Kevin Werbach’s “Only Connect” got quite a bit of attention in the blogosphere when it was unveiled, including a post here on TLF. The attention was well deserved. The paper does an excellent job of explaining what’s at stake in the network neutrality debate and elucidating the positions staked out by each side. His discussions of the complexities of discrimination, access tiering, quality-of-service, etc in sections III(B) and III(C) are especially well done. He seems more keenly attuned than most scholars to the challenges that a regulator tasked with enforcing a non-discrimination rule would face.
With that said, I think the paper suffered from a fundamental conceptual weakness that left me unpersuaded by his ultimate thesis: I wasn’t ultimately convinced that interconnection and non-discrimination are separate and distinct regulatory issues. To the contrary, I think the two are often intimately connected. An effective interconnection mandate almost always depends on ensuring that the terms of interconnection are non-discriminatory. If network owner A is forced to interconnect with network owner B against its will, there are a variety of ways A can retaliate by charging B unreasonable prices, dropping B’s packets, dragging its feet on installing B’s equipment, etc. In practice, a practical interconnection mandate will invariably require some network-neutrality-like regulations to make it effective. The converse is equally true: a legal rule mandating non-discriminatory routing policies is likely to require some regulation of interconnection terms in order to ensure that the regulated carrier doesn’t discriminate through the back door by only offering low-quality links to those carriers against whom it wishes to discriminate.
I think this point is made most forcefully (albeit unintentionally) in his final section. It’s much shorter than the other sections and a significant portion of the section is devoted to the Atkinson-Barnekov proposal I discussed in my last post. But anti-discrimination would be at the heart of any practical effort to implement Atkinson and Barnekov’s proposals. In order to make the FCC-determined prices stick, the FCC would have to monitor the behavior of the larger carriers to ensure that they did not use covert methods to advantage the traffic of their own customers over those of their rivals. Otherwise, the threat of discriminatory routing policies would always give the incumbents the leverage they needed to extract larger payments than they would receive under the FCC’s formula.
Indeed, discriminatory policies were a big part of the reason that the FCC’s telco unbundling experiment failed. The problem in this case wasn’t with discriminatory routing policies so much as the discriminatory foot-dragging when competitors needed access to their facilities to install equipment. The Baby Bells successful bled the CLECs to death by simply making it a nightmare to work with the Bells’ facilities. The CLECs complained to the FCC, but the FCC didn’t move quickly or decisively enough to save the CLECs.
Anyway, my point isn’t to re-hash that debate, but to point out that an interconnection mandated necessarily implies a nondiscrimination mandate, and vice versa. When you force two companies to work together against the will of one of them, you necessarily have to monitor the unwilling company’s actions to ensure they’re not trying to sabotage the coerced relationship.
To put this in slightly more concrete terms, it’s not at all clear to me how mandating interconnection on the ‘net would create any fewer headaches for regulators than mandating non-discrimination. Mandating interconnection would necessarily mean setting rates and monitoring routing policies to ensure they were treating their voluntary and coerced partners equally. Moreover, as Jerry pointed out, the threats of non-interconnection seem even more remote than the treat of discrimination. Internet service that won’t reach all of the Internet is significantly less valuable than Internet service that allows you to reach everything, so I don’t think we should worry that we’ll wake up some morning and discover that our ISP can no longer reach Google or Wikipedia.
The Technology Liberation Front is the tech policy blog dedicated to keeping politicians' hands off the 'net and everything else related to technology.