Shutting Down The Phone System Gets Serious: The Implications of AT&T Upgrading to An All IP Network

(This is a reprint of a piece I wrote for the Public Knowledge blog)

I believe AT&T’s announcement last week about its plans to upgrade its network and replace its rural copper lines with wireless is the single most important development in telecom since passage of the Telecommunications Act of 1996. It impacts just about every aspect of wireline and wireless policy.

For those who missed it in the morning-after blur of the election results, AT&T announced that it will invest an additional $14 billion to upgrade its wireline and wireless networks, so that it projects investing $22 billion a year for the next several years in capital expenditures (“CAPEX” as they say on “The Street”). At the end of the three year time frame, AT&T expects to have converted its existing “time division multiplexing” (TDM) phone network entirely to an IP-based network which will seamlessly mix its wireless, remaining souped-up copper, and fiber (but not fiber-to-the-home). Since all existing phone regulation governing universal service, consumer protection, and competition rest entirely on the existing TDM/copper network, AT&T simultaneously filed a petition with the FCC to “begin a dialog” on how to address the regulatory issues raised by this shift and proposing some entirely deregulated “pilot programs” to determine what regulations are “really” necessary.

Setting aside my skepticism that these pilot programs offer anything of value, I thank AT&T for beginning with an offer to talk. At the same time, I’m mindful we need to get the key elements of the new framework down over the next year or two – which is practically nothing given the complexity of the issues and the number of stakeholders involved. It puts a premium on communities working quickly to come to internal consensus and on trying to bring as many allies to the table as possible. Ideally, we would set universal rules for all IP networks, but this would meet fierce resistance from existing IP-providers. Nevertheless, AT&T raises a valid point of concern if the rules for the TDM to IP apply only to it and other Local Exchange Carriers (LECs) upgrading their networks. The FCC must balance these concerns about competition and fairness with the broader questions of what happens when our 100-year-old copper safety net gets replaced by an essentially unregulated IP-based networks.


What’s At Stake? Everything In Telecom Policy.

To list just the headline questions:

  • What happens to the concept of universal service, particularly in rural areas? AT&T itself says in its announcement that its new combined 4G LTE and wireline IP footprint will cover “Ninety-nine percent of existing customer locations.” That loss of 1%, while small in absolute terms, potentially means many thousands of people losing access to basic phone service.
  • Even if AT&T’s wireless footprint precisely overlapped its rural copper footprint, there would still be significant questions about reliability and price.  Traditional phone service has minimum standards of quality enforceable at the local level, and in most states a requirement to offer basic voice service at a regulated price. This combination of a requirement to serve everyone in the service territory, at a minimum standard and to offer a basic, affordable voice option is generally thought of as “Carrier of Last Resort” (CoLR) regulations. If AT&T transition to IP-based networks and eliminates its CoLR obligations associated with its traditional telephone service, what happens? In rural areas, wireless signal might not have the same quality as existing copper, or it might become unaffordable for poor subscribers in rural and urban neighborhoods where low-income families rely on a low-cost basic voice service.

Nor can we ignore the impact on the hard of hearing and the visually impaired. Basic copper voice service has built-in capabilities for the deaf, and home equipment designed for the visually impaired. How will the changeover impact these communities?

In other words, while the new network will definitely be a better network, offering a more valuable bundle of services, it may not be either available or affordable to all Americans in the way existing phone service has been for decades. We could become the first industrialized nation to movebackward on availability of basic voice service to all people.

  • Copper is built to 99.999% reliability, the “five 9s” of a public utility. Wireless and IP-based networks do not come anywhere close to that. In the aftermath of Hurricane Sandy, New Yorkers without power for IP-based services or cell service went combing their neighborhoods for payphones – which continued to operate because they connect through independently powered copper networks. What happens in 22 states when that older, more expensive but more reliable technology is replaced with better, faster but more fragile technology?


  • What happens to the Universal Service Fund? The current AT&T telephone network provides a significant portion of the existing contribution to the fund because only traditional telephone networks are (at the moment) required to pay into the fund. DSL and other IP-based networks do not pay into the fund.

For years, contribution to the Fund has been eroding. AT&T’s proposed rapid transition of its entire network will dramatically accelerate this trend. Many states have separate state-based USF funds. AT&T territories, this will be even more impacted.


  • A host of policies remain confined to the telephone world and not the IP-space. These include pro-consumer policies on privacy and billing, and pro-competitive policies such as special access and unbundled network elements (UNEs). To the extent these policies apply at all to IP-based networks, they do so as “ancillary” to existing telephone networks. Even access to phone numbers is technically limited by statute to “telecommunications providers,” and is assigned to IP-based networks by regulatory fiat. What happens to these policies when the telephone network that supports these policies disappears? What protections will consumers or competitors continue to enjoy with regard to things like privacy, truth-in-billing, and signal quality?
  • Most importantly, the FCC must decide the question of whether IP networks must interconnect with one another. Right now, they do not. And every now and then we have a “peering dispute” where networks refuse to exchange traffic because they cannot agree on terms. What happens if AT&T and Comcast cannot agree on terms, and several million AT&T Wireless subscribers can no longer call home? If this seems unlikely to happen, I remind folks that no one ever imagined that cable systems or DBS providers might go for weeks or even months without local broadcast channels because of “retrans fights.” However, when DISH subscribers must make do without AMC or when Cablevision subscribers can’t watch Food Network, it’s merely annoying. When people with Comcast phone service can’t call AT&T or AT&T wireless subscribers, the impact – for the economy, for public safety, and for the individuals involved – is a hell of a lot more significant than missing “Mad Men” or “Chopped.”
  • How will AT&T’s investment impact the digital divide and problems of digital adoption, particularly in minority communities? There is no question that, handled correctly, this massive investment in infrastructure could prove a tremendous boon to communities that have until now been in danger of marginalization. At the same time, we must also recognize the danger that a poorly handled transition to all-IP networks could cement the status quo, or even worsen existing digital inequities.

This is a daunting list of questions to resolve. Many of them have been the focus of quiet conversation in telecom policy circles for years. AT&T itself raised these questions during the National Broadband Plan in 2009.  But while everyone could see that someday we would need to address these issues, the all-IP future always seemed comfortably enough away that the FCC could defer these questions for another day.

That day has now come.  AT&T’s announcement that it will phase out its traditional phone network and convert entirely to an IP-based network over the next three years forces us to address these questions in a relatively short period of time.  AT&T, to its credit, has acknowledged that it must undertake a dialog with stakeholders and regulators. For our part, we at Public Knowledge welcome this opportunity.

Real Investment That Creates Stronger Networks Is Always A Good Thing.

Let us be clear. We want this investment to happen. This investment will create a combined wireless and wireline network that is truly greater than the sum of its parts. This is how competition is supposed to work. AT&T invests to meet the challenge from cable operators on the landline side and from its wireless competitors. This, in turn, forces the cable operators and wireless carriers to respond with improvements and lower prices of their own.

Those dismissive of the upgrades to copper overlook two critical aspects. First, it was only a few months ago when all copper was supposedly destined for the ash-heap of history, and the war between cable and copper settled entirely in cable’s favor. This upgrade potentially restores DSL as a viable, low cost alternative able to put price-pressure on cable systems and force cable operators (particularly in mid-size markets) to accelerate their own upgrades.  Second, and more importantly, this network is designed to fully integrate its wireless and copper components. This will have major impacts on spectrum policy and vastly accelerate the shift to small cell architecture and the trend of enhanced reliance on unlicensed spectrum to augment licensed spectrum.

More importantly, this investment and network upgrade will happen whether we want it or not. We at PK do not doubt that AT&T intends to invest the promised $22 billion per year over the next three years. In the past, we have not hesitated to express hearty skepticism where AT&T (and other companies) have promised to invest in exchange for regulatory relief. But unlike previous announcements made to regulators and made explicitly contingent on receiving regulatory relief, AT&T’s announcement here is wholly different in nature and character. First and foremost, AT&T made this announcement not to regulators alone, but to a conference of Wall Street analysts. No publicly traded company tells Wall St. it plans to make enormous investments in capital expenditures that will take several years to pay off unless it has absolutely no choice.

And AT&T has no choice. Competition is forcing AT&T to invest in its networks or risk obsolescence. Cable providers have already taken AT&T’s residential wireline business, and are eating into AT&T’s commercial enterprise customers.  Verizon Wireless has a superior wireless network, and both T-Mobile and Sprint are pouring billions into network improvements and upgrades. AT&T either upgrades or goes under. This is why AT&T’s filing makes it clear that AT&T is going ahead with this investment whether or not the FCC grants it any kind of regulatory relief.

Where Do We Go From Here?

AT&T has called for a dialog to begin in earnest. While all stakeholders should welcome this opportunity, we must not lose sight of the fact that AT&T – as a profit-maximizing firm seeking to minimize its costs – will look to cut the best deal it can. I confess I am deeply skeptical that AT&T’s proposed regulation free zones in the form of a “pilot program” can teach us anything of value. Further, while AT&T has acknowledged that the shift to IP networks should not be entirely “regulation free” and must acknowledge the historic duties associated with its traditional copper network, I recognize that AT&T (in this filing and in past filings) has also suggested that the FCC’s authority, if push came to shove, is extremely limited.  AT&T is also at pains to observe that cable operators and other competitors do not face the same traditional regulatory burdens, and to sing from the usual hymnal the paeans of praise to the deregulated world of IP-based networks.

I do not say any of these things to suggest that AT&T is not approaching this with good faith. But no one should imagine that AT&T is a charity. This network upgrade raises critical questions that go to the very basis of the FCC’s authority over core communications in this country and the 100-year old social contract that made voice service in the United States universally available, affordable, and reliable. Even were the impact confined to the millions of customers in 22 states that subscribe to AT&T’s wireline service, this would be a critically important and monumental set of proceedings. The decisions made here will ultimately impact everyone, regardless of what service they use to communicate. This conversation will happen at the local level, the state level and the federal level. It will happen at the FCC and in Congress.


Every stakeholder community needs to carefully consider its position and come ready for some hard bargaining. The traditional battle lines and positions need to be carefully reexamined. The world is changing, and it will make a radical change like this only once. This is not the time to repeat the rote responses of the past. This is not going to be some Libertarian nirvana where the regulatory state withers away and we shall move from each according to his need to each according to his ability to pay. Neither is traditional regulation going to remain unchanged. Anyone taking either position is effectively removing him or herself from the real conversation that must take place.

What we need to do, individually and collectively, is figure out how to take the values of the 100-year old social contract in telecom that has served us so well as a nation and decide how to express those values in concrete terms for the next generation of networks. That won’t be easy. But stuff that matters never is.


  1. “Even if AT&T’s wireless footprint precisely overlapped its rural copper footprint, there would still be significant questions about reliability and price. Traditional phone service has minimum standards of quality enforceable at the local level, and in most states a requirement to offer basic voice service at a regulated price.”

    Again Harold your lack of technical understanding shows through. One may have QoS in being equally able to reach a CSR to complain about service but that has nothing to do with the actual service itself. There are still many areas still where phone service is run over barb wire (West Texas) the rancher installed (and darn glad to use it), microwave towers for backhaul, (TX, AZ, NM) and installation of new copper/fiber by horse (central Vermont). The point is that even now we have 19th century layer 1 tech being used due to fiscal constraints. Under those conditions one CANNOT quote let alone meet a uniform QoS.

    “In rural areas, wireless signal might not have the same quality as existing copper, or it might become unaffordable for poor subscribers in rural and urban neighborhoods where low-income families rely on a low-cost basic voice service.”

    Then let it go. Let AT&T off the hook then at the same time remove AT&T’s control of the LATA space and the wireless coverage it has agreed to abandon. Let the community replace it —, Given the choice, most would embrace the more local, responsive community set up. No regulations, just get the job done.

    “Copper is built to 99.999% reliability, the “five 9s” of a public utility. ”

    In 1980 that would have been true. That is not true today, at least with Verizon, and the 9’s are shrinking all the time. The fact is budget for wireline maintenance has been slashed to the bone and has been for years. I would just as well point you to the experience of Katrina where entire sections of the wireline plant was no longer in service.

    “Most importantly, the FCC must decide the question of whether IP networks must interconnect with one another.”

    An interesting question that does deserve an answer. However the question needs to be positioned differently — An explicit contract exists between the end customer and ALL providers to the service(s) in question. No interruption may ensue that inconveniences the end point buyer. Treble damages otherwise. It would induce the behind the scenes offerers to either work it out or work an alternative. It would also over the long haul transition away from the channel bundling world we now live in to an ala carte world that most end customers would prefer.

    “At the same time, we must also recognize the danger that a poorly handled transition to all-IP networks could cement the status quo, or even worsen existing digital inequities.”

    To a point I say so be it. The reality is somebody would service the community if not the community itself. Its a corollary to my observation above. Release the legacy carriers from both their obligation and their opportunity (land line and wireless). Then permit a period of land rush to occur followed by inclusion into an reg-lite regimen suitable to the environment that low served communities face.

    It is only a daunting list of questions if one is attempting to shoehorn an entire nation into one solution for all. That is the fundamental flaw with regulatory regimes. The fact is there are multiple solutions available, most of which will work if only they are allowed to.

  2. Pingback: Wetmachine » Tales of the Sausage Factory » Wheeler Blog Post Shifts Phone Transition Into Second Gear.

Comments are closed