Economics and Technology, Inc. -- Views and News, June 2011

June 2011

Continue reading ETI Views and News at econtech.com, or download a printer-friendly version.

Understanding the economics of IP-based voice telephone services

A great deal of attention in recent years has been directed at the large-scale "cord cutting" by residential telephone consumers as they abandon their traditional wireline local phone service and become "wireless-only" households. The Center for Disease Control and Prevention (CDC) has been tracking this trend for a number of years, and in the most recent CDC study it estimated that some 26.6% of US households were relying upon wireless for their primary residential phone service. FCC data indicates that between June 2005 and June 2009, the number of (ILEC+CLEC) wireline residential exchange service connections had dropped from 111.7-million to 92.9-million.

Business use of wireless services has also grown rapidly but, unlike residential, business "cord-cutting" is far more limited. In June 2005, there were roughly 66-million (ILEC+CLEC) wireline business exchange service lines in place; by June 2009 that figure had decreased only slightly, to 63.7-million. But is this drop in business access line demand also due to migration to wireless, or is there something else going on here? Indeed there is.

Traditional circuit-switched business exchange service is being replaced by IP-based voice telephone offerings that "ride" on the customer's broadband connection, often the same or a similar one that is used to obtain Internet access. "SIP" (Session Initiation Protocol) is used to provide Voice over IP (VoIP) signaling, facilitating telephonic connections over IP-based networks (such as the public Internet) much as Signaling System 7 (SS7) supports an analogous set of functions in the circuit-switched Public Switched Telephone Network (PSTN). SIP supports the same types of calling features found in traditional telephone systems (plus some additional ones), but operates as a peer-to-peer protocol that can support non-voice communications applications as well. SIP-based services provide a functional alternative to traditional PBX trunks and Centrex lines, and enable business multiline users to "cut the cord" with respect to their telco-provided services just as residential consumers can abandon wireline in favor of wireless.

How SIP works

In the broader context of the Internet and IP-based services generally, SIP is simply a packet-based VoIP "application" that travels over the same broadband facilities as other IP-based apps. SIP can be used in place of traditional PBX trunks (so-called "SIP trunks") to provide dial-up access to an "IP PBX" physically maintained on the customer's premises. Providers of SIP trunking services, like providers of most other Internet-based applications, typically utilize broadband facilities obtained by the customer from the local phone company, cable company, or other source. SIP "trunks" are not discrete channels (as in the case of traditional Time-Division Multiplexing (TDM) trunks) but utilize the customer's broadband bandwidth as needed. SIP trunking services are typically priced in terms of the number of simultaneous voice calls that can be supported, as well as charges for usage and for direct-dial phone numbers.

The other main form of SIP services is analogous to Centrex. These "Hosted PBX" services utilize remotely located IP PBX platforms and support individual handsets connected (registered) to the remote platform over the customer's broadband connection. Larger organizations would typically utilize broadband services specifically dedicated to the SIP services; small and medium businesses may be able to integrate their SIP phone services with their access to the public Internet.

IP Telephony is often less expensive for the customer

From the business user's perspective, IP phone services typically cost much less than comparable circuit-switched PSTN services, for several reasons. First, the market for IP-based phone service is far more competitive that the market for circuit-switched services, which consists mainly of the incumbent local phone company and at most a handful of CLECs that are themselves dependent upon facilities obtained from the incumbent local phone company. Because the SIP provider does not need to construct or even lease its own broadband network, barriers to entry into the IP phone service business are minimal. Customers can easily switch providers in response to better pricing offers. SIP handsets, which cost about the same as traditional business multiline phones, are connected to the broadband service using the same ethernet wiring infrastructure that a firm uses for its LAN, obviating the need for separate telephone wiring. Also, because all of the addressing information is maintained within each individual handset, off-premises locations (e.g., branch offices, telecommuting employees) can be served from anywhere that a broadband Internet connection is available.

The second principal source of the cost differential between IP and PSTN telephone service may be more illusory than real. Where the various factors summarized above represent substantive operational efficiencies and the effects of competition, SIP services are also less costly to provide because of differential regulatory treatments afforded IP vs. PSTN telephony. PSTN services have long been subject to a complex array of cross-subsidies, surcharges, fees and other non-cost-related factors imposed by regulation to support one or more "public interest" goals. Long distance services were traditionally priced above cost to subsidize local service; business services were priced so as to provide a subsidy for residential service; urban services were priced so as to subsidize rural services, and premium or "optional" features were priced so as to subsidize "basic" dial-tone access. IP-based voice services have thus far escaped much of this treatment and, while the FCC has for a number of years sought to create technology-neutral parity among all services offering the same or equivalent functionalities, there is still a considerable gap. Moreover, since technology seems to move a lot faster than regulation, it is not clear that this gap could ever be fully eliminated.

Is it actually less expensive to provision IP-based telephony?

PSTN vs. IP operating parameters. The PSTN employs a circuit-switched architecture in which a physical or logical transmission path is established between caller and recipient that exists for the duration of the call. The various PSTN resources are thus dedicated to a specific call and are not available to support any other use until the call has been disconnected. It is for this reason that PSTN services are typically subject to some form of duration-based pricing. In the case of IP, traffic is measured in terms of bits transmitted rather than duration of the call itself. In its recent Universal Service/Intercarrier Compensation NPRM (Views and News,March 2011) the FCC suggested that for this reason VoIP traffic should be measured and priced on the basis of bits rather than time. While that approach might have some theoretical appeal, as it turns out, VoIP traffic actually embodies cost attributes that are not all that different from those that control PSTN operation, making duration-based measurements a reasonable – and far more understandable (from the consumer's point-of-view) – proxy.

While the customer operational efficiencies associated with IP telephony are real, it is far less clear that VoIP transmission is actually less costly to provide than legacy TDM transmission that is used in the PSTN. Both require periodic sampling of the analog voice signal for conversion into digital form. However, where TDM involves the assignment and dedication of a specific "time slot" in the synchronous TDM bitstream, VoIP traffic is transmitted asynchronously in packets. The voice conversation must still be continuously sampled for the full duration of the call, so the number of bits involved for a given call duration is not dependent upon whether actual conversation is taking place. Put differently, there is a direct and predictable relationship between the duration of a VoIP call and the number of bits that will need to be transmitted – in both directions – over the Internet or other IP transmission facility. Additionally, in the case of interconnected VoIP – i.e., VoIP traffic that either originates or terminates on the PSTN – the resources involved in the PSTN end of the call are, like any other PSTN traffic, duration-sensitive.

There are many reasons why regulatory policy must be technology-blind. Existing regulation deliberately or inadvertently operates to treat TDM and VoIP services differently across a broad range of policy areas – support mechanisms, intercarrier compensation, jurisdiction, among others – and these disparities need to be corrected. But the "corrections" need to be in the form of eliminating uneconomic cost shifting and subsidies that currently apply to PSTN services, not by extending these same treatments to VoIP.

The FCC seems to agree in principle, although not necessarily for the right reasons. In the NPRM, the Commission posits, but without any factual basis or support, the notion that "because most [PSTN] intercarrier compensation rates are set above incremental cost, they create incentives [for ILECs] to retain old voice technologies and engage in regulatory arbitrage for profit." However, in reality, PSTN carriers – primarily ILECs – are retaining "old voice technology" precisely because of the decline in voice MOUs that has been taking place over the past decade. As the NPRM observes, switched access MOUs peaked at roughly 550-billion in 2000 and by 2008 had plummeted to just above 300-billion. This rapid drop-off in switched access demand was further compounded by an even larger decrease in dial-up ISP MOUs over the same period as consumers migrated to broadband. The result is massive amounts of excess switched service capacity the costs of which are largely sunk at this point. There is little economic justification for tossing this in-place and perfectly serviceable capacity aside merely because IP exists as a technological substitute. Indeed, when correctly viewed in the context of our current national economic recession and budgetary crisis, unnecessary replacement of these assets is antithetical to our national interests and to our economy generally. Moreover, due to the disparate treatment of VoIP and TDM for purposes of intercarrier compensation, there has never even been a market-driven, level playing field test of these two alternative voice telephony technologies. We don't know for certain that VoIP is actually more efficient than TDM, but what we do know for certain is that using existing TDM capacity whose costs are entirely sunk is certainly more efficient than incurring new capital investment costs to replace these assets with IP technology.

There are, to be sure, numerous deficiencies and inefficiencies in the existing ICC system that must be addressed and resolved, but not for the purpose of incenting incumbent carriers to make investments that serve no valid economic purpose. Indeed, "correcting" the existing ICC regime so as to "incent" carriers to prematurely and unnecessarily abandon their TDM assets should be seen as antithetical to the Commission's overarching goals.

For more information, contact Colin B. Weir at cweir@econtech.com

Read the rest of Views and News, June 2011.

About ETI. Founded in 1972, Economics and Technology, Inc. is a leading research and consulting firm specializing in telecommunications regulation and policy, litigation support, taxation, service procurement, and negotiation. ETI serves a wide range of telecom industry stakeholders in the US and abroad, including telecommunications carriers, attorneys and their clients, consumer advocates, state and local governments, regulatory agencies, and large corporate, institutional and government purchasers of telecom services.