This is cross-posted from Harold Feld's personal blog on WetMachine.com.
As reported by Brian Fung in the Washington Post and others, a company called Commercial Network Services (CNS) has filed the first network neutrality complaint under the FCC’s new rules — which went into effect June 12 after the D.C. Circuit denied a stay request. You can read the complaint here. While I probably should not prejudge things, I expect the FCC to deny the complaint for the excellent reason that — accepting all the facts alleged as true — Time Warner Cable did absolutely nothing wrong.
I elaborate on what CNS gets wrong, why this differs from other high-profile disputes like Cogent and Level 3, and why such an illustration is good for the FCC’s rules as a whole, below…
Basically, CNS wants settlement free peering (aka free interconnection) with Time Warner Cable (TWC). According to CNS’ complaint, they have exchange points where TWC also has a point of presence. They want TWC to exchange traffic with them for free, and they allege it is unjust and unreasonable for TWC to offer them the standard deal for entities of their size, i.e. a paid transit agreement.
Entities like CNS haven’t gotten free peering as a matter of course with major carriers since the first great upheaval in the peering market in the mid/late-1990s. CNS does not provide any evidence that it has been treated any differently from any other comparable entity of its size and function. TWC considered their request for settlement free peering under its standard procedures and found that CNS did not qualify.
CNS recognizes that it wants a change in the traditional way TWC has done business, but maintains it is unreasonable under Section 201 and the FCC’s rules for TWC to refuse free interconnection when CNS maintains a point-of-presence within arms reach TWC. CNS sums up its argument as follows:
“It is not possible for an industry, only in existence for 2 weeks, to have ‘well-established industry standards.’ TWC fails to understand they are no longer an ISP — they are a ‘BIAS’ with different obligations to the BIAS consumer.”
Unfortunately for CNS, this pretty much misconceives what the FCC intended to do by reclassifying broadband as a Title II service. The objective of the FCC was not (and the effect of the reclassification is not) to upend the existing business arrangements of ISPs for interconnection and impose a new regime of settlement free peering that hasn’t existed for 20 years. The objective of the FCC in reclassifying Title II (in addition to recognizing the reality that the way people use broadband is reflective of a Title II service), was to prevent a handful of giant ISPs from unilaterally dictating the business model for the entire broadband ecosystem to suit themselves.
The Difference Between This Case and Cogent/L3/Netflix.
Those not familiar with how the transport of bits across the “cloud” from wherever they originate to wherever they end up may not understand why this case is so much the polar opposite (at least on the pleadings) of the more high-profile cases involving Cogent and Level 3 (and, by extension, Netflix). Rather than an elaborate economic treatise, let me try to sum up the differences.
The first difference, of course, is we never really got any resolution on who was right in the disputes between the big transit ISPs (the companies that carry huge amounts of data to other networks) and the last-mile broadband access providers. But Cogent and Level 3 alleged a very different case than that alleged by CNS, and the differences are critical.
The gravaman of Cogent’s and Level 3’s complaints was not that the big broadband providers maintained their traditional policies, but that they dramatically changed them in response to taking on a single, large customer — Netflix. Cogent and Level 3 argued that the largest broadband access providers, i.e., those with market power, refused to upgrade the existing interconnection points in the manner they had previously been doing for years to accommodate settlement-free peering (for which Cogent and Level 3 had previously qualified). Now, because they carried Netflix (a major rival with these same carriers’ traditional video services), these carriers demanded lots more money. While no one can say publicly how much more they demanded (it being proprietary and all), Cogent and Level 3 allege that the demands for payments to upgrade the interconnection point and receive Netflix’s traffic well exceeded any reasonable cost and demonstrate an abuse of market power. Because Netflix’s video traffic is particularly vulnerable to the delay caused by congestion, this refusal to upgrade the interconnection point worked very effectively.
As an added blow, allege Cogent and Level 3, once the largest broadband carriers discovered this vulnerability, they wanted to get higher payment (“tolls” as Cogent and L3 describe them) for all traffic — even after Netflix cut a separate deal for direct interconnection and took its high-volume latency-sensitive traffic off their systems.
The broadband carriers deny that they were the ones changing the accepted standard for exchange of Internet traffic. They argued that it was Netflix that changed everything by taking their traffic that had traditionally gone through a content distribution network (CDN) and moving it to a peering network. In other words, Netflix (and its transit providers) were the ones changing how things worked and therefore the broadband access providers were totally reasonable in demanding they renegotiate terms. Furthermore, they argued that the economics of traffic exchange have constantly evolved over the course of Internet history and there was no proof that the access providers were doing anything nefarious by responding to a changing market.
I’m not going to revisit who I think has the better argument. Anyone who has actually read anything I (or my employer Public Knowledge) has ever written on the subject will show that we don’t have an opinion on any specific interconnection/transit dispute. What we do say, and continue to say, is that we can’t allow Internet traffic to millions of people get disrupted because of a commercial dispute between companies. We also say that, if it turns out that Cogent and Level 3 are right and this is about market power, then the FCC ought to do something about it. At the same time, carriers do need flexibility to adapt to the changing nature of Internet traffic.
Hence a rule that allows the FCC to act as a referee and make sure that all the companies play by the rules. It seems to have worked out reasonably well for the pending disputes between Comcast and Level 3 and the largest providers. Companies will still negotiate, but consumers will not get crushed in the crossfire when a “peering dispute” happens and traffic gets congested. The idea of such a rule is not (contrary to the non-stop allegations of opponents) to regulate peering arrangements. The idea is to provide a method of resolving interconnection disputes that does not congest traffic for lots of people totally helpless in the dispute and utterly dependent on the Internet for the smooth functioning of their businesses and daily lives. A referee calls fouls and makes sure games are played fairly, but they don’t fix the scores of games or tell players what plays to run. Even those who are angry about a particular call will generally agree that, for a game to work, professional teams should not be allowed to call their own penalties.
CNS Alleges None of These Things.
Which brings me back to the CNS complaint. CNS doesn’t allege that TWC suddenly changed their policy or that a facially neutral policy discriminates against a particular application. They basically say, “Hey! I want free peering because that would be good for TWC’s subscribers.” We had that argument in the mid/late 1990’s when differentiation in the network moved us from free peering for everyone that could plausibly lay claim to being a “backbone” provider to formal peering policies that created the modern peering and transit market.