Regulating peering

Thanks to Barack Obama, the net neutrality debate has now morphed into a bare-knuckles political brawl over the technical legal question of whether to reclassify broadband service as a Title II common carrier. Since Obama publicly endorsed reclassification this week, his political opponents on Capitol Hill, to say nothing of those who hope to succeed him, have sought to make opposition to reclassification into a rallying cry for the conservative faithful (“Obamacare for the Internet!”).

Somewhat lost in the hubbub over reclassification, however, was another significant recommendation by the president that, from the perspective of most internet users, might actually have greater impact.

“The connection between consumers and ISPs — the so-called ‘last mile’ — is not the only place some sites might get special treatment,” the president said. “So, I am also asking the FCC to make full use of the transparency authorities the court recently upheld, and if necessary to apply net neutrality rules to points of interconnection between the ISP and the rest of the Internet.”

That had to be music to Reed Hastings’ ears. The Netflix CEO has been clamoring for months for the FCC to include regulation of interconnection fees in its net neutrality rulemaking proceeding, despite agency chairman Tom Wheeler’s clear statements that the FCC would address peering separately, if at all.

In fact, Wheeler himself lately seemed to be moving in the direction of regulating interconnection agreements. His hybrid proposal, floated last month, which would reclassify ISP’s “wholesale” dealings with third-party edge providers as Title II services while leaving their consumer-facing “retail” operations as more lightly regulated Title I services, seemed to be as much about establishing a predicate for regulating interconnection — if not immediately than in the future — as it did implementing net neutrality rules.

One of the most puzzling aspects of the peering disputes that have arisen — principally between Netflix and a handful of the largest ISPs — is how little money appears to be involved. As Streaming Media’s Dan Rayburn has argued, in fact, it’s quite possible Netflix is actually saving money by paying to connect directly with Comcast, Verizon and AT&T compared to what it was spending to have Level 3 and Cogent handle delivery.

In a fascinating panel discussion on interconnection hosted earlier this month by Comptel (h/t Rich Greenfield) AT&T’s Hank Hultquist acknowledged that the revenue AT&T receives from paid peering, including with Netflix, is too small to be material.

Asked by Google Fiber exec Milo Medin why AT&T charges content providers for an arrangement that actually reduces costs for the network, Hultquist was admirably frank. “If you were to look at the revenue involved here to us, it makes me want to ask the same question, it’s not a lot of revenue,” he said.

In his next breath, however, Hultquist gets to what I think is the real rub, although he may not have realized it. “The view at least that I hear from our network people is they do not want to go to — they especially do not want to be forced into a model that when they increase capacity wherever they may increase it in the network, the only place that they can go and get recovery from that is directly from the consumer,” he said. “And, as was mentioned before, everyone recognizes that the consumer is paying in the long run.”

AT&T’s underlying fear in other words, however obliquely expressed by Hultquist, is that over time, as the internet increasingly becomes the primary pipeline for TV and other video programming, content providers will gain enough leverage over network operators to impose unrecoverable costs on them.

Nearly all major ISPs are also in the pay-TV business in one way or another. The cable ISPs started in the pay-TV business. They have all seen first hand how content implacably accumulates leverage over distribution and have felt the impact that ever-rising programming costs — which can only be recovered from the consumer — can have on their margins. As the pay-TV business begins to migrate to the internet, ISPs are rightly of recreating pay-TV economics on a platform that, up to now, has been free of programming costs.

The purpose in charging large video providers like Netflix for connection is not to raise significant revenue but to try to stave off pay-TV economics on the internet by establishing a regime in which content pays for distribution, instead of the other way around.

With Obama putting a very big thumb on the scales in favor of regulating peering agreements (and Wheeler clearly leaning in that direction) the risks for ISPs are growing. Banning paid peering might not have a major short-term effect on their bottom lines. But if it leaves the door open for internet-delivered video to evolve as most other video markets have, the costs of adding capacity to carry the increased video load could eventually seem trivial compared to the cost of carrying the programming.