The FCC (News - Alert) recently released an order addressing a number of long-running dockets, the most prominent one establishing a new Connect America Fund as a way to add explicit “broadband” support to the Universal Service Fund. It also reforms intercarrier compensation (ICC), the astonishingly complex set of rules by which telephone companies pay each other for carrying part of a call. Both of those would be complex enough under any circumstance.
But in releasing a 751-page document, this veritable “War and Peace” addresses far more than the original scope of the proceedings. Indeed many of the most contentious, difficult portions of the original dockets are still not settled. Setting aside the happy talk of “broadband for all”, the Commission has also adopted a Further Notice of Proposed Rulemaking (FNPRM) which does more than address the many remaining issues from the original dockets. It actually opens the possibility of directly regulating the Internet itself, making backbone peering and ISP-to-ISP connections subject to direct regulation similar to how telephone carriers interconnect.
The Internet is not just a packetized PSTN
The beauty of the Internet is that it developed in the US as a free market service that took advantage of what was then a regulated utility, telecommunications service, for its basic connectivity. The telephone companies never liked the Internet idea, which flouted their devotion to distance-based charges. The incumbent telephone companies had nearly-absolute protected monopolies on local service until 1996, and still have substantial market power. No single ISP, however, ever had such power over the Internet market, so a very different structure, one based on often-worshipped, rarely-seen free market principles, developed rapidly in the early 1990s.
ISPs have never been treated as regulated common carriers. ISPs operate at a higher layer than telecom carriers. They are its content, not the carriage itself; that’s why it’s called “information service” instead of “telecommunications service”. There is no license required to start an ISP in the US, nor is there a bright line between who’s an ISP and who’s a customer. (How, for instance, would you classify universities?) Most importantly, it doesn’t matter. Because it has been an unregulated market, just who is and isn’t an ISP is irrelevant. Likewise, who is or isn’t entitled to “peering” is a free market issue, a matter of price.
Contrast this market approach with the PSTN, with its rigid distinctions between IXC, LEC, and subscriber. In the PSTN (public switched telephone network) world, wholesale rates – ICC – are regulated. This gives the little guys a fighting chance against the former monopolies, and also provides rural carriers a bit of an implicit subsidy, to cover their higher unit costs. Indeed the reason the ICC and USF questions are so closely linked is that USF was created to replace non-cost-based ICC subsidies to rural carriers, a process that it still under way. Under PSTN rules, strengthened by the recent Order (it now applies to interconnected VoIP providers), telephone service providers can’t choose to not deliver calls to a carrier whose termination rates are unpleasantly high. If your customer dials the number, it must be delivered.
The Internet isn’t like that; all interconnection is voluntary. Indeed if an ISP is accused of being spam-friendly, the unwritten but well-understood “Mutually Assured Destruction” policy says that other ISPs must block them. Not just the spam, but the whole ISP, as they are unwelcome on the Internet. This hasn’t prevented all spam, but has helped force it underground.
“VoIP” refers to a broad range of technology and services
VoIP confuses things. As I wrote in a recent article There’s No Such Thing as VoIP, Voice over IP is raw technology, not one specific service. It can be used in many different ways. Since regulation should be technology-neutral, it should look at what’s being done with VoIP, not at the “magic pixie dust” of having an IP header somewhere in the stream. Skype (News - Alert), Vonage (and its many “over the top” Interconnected VoIP competitors), VoIP centrex, VoIP IXC backbone trunks, and PacketCable all use some form of VoIP, but are functionally very different.
The Order, though, talks about most VoIP as if it were Vonage (News - Alert), which is a pretty small share of the total, or at least will be soon. Or should be. And it shows a remarkable ignorance of how real-world VoIP actually works. The Order never mentions MPLS, which most enterprise VoIP networks depend on to provide the necessary Quality of Service and isolation from the Internet. It never mentions Session Border Controllers (SBCs), the application-relaying (call by call) systems needed to pass VoIP calls between otherwise firewalled IP networks.
It falsely assumes that most VoIP actually traverses the Internet backbone, when in practice only low-tier services like Vonage do. (So does Skype, but since it doesn’t touch the PSTN or use telephone numbers, it’s still outside of the FCC’s jurisdiction.) VoIP is a tiny share of ISP peering traffic; likewise, ISP-peered traffic is a tiny share of VoIP. Even Vonage-type services seek to avoid peering as much as possible, as peering points are a potential source of packet loss.
What really had been on the table was a perfectly reasonable request from a number of carriers, that they be allowed to interconnect with regulated ILECs using VoIP technology. They now must use TDM trunks in conjunction with Signaling System 7, but would prefer to use SIP. Modern telephone switching equipment supports both TDM and IP ports; even ILECs are beginning to replace their 1980s-vintage TDM-only gear. This IP interconnection would still be using dedicated voice trunks; all that would change are the multiplexing (TDM to IP) and signaling (SS7 to SIP) protocols.
Regulating the Internet backbone
So what does the FCC do in the FNPRM? It formally proposes moving the Internet backbone under PSTN regulation, under extremely questionable authority. It begins by falsely assuming that “IP-to-IP interconnection” of telephone calls necessarily crosses the Internet. It goes on from there:
For example, all carriers [note that ISPs are not carriers; this is a reference to the PSTN] are subject to a general duty to interconnect directly or indirectly, with LECs also subject to certain rate regulations, and incumbent LECs subject to a more detailed framework. In other contexts—notably, interconnection among Internet backbone providers—the Commission historically has chosen not to “monitor or exercise authority over” such interconnection on the grounds “that premature regulation ‘might impose structural impediments to the natural evolution and growth process which has made the Internet so successful.’”
So they are revising history to say that their non-regulation of the Internet was merely a “choice” because it would have been “premature”. A bit later they go on [emphasis added]:
We stated in the Order that we expect carriers to negotiate in good faith in response to requests for IP-to-IP interconnection for the exchange of voice traffic. But, we note that various types of services can be transmitted in IP format, and commenters recognize that many pairs of providers are exchanging both VoIP traffic and other IP traffic with each other. [Note: The “providers” here encompass both PSTN carriers and ISPs; these functions are usually performed separately. They are conflating the two.]
Further, different commenters appear to envision IP-to-IP interconnection policy frameworks encompassing different categories of services provided using IP transmission. We seek comment on those issues below, along with any other recommendations commenters have for defining the scope of an IP-to-IP interconnection policy framework in this context. For any proposed scope of IP-to-IP interconnection, we also seek comment on whether it is necessary, or appropriate, to address classification issues associated with particular IP services.
That they are unaware of the way telephone carriers exchange VoIP traffic is made even clearer in this question, which questions the feasibility of what we know to be the norm:
Does it make sense as a policy matter to adopt an IP-to-IP interconnection framework focused specifically on voice service, and how would such an approach be implemented? For example, would this approach have the result of compelling providers to exchange VoIP traffic under a different technological or legal arrangement from what those providers use to exchange other IP traffic?
Because they “seek comment”, this is formal notice that such rules are under consideration. The Internet itself is no longer sacrosanct. Earlier this year they ordered regulating the content that ISPs must provide on grounds of “network neutrality”, again conflating content and carriage. (I doubt that will survive court challenge.) The FCC has sniffed so much IP magic pixie dust that they’ve lost sight of the difference between the PSTN and the Internet, assuming that if they both use a common protocol header, they must somehow have merged into one blob.
This thinking shows up in a number of other places, including an assumption that when carriers transition from TDM to VoIP transmission of telephone calls, their points of interconnection somehow must move from an existing Point of Interconnection within each LATA to an ISP carrier hotel or NAP somewhere else. Somehow it’s assumed that IP headers cannot flow through a building that also houses TDM traffic.
Or just deregulate and throw PSTN competition under the bus?
One should not get the impression that the sole proposal on the table is to regulate the Internet backbone. Given that the FCC cannot tell the difference between PSTN and Internet any more, they are also looking at deregulating the PSTN as if it were the Internet. The new Order explicitly sets new ICC rates as a “default”, allowing privately negotiated deals. But it raises the question – AT&T (News - Alert) has been pushing for this for a while now – of fully deregulating PSTN interconnection.
This would come about by declaring that all IP-IP interconnection is deregulated. The Internet grew up with no monopoly using IP, and once the PSTN uses IP, apparently the magic pixie dust is assumed to do away with the consequences of the ILECs’ monopoly power. Then to finish the job, it would mandate rapidly moving all PSTN interconnection to IP, so that the transition to unregulated IP would happen even before the new regulated ICC rates take effect. Those rates would thus be moot.
An unregulated PSTN may sound like a good idea, but it really means that the big carriers can simply force the little ones out of existence by refusing to interconnect with them, or charging exorbitant rates for their connections. They’ve already shown how they’d behave given the chance: ILEC provision of wholesale DSL has already been deregulated. While “commercial agreements” for wholesale DSL still exist, the wholesale price is usually than the ILECs’ own retail price. Hence Verizon (News - Alert) is down to around 75,000 third-party ISP DSL lines nationwide.
Deregulation does not mean competition; it means that monopolies are more powerful. Europe, in contrast, has achieved true retail competition, and much lower rates than the US, by reregulation, in which wholesale services (primarily unbundled loops and interconnection) are the focus of regulatory oversight.
A few answers; a myriad of questions
So what did get settled in the FCC’s lengthy order? To be sure, a number of changes were made to both Universal Service and ICC. Not without controversy – appeals have been filed in most of the federal circuit Courts of Appeal, which will keep the lawyers happy for a few more years. But the general outline is clear. I did a rather extensive write-up, in 18 pages, here, for those who want more detail.
The High Cost Fund, which subsidizes rural carriers, transitions into the Connect America Fund. The old fund subsidized rural carriers to provide telephone service, allowing them to do so by building costly broadband-capable plant such as fiber to the ranch. The new approach subsidizes them to build costly broadband networks, so long as they also provide telephone service. See the big difference? It does away with funding for competitive carriers, though. If the incumbent LEC doesn’t accept the FCC’s offer, which will be based on a to-be-built cost model, others can bid to take over the turf.
There’s a new $4.5B cap on all high cost funding, though the USF tax rate rises in 1Q2012 to an unheard-of 17.9 percent. This may be approaching death-spiral levels, so the cap may be too little too late. Separate mobility funding is also provided to subsidize a single mobile carrier in currently-unserved areas. The Schools and Library Fund and Rural Health Care Fund are not changed.
The good news for competitive providers, such as WISPs, is that USF subsidies will not be provided to areas that already have “unsubsidized competition”, as of certain still-undetermined milestone dates in 2012. However, this has to be both “broadband” and voice service, so WISPs may need to quickly team up with VoIP operators or CLECs to get telephone numbers in their service areas. WISPs and satellite operators may also bid for pieces of the $100M “remote areas fund”, intended to pick up areas so costly to serve that ILECs don’t want them, even with allowable subsidies.
Intercarrier compensation is expected to eventually, over 6-9 years, move to a system where call termination charges are zeroed out (“bill and keep”). However, intercarrier compensation consists of both transport and termination charges, and the transport charges and rules are still largely left to the FNPRM. The sticky issues of “Virtual NXX” and Feature Group A (local lines deemed to be switched access) are not even mentioned.
In the meantime, the FCC’s baroque system of call classification is strengthened, with new rules to prevent “phantom calls”, where the originating location of intrastate toll calls is masked in order to get lower interstate rates. VoIP operators are expected to redesign their networks to provide originating caller location, although intrastate rates are due to be reduced to interstate parity by mid-2013, and VoIP will be exempt from intrastate access in the meantime.
Interconnected VoIP is deemed subject to interstate access charges (even for intrastate non-local calls) on a going-forward basis (no true-up for past calls, so Level 3 does not owe AT&T and Verizon billions of dollars in “agree to disagree” accruals). So the Interconnected VoIP wholesale price advantage (avoiding all access charges) is reduced, and goes away within two years. New rules also attack so-called “access stimulation”, by which free conference call services are funded, but they are not explicitly banned. Carriers who are found to be supporting them will however have their access charges reduced to that of the lowest-cost ILEC in the state; today those services are largely funded by sharing call termination charges with high-cost rural carriers.
All told, the FCC’s Order and FNPRM is Tolstoy-like in size and scope, but in terms of vision, it’s just more Beltway deal-making combined with dangerous over-reach. ISPs and others should take advantage of the opportunity to file Comments by the February 24 deadline, or Jan. 18 for USF issues. (I plan to help my clients do this.) The FCC needs educating; this could be the last time, before it’s too late.
Fred Goldstein, principal of Ionary Consulting, writes the Telecom Policy column for TMCnet. To read more of Fred�s articles, please visit his columnist page.
Edited by Jennifer Russell