Captive Audience: The Telecom Industry and Monopoly Power in the New Gilded Age (41 page)

AT&T argued that the deal should be scrutinized on the basis of local markets, many of which have three or four carriers from which consumers can buy service, and it claimed that this meant wireless service was highly competitive.
⁴²
But this is like asserting that Washington, D.C., has several football teams: the NFL Washington Redskins, the Georgetown University Hoyas, and the Gonzaga College High School Eagles. Only AT&T and Verizon provide reliable nationwide service. They own the physical lines that connect to the wireless towers and carry data to the Internet backbones. Smaller carriers like Sprint and TMobile have to pay AT&T and Verizon for the privilege of roaming services in areas where they do not have their own towers. But AT&T went on hiring scores of influential lobbyists, muscling its way through Washington, and repeating the mantra that all relevant markets were local.

Its timing and political planning seemed to be pitch-perfect. The Obama administration, busy touting the vital importance of wireless access to
America's future, was put into an even tougher position than it had been with the Comcast-NBCU merger. Blocking the deal would allow Republicans to paint the administration as antibusiness. Given that no large businesses other than Sprint publicly opposed the deal, that it seemed to dovetail with the administration's own interest in improved wireless high-speed data access, and that the administration could use the deal to exact public policy improving conditions, it looked initially as if AT&T would win the day.

Why did TMobile decide to sell itself to AT&T? One word: spectrum.

Comcast can keep competitors at a distance because it has snapped up exclusive franchises—given by the government—around the country, and then clustered its operations so that it owns the whole of market where it chooses to sell services. The programming it controls—most important, sports—is another barrier to entry; Comcast's competitors have to have this content, and Comcast can charge whatever it wants for it. No one is likely to have the resources to enter the market to compete seriously with Comcast in the distribution business—a competitor would have to offer great video programming and, at the same time, shoulder the enormous up-front costs of building a network.

AT&T has its own ace, and it also came from government. Instead of licenses to do business (like a cable franchise), AT&T has licenses to transmit signals, granted by the FCC.

AT&T and Verizon inherited an enormous amount of beachfront—lower frequency—spectrum from their corporate ancestors. TMobile acquired its spectrum holdings through Deutsche Telekom's acquisition of Voice-stream, which had bid successfully in the FCC's 1994 PCS auctions, and through its own $4.2 billion bid in a 2006 AWS (Advanced Wireless Services) auction. Both PCS and AWS are higher-frequency bands than AT&T and Verizon's 700 MHz spectrum (for which they paid a combined $15 billion in 2008) and 850 MHz spectrum (granted free to AT&T's and Verizon's corporate ancestors).

The amount of information that can be conveyed goes up with higher frequencies, but the distance data can travel goes down, and the data are more easily interrupted by physical objects in the way. The high-frequency bands can thus indeed carry gigabits of information, but they require cell
towers every hundred yards or so. This is what makes WiFi faster than commercial wireless but limits the signal to the area around your house. A carrier with licenses to lower-frequency spectrum can thus serve a territory better than a carrier with higher-frequency spectrum, because it can build far fewer base stations. This is a major cost advantage held by AT&T and Verizon. It would take significantly more cell sites to serve TMobile's customers using its spectrum than to serve the lower-frequency spectrum customers of AT&T and Verizon, and even then, indoor coverage would probably suffer.

These spectrum issues have created a gap between TMobile (and Sprint) and the two big wireless carriers, with their broad, unchallenged holdings in the 700 MHz and 850 MHz bands. Nonetheless, TMobile's management told investors in a January 2011 conference call that the company was “a very good asset” that was generating positive annual free cash flow of between $2.5 billion and $3 billion a year. TMobile, claimed its management, had a strong network architecture in which half its towers were already fed by fiber, terrific smartphones, the best value for consumers, great customer service, and higher (and growing) margins on revenue than Sprint. Why the optimism? TMobile's management also said during that call: “We're absolutely positive and optimistic about [the] commercial option in [the 700 MHz] D block.”
⁴³
TMobile, in other words, was hoping to buy additional beachfront, low-frequency spectrum that the FCC had planned to make available: the nationwide “D block” of 10 MHz within the 700 MHz overall allocation. If it got access to that spectrum, it could catch up with Verizon and AT&T. And there was another block of federally used spectrum that TMobile planned to pair with its own existing holdings. More spectrum would give TMobile a chance to compete.

When the administration failed to follow the FCC's lead and appeared likely to take another auction for commercial use of the 700 MHz D block off the table, Deutsche Telekom apparently could not see a path forward and decided to pursue a merger. In a sense, the Obama administration itself caused TMobile to seek a deal. The company could see the writing on the wall: it would never have the beachfront spectrum holdings or scale to enable it to compete effectively with AT&T and Verizon, so it might as well fold its tents.

The merger announcement had an immediate effect on the viability of Sprint, the other also-ran wireless carrier. Talks between Sprint and TMobile had been going on for months before AT&T swooped in. Given the deal's apparent inevitability, Sprint's market valuation plummeted. John Delaney, research director with the analyst firm IDC, noted, “With Sprint's position looking increasingly difficult, and with one of Sprint's options for improving its position, a merger with TMobile USA, now off the table, the US looks like it is heading towards a duopoly of national mobile operators.”
⁴⁴

AT&T is fond of pointing out that devices like smartphones and iPads generate twenty-four times more data than conventional wireless phones, and that AT&T's mobile data volumes climbed by 8,000 percent from 2007 to 2010. AT&T's networks cannot bear the strain, and the company says that the answer—the only answer—is more spectrum.
⁴⁵
Eager to help, the Obama administration in 2011 launched a search for 500 MHz of spectrum to be reallocated to wireless data use from less efficient uses.
⁴⁶
Every closet in every federal agency is being scoured for spectrum. This process will take years—on average, it takes six years to identify, reallocate, and distribute spectrum—and a scores of civil servants will invest a great deal of energy in the spectrum hunt.
⁴⁷
(Prediction: the Department of Defense will fight giving up any spectrum it has been allocated using every lobbying tool it has available. The military-industrial complex in this country is enormously effective at avoiding sharing its spectrum assets with commercial actors.)

In the meantime, it is worth continuing to ask whether the problem is solely, or even mostly, spectrum. The large wireless carriers could also increase the information-carrying capacity of their networks by building more towers and connecting them to fiber rather than copper wires. Today, even though 97.8 percent of the U.S. population has 3G coverage, more than 80 percent of cell sites are still connected to copper wires.
⁴⁸
But since the goal of any private company seeking Wall Street investment is to achieve the same levels of revenue (or more) while laying out less money, spending on “backhaul” (connections between towers and Internet access points) has not been a high priority. The problem in wireless transmission, therefore, is probably the wires and the towers, not spectrum. Executive compensation and quarterly results trump higher-quality service every time.

Avoiding capital-intensive fiber installations where possible will both please investors and continue to shape users’ expectations; Americans will be used to slow, crippled, heavily curated, and compressed mobile services. AT&T can also use the usage-based billing model to ration usage of its network while reducing the attractiveness of potentially competing data services and other nonaffiliated products crossing its wireless networks; consumers will not want to use competing services, even if they technically can, because they will thus be subject to large overage charges, service cut-offs, or other remedies imposed at the carrier's discretion.

All in all, AT&T's effort to merge with TMobile made sense; it was trying to force a utility communications service (similar to the water and electricity companies), with its extraordinarily high up-front costs and sharply declining cost curves, into a private, profit-making, model that would be attractive to investors, and the only way to do that was to continue to scale, tightly ration capacity, employ price discrimination, keep capital costs down, and eliminate competitive ideas that would undermine the model. TMobile's low-priced services and open development platforms were one such disruptive idea. Acquiring TMobile would have eliminated that alternate mindset while letting AT&T add subscribers without increasing its employee headcount.

Thus, although “more spectrum!” became the call of all of the carriers in 2010, it was unclear whether spectrum was the problem—just as it had been unclear in the Comcast-NBCU transaction whether “saving the NBC Peacock” was a strong enough reason to approve the deal. TMobile had hoped that new spectrum would be made available for it to bid on so that it could compete. AT&T and Verizon already had a lot of spectrum. AT&T had more than anyone else in the top twenty-one markets in the United States, and it was using less than a third of what it had.
⁴⁹
Verizon said it did not need any more.

But with Jim Cicconi shaping the story line, “more spectrum” would take the form of more capacity offered by the same player. The merger between AT&T and TMobile would solve the country's broadband problem.

When the Department of Justice sued to block the merger in August 2011, arguing that the national market was relevant and that getting rid of TMobile would reduce horizontal competition, AT&T was shocked. The
company had been supremely confident of approval, and its response was remarkable: “We are surprised and disappointed by today's action particularly since we have met with the Department of Justice and there was no indication from the DOJ that such an action was being contemplated.”
⁵⁰

If anyone in AT&T's enormous communications–public relations shop had known the story of J. P. Morgan's meeting with Theodore Roosevelt, the company would not have made that statement. It echoes Morgan's words when Roosevelt's attorney general sued his railroad trust under the Sherman Act. AT&T, like Morgan, apparently thought of itself as the government's equal.

What made the difference between the Comcast-NBCU and AT&T–TMobile cases? AT&T had brought even greater political pressure to bear on the administration: even more letters from state governors and nonprofits had been filed in support of the merger, even more lawyers and lobbyists had been deployed, and even more adept political messaging had taken place at the highest levels. But this time the merger-specific merits of the antitrust case seemed clearer; the Antitrust Division's objection to the increase in horizontal concentration seemed more likely to be upheld by a court. Besides, the department had just been through the deep political waters of the Comcast-NBCU deal and wanted to remind the country that it was a law enforcement agency. AT&T's CEO, Randall Stephenson, took a slap on the hand as a result of the failure of the deal, losing $2 million of his total compensation for the year—which left him with $22 million in his yearly pay package.
⁵¹
AT&T took a charge on its books to cover the breakup fee ($1.4 billion after the tax write off).
⁵²
But its stock price had gone up more than 8 percent since the deal had been announced—more than Verizon's had climbed. AT&T had also won: it had severely damaged TMobile, undermined Sprint, gotten access to a lot of TMobile's strategic planning documents, and stalled any potential change in the regulatory structure for its sector.

Better regulatory policy on the wireless side would help foster competition, innovation, and lower prices: the government could make spectrum available to more players for less money by capping participation in auctions or charging fees for spectrum; it could require existing carriers to share their towers, thus lowering the costs of doing business for new
competitors; it could make more unlicensed spectrum available; and it could oblige wireless providers to act as common carriers when it comes to the Internet data passing over their airwaves. But there is little or no impetus for government intervention in wireless. And at the same time the government is relying heavily on wireless as the answer to America's Internet access problem.

Will wireless access help America reach the president's goal of one gigabit to every community? No. It bears repeating: wireless access cannot be a direct substitute for high-speed wired services (other than the legacy DSL services, which have already become irrelevant). Ever since the dawn of the digital age, wireless-technology speeds have lagged behind wired speeds by substantial margins. Published data rates of post–third generation standards—Verizon claims peak speeds of around 25 Mbps, and average speeds of around half that—seem high. But these numbers need to be understood in context. They assume optimal (and unlikely) conditions, such as a single user or a few users who can use all the available bandwidth without sharing and are in close proximity to a base station. In addition, it is possible to achieve these rates only by using large amounts of spectrum, generally more than is available for current 3G systems, and by using relatively small cell sizes—which means building lots of towers or deploying many base stations and serving them with fiber, which AT&T and Verizon have not done and are experiencing no competitive pressure to do. The laws of physics make it extremely unlikely that wireless connections over long distances will ever be capable of delivering the hundreds of gigabits per month that users will want to consume over their data connections.