Does AT&T Really Want to Get Rid of the PSTN?

When AT&T responded late last year to the FCC's request for comment on the transition from circuit-switched to all-IP public telephone infrastructure, it seemed to mark a turning point in the history of telecommunications. The 30-page letter the telco filed on December 21 urged the total phaseout of POTS (traditional "plain old telephone service") and the PSTN (public switched telephone network) on which it runs. It also recommended that the agency set a firm date for the transition in order to ensure that it happen as quickly as possible. Thus the company whose name is virtually synonymous with traditional telephony seemed to cast a clear vote in favor of the all-VoIP future. In reality, though, it was just as much a vote against AT&T's traditional service obligations. It was also a shrewd attempt to bind the phone company's interests to the high-profile issue of universal broadband access.

A central argument of the letter was that the dominant local phone companies known as ILECs, or incumbent local exchange carriers, currently have to support two networks. One of these is the traditional PSTN necessary to provide POTS service. The other is the broadband network that currently carries Internet traffic, and will eventually carry all voice traffic. The letter noted that while POTS usage and revenues are dropping precipitously, costs for maintaining the PSTN are in fact rising on a per-customer basis. That meant, it claimed, that having to support the PSTN hindered investment in broadband infrastructure.

Given the FCC's responsibility to come up with national broadband plan, and Washington's intense interest in the subject, focusing on broadband was politically savvy. But AT&T's professed concern for broadband was just one aspect of a deeper concern: getting rid of a set of regulations to which ILECs have been subject since they were government-sanctioned monopolies. The regulations, described in terms such as universal service, carrier of last resort and obligation to serve, force traditional carriers to provide telephone service even where it's unprofitable, in order to ensure that everyone has access to such a basic necessity. Currently only ILECs have to bear this burden, which puts them at a disadvantage compared to the companies that are increasingly cutting into their POTS businesses, including cable or other VoIP provides as well as cellular operators. AT&T unsurprisingly wants to at least share the burden.

The letter used a carefully constructed argument to bind the service obligations issue to the broadband issue as tightly as possible. The foundation of the argument was that the regulations imposing such obligations define them in POTS/PSTN terms. That meant, the argument went, that the regulations essentially forced ILECs to support and maintain the PSTN. And because both supporting the PSTN and building out a broadband network is more expensive than just doing the latter, it concluded, the service obligations as they are seriously harm the national goal of broadband buildout.

The argument was misleading in at least one important way: It claimed that the PSTN and the broadband network are two distinct networks, making it seem as if each required its own process of construction and maintenance. In reality, the two share significant portions of their infrastructure. For example, most long-distance voice traffic travels over IP links, sharing fiber connections with broadband Internet data traffic. And voice takes up a lot less capacity than most other types of Internet traffic.

In addition, DSL, a widely used broadband service, travels over copper wire to the customer home or business. That copper wiring is the same that carries PSTN voice traffic, so that POTS and DSL service share the same physical wires. Thus supporting the PSTN infrastructure is often an integral part of building a broadband network, not a separate activity. That in turn means maintaining the PSTN does not directly detract from the efforts of those bent on providing broadband service.

That's not to say it doesn't make any difference. Anyplace a traditional phone company has to provide POTS service, it needs PSTN equipment such as TDM switches located near the customer. Although that equipment too is getting less expensive as IP-based softswitches replace traditional Class 5 TDM central office switches, it's an additional expense service providers wouldn't have if all voice traveled over an all-IP broadband network.

Despite its flaws, the argument will likely prove at least partially effective in advancing AT&T's main goal, which is to change the definition of service obligations. Rather than something only PSTN-based providers can fulfill, AT&T wants the definition to refer to services that providers using any number of other types of infrastructure, especially broadband-based VoIP networks, can offer.

Once that happens, the carrier can push to share the burden with all voice providers, particularly the cable VoIP operators that are taking substantial market share from ILECs. If it succeeds in that, it won't have to provide PSTN service where it's unprofitable, but only where it makes money. At that point, its sense of urgency to move to all-broadband telephony will undoubtedly lessen.