lot of these deals in the market and have been for many, many years. Some ISPs in the U.S. have dozens of paid interconnect deals in place. Multiply that by the number of ISPs in the U.S., and the number of paid interconnect deals is probably close to 100. So for Netflix’s CEO to call their paid interconnect deal with Comcast “unprecedented,” is intentionally misleading.
While my sympathies lie with the content guys and not the ISPs, I’m citing Rayburn (an expert on network streaming) because his perspective adds weight to my idea that paid prioritization is a red herring in the war over the open Internet. But that perspective leaves a big question unanswered:
When ISPs set up a paid peering arrangement with a major content provider, what if anything should be done to ensure that end-users benefit, by getting reliable service within the speed range they’re paying for?
While this media obsession with paid prioritization plays itself out, we can only hope the FCC will make headway on the many other issues that have come to light. Some progress should be registering, as long as the FCC staff are doing what Wheeler promised in his June 13 statement on broadband consumers and Internet congestion (pdf here):
Recently, at my direction, Commission staff has begun requesting information from ISPs and content providers. We have received the agreements between Comcast and Netflix and Verizon and Netflix. We are currently in the process of asking for others.