“Who” should be covered by common carrier regulation relating to network interconnection has gotten murkier in the internet era, as have many other older concepts. We used to leave “data services” largely unregulated. Telcos were highly regulated, though less so these days than in the past.
Debates about “sending party pays” policies for interconnection are an example. In the past, only retail-facing telcos were subject to clear interconnection obligations, Internet domains interconnect on a voluntary basis.
But “sending party pays” rules extend those obligations to new parties: a few hyperscale content or app providers. That moves beyond public common carrier interconnection and towards rules for internet domains with no obligations to serve the public.
Such arguments also implicitly raise issues about which regulatory regime ought to hold: common carrier or data networks; internet or "telecom.
All arguments about universal service and now access network infrastructure upgrades necessarily entail rules about who should pay. To an extent, the answer could be “shareholders.” More often the answer is “customers” but sometimes the answer is “business partners.”
The point is that universal service or network upgrades typically entail some contribution by all customers of communication networks; payments by service providers (whether they pay or simply collect is an issue), financial support from all taxpayers or shareholders. In some cases, debt holders can wind up paying, especially if a firm goes into voluntary bankruptcy.
Basically, the argument that a handful of hyperscale app providers should make payments to access providers is of the “make business partners pay” argument, even if, in practice, there are no formal business relationships between app providers and ISPs.
ISPs do have such direct relationships with other internet domains and connectivity providers. Adding a few hyperscale app providers to the interconnection framework essentially treats those hyperscalers as though they were “carriers.” Historically, such agreements were between “telcos” and other public communications service providers.
Perhaps the difference now is that both ISPs and content domains now interconnect--directly or indirectly--to the internet fabric. So one way of framing the “sending party pays” discussion about internet traffic is to view the issue as a new form of debate over interconnection.
The idea is not unprecedented. Internet domains have used both settlement-free peering and transit fees when setting up business relationships. Peering is easy to justify when traffic flows are roughly equivalent.
It is harder to accomplish when traffic flows are unequal. And that is among the problems with proposals to charge a few hyperscale app providers for unequal traffic exchange volumes.
The bigger question is whether rules for common carriers should be extended to data service providers. In the past, the answer has unequivocably been "no." But that seems to be changing.
No comments:
Post a Comment