Comcast vs. Level 3 vs. Network Neutrality

My preceding article on the bafflegab that follows the network neutrality debate like a malevolent black cloud only briefly touched on the brewing dispute between Comcast and Level 3. Within hours it has blown up into a widely-covered issue that, in the spirit of the bafflegab I talked about, is being used by many to trumpet their own entrenched position in the internet food chain. I had intended today to build upon the theme I introduced in yesterday's post, but this dispute simply provides too good opportunity to highlight a few key items in these companies' war of words that fit well with the theme I introduced.

First off, I do not intend to do what others are already doing; there are a few (maybe more) excellent articles that dig deeper into what the dispute is really about, but that may get drowned out in the flood of media coverage. There is one in particular I want to recommend, which was published by Ars Technica, that is short, lucid and does a good job of covering off the deeper nuances of the dispute and also provides references to some more comprehensive background material. If you want to learn more rather than merely pick sides in the fight, go read it.

There is only one sentence that I want to mention here since it goes to the heart of just what network neutrality is all about:

The DC group Public Knowledge blasted Comcast's stance as a net neutrality violation.

I like this since it is such pure nonsense. The thing is, there is no commonly-accepted definition of network neutrality when it comes to the internet -- which includes access, transport and services -- and there is certainly no law on the books that anyone is violating. That statement is mere misdirection with the apparent intent to sway others towards Public Knowledge's position.

To understand this more deeply we need to look at just how the internet is assembled -- the Ars Technica article provides better references so look there if you want more than the following brief and somewhat superficial description. The internet was originally constructed of autonomous systems that interconnected by mutual agreement for the mutual benefit of their users, irrespective of each system's size, public or private or government, using data connections whose cost was either shared in some fashion or covered by one of the two parties. The telco monopolies, which were common carriers, that provided those long-haul and short-haul transport facilities had no interest whatsoever in what was being carried; they billed for the transport and never even saw what was inside those pipes. This is peering in its simplest form.

As the internet opened to the public, the structure had to evolve. First, ISPs came on the scene with their racks full of modem banks and oodles of telco business lines that users dialed to access the internet and local ISP services like email. Those phone lines were subscribed by public tariff and were subject to the telco's common carrier responsibilities; that is, the telco had no choice but to offer those business lines as long as their network equipment wasn't harmed. At first this was good business since it brought a lot of new revenue, both from the ISP lines and all the second lines that residential customers installed for their computers.

The ISPs negotiated peering agreements with other ISPs to mutually terminate traffic to their customers, which included both users and services, and to route traffic to other ISPs further removed, even across the globe. The industry moved toward more stratification between access ISPs that served users, backbone providers that only provided transport and routing between access ISPs and other backbone providers, and service providers that provided the content that users wanted to access.

Backbone providers found it easy to peer among themselves since their traffic flows tended to be similar in both directions, and not charging for that traffic made sense since the burden of accounting could be dispensed with and the monthly net tended towards zero. The traffic differential between access ISP and backbone providers increasingly became unbalanced since as users accessed services that utilized the increasing access bandwidth they did not transmit much data upstream. Besides, since the backbone providers could not cover their business costs by peering, they charged the access ISPs for their services; but not for the raw transport, which the ISP alone was responsible for by contracting with a telecommunication provider, which was usually a telco or a raw transport provider such as MCI.

Notice that in none of this did I mention regulation once. That's because there was none. The data transport facilities that everyone used came from common carriers of one sort or another, but while the common carriers were tightly regulated the users of their transport services were not. It's the same whenever you pick up the phone and call someone; the telco provides a tariffed service as a common carrier but you are not regulated. The common carrier does not and, importantly, must not concern itself with how you are using that tariffed service. If they do get involved then they are almost certainly breaking one or more conditions of their licenses or even the law.

Because unlike network neutrality, common carrier has a definition in law and is enforced through government licensing and oversight, complete with penalties for non-compliance. However there is a benefit to the common carriers because even if you use a common carrier service to commit a crime (such as making a drug deal or arranging a murder) the carrier is protected by law from any liability; you can't sue Bell Canada because someone used their services to commit a criminal or civil offense that causes you harm. Similar legal protections have been extended to  ISPs and related services in the US for nearly 15 years by means of Section 230 of the Communications Act, although they are not common carriers.

Getting back to the evolution of the internet, as the business potential grew large and broadband replaced dial-up, through a multi-year frenzy of mergers, acquisitions and emergence of new business models, we now have the carriers in an enviable dominant position where they have the power to dictate terms to others. They have a pretty solid lock on access and transport, wireline and wireless, where they are (as Broadband Reports puts it) the "troll-under-the-bridge. They almost always stand between end users and web-based services, and they are unencumbered by common carrier regulations: the regulations are not applied to corporate entities, but selectively to each line of business that provides those particular services. For example, Verizon is a common carrier but also an ISP. The same applies to Level 3 with their transport and telephony services distinct from their routing and CDN services.

This last point is where an important nuance comes up in the dispute between Comcast and Level 3. Level 3 has a financial advantage over pure CDNs like Akamai since they can "sell" transport and routing between business units at a discount to what they offer to other companies. This may have helped them win Netflix business from Akamai. From Comcast's perspective, they do have a legitimate issue with regard to peering with Level 3 since, due to their CDN business, we would expect the traffic imbalance to be greater than with other backbone providers and ISPs.

Now, Level 3 proposes to send traffic to Comcast at a 5:1 ratio over what Comcast sends to Level 3, so Comcast is proposing the same type of commercial solution endorsed by Level 3. Comcast is meeting with Level 3 later this week for that purpose. We are happy to maintain a balanced, no-cost traffic exchange with Level 3. However, when one provider exploits this type of relationship by pushing the burden of massive traffic growth onto the other provider and its customers, we believe this is not fair.

Of course there is already an imbalance that any pure access ISP like Comcast will see since the bulk of their customers are end users that primarily download from content served by other ISPs. There is nothing to stop Comcast from getting into the content hosting business, it just isn't what they've chosen to do. The traffic imbalance is not due to any nefarious action on Level 3's part, just a natural consequence of Comcast's and Level 3's respective business priorities. Recall that none of this is regulated; while each company does have some FCC regulated business (Comcast's cable TV business and Level 3's transport business) their internet and broadband businesses are pretty much unanswerable to government regulators. Each company chooses its business activities as it sees fit with regard to internet.

To be more blunt, this dispute, which has nothing to do with network neutrality and where peering is optional not mandated by law, is almost entirely a private commercial contract negotiation between two companies. Neither is above slinging around terms like network neutrality, monopoly and fairness if doing so wins them political and public allies to buttress their side of the negotiations. Here's part of what Level 3 has to say:

John Ryan, Assistant Chief Legal Officer of Level 3 Communications, Inc.: "...the fundamental issue is whether Comcast, as the largest cable company in the country with absolute control over access to its cable TV and broadband access subscribers, has the right to unilaterally set a 'price' for that access that effectively discriminates against competitors of Comcast’s cable and Xfinity content..."

Notice how Level 3 tries to drag in Comcast's regulated business activities by insinuating that there is "leakage" between their regulated and unregulated business units which, if true, would justifiably draw government investigation.

The stakes for Comcast are quite high in this game of brinkmanship. While they do have a dominant position as gatekeeper to a huge body of internet access customers, just as they do for TV content distribution, and they are making inroads against the telcos by winning away telephony business, they are in fact surprisingly vulnerable. First, there is the talk about "cord cutters", which are customers, especially 20- and 30-somethings, that are showing signs of abandoning cable services -- TV, broadband and telephony -- since prices keep rising, customer service is awful, and their services are increasingly redundant with wireless voice and data. In other words, why pay for both if you favour internet media over TV and one device, the smart phone, gives you everything you need everywhere you go? Comcast doesn't have a wireless business. Further, even at the high prices charged for tethering, it can be the superior and cheaper choice to connect a PC or netbook to the internet (when a larger screen makes sense) in comparison to paying two broadband bills.

Even worse for Comcast is if the FCC and Congress feels that companies like Comcast are becoming overly aggressive their broadband billing and content practices for a service that is now seen as a pretty essential utility. Political influence goes only so far, and that is when the voters start shouting for Comcast's blood and want the government to take action.

Once they complete the acquisition of NBC Universal, if they continue to jack up cable and broadband rates, offer poor service and use their growing control over content, both TV and internet-based, to their own advantage, they might find they've painted themselves into a corner they can't their way out of. Not only is it more likely that the FCC will pursue turning broadband into a common carrier service, they may get the Congressional support that they need and currently lack. Worse, it is even conceivable, if still very unlikely, that the FCC could invoke Title VI of the Communications Act  and mandate "must carry" for internet media content (Hulu, YouTube, Netflix, etc.) in a manner similar to cable channels if Comcast is too aggressive in prioritizing content of those willing to pay them a premium. Companies with a dominant or monopoly position providing an essential service do have to take care to not cross that invisible and shifting line in the sand that will signal an end to government indifference.

Ultimately it is the consumer that will pay for all of this since all business costs are passed along in one way or another to end user, and governments know this:

This is a problem the Congress and regulators cannot ignore. Just as in the recent retransmission fights in the pay TV world, these rumblings between giant companies leaves consumers in the lurch, even though they’ve actually paid for access to the Internet — that is, the whole Internet, not one approved by Comcast or some other company. The problem, of course, is lack of competition in the broadband markets.

Network neutrality is just a word (or two), and no matter how much it is used to misdirect and obfuscate and attack a company's competitors, the real metric is when the howls of the public become audible in Washington. This is about politics and voter dissatisfaction; network neutrality is a definitional sideshow that should not distract our attention from the real issues at stake in disputes such as that between Comcast and Level 3.