Bumps in the road for AT&T-BellSouth merger?

The conventional wisdom may be wrong. There is a twist that ought to give the regulators pause. Both BellSouth and AT&T own considerable chunks of prime wireless spectrum that is unused and that could quickly and relatively cheaply be used to provide broadband services that would compete with telecom and cable wireline broadband services--in other words, the long-sought "third pipe" to the home or business.

Competitors will no doubt argue that the Federal Communications Commission and the Justice Department should worry about the loss of intermodal competition that will result if the two companies are permitted to retain this valuable spectrum.

Over the last year, the development of worldwide standards for WiMax and improvements in equipment performance and cost have made the 2.3GHz and 2.5GHz frequencies more attractive. They are now among the most desirable for use in providing broadband voice, data and video services that will compete with telecom and cable wireline broadband services. These spectrum bands can also be used to compete in the backhaul and special access (both last-mile and interoffice transport) markets where the regional Bell operating companies are also dominant. Soon. they may even be used to compete with cellular and other mobile voice and data services. Equipment costs will be low and will keep falling as economies of scale kick in.

Intermodal competition cannot develop if two of the three modes--wireline and wireless--are in the hands of a single, gigantic company.

Off-the-shelf WiMax products for the 2.5GHz band will be available from numerous manufacturers later this year, and a 2.3GHz version of WiMax may be developed. Mobile WiMax equipment that provides voice and high-speed data is now being tested on KDDI's network in Japan and should be widely available by early 2007.

The competitive importance of this spectrum was acknowledged by the FCC last year. As a condition to its approval of the Sprint-Nextel merger, the FCC required that Sprint meet certain milestones for offering broadband service in the 2.5GHz band. Sprint agreed to offer such service to at least 15 million Americans within four years of the order (for instance, by fall 2009), and to offer it to an additional 15 million customers by two years later. Sprint controls the most 2.5GHz spectrum in the U.S., and its planned use of this spectrum to provide services competitive with (or as Sprint claims, better than) wireline broadband services is one of the key drivers of Sprint's ongoing spinoff of its wireline operations. But Sprint's spectrum holdings, large as they are, do not provide adequate coverage in many areas, particularly in the BellSouth footprint.

BellSouth, the second-largest owner of 2.5GHz spectrum in the U.S., controls spectrum in most of the 50 largest markets, according to published reports. It also has substantial 2.3GHz spectrum (acquired in auctions in 1997). SBC Communications also gained a large amount of 2.3GHz spectrum when it acquired AT&T. According to publicly available information, neither company has yet developed any of the spectrum into a commercial line of business; rather, the spectrum is effectively warehoused at present. (However, BellSouth has deployed test broadband service in nine in-region markets using its 2.3GHz spectrum, and it announced in March that it would use the 2.3GHz spectrum to provide "backup" broadband for its wireline broadband service.)

The proposed merger will concentrate huge blocks of spectrum in one company with every incentive not to use it. Assuming that AT&T's licenses are spread equally around the U.S. and that BellSouth's licenses are located primarily in the BellSouth service area (as appears to be the case), then prior to the merger, approximately 40 percent of the BellSouth and 30 percent to 35 percent of the AT&T licenses are located in the parent Bell operating company's (BOC) service area and are not likely to be used to compete with companies' wireline broadband services.

The postmerger wireline service area will be much larger, however, and substantially less than 50 percent of the spectrum would be out of region. It is unlikely that a postmerger AT&T would have any incentive to develop this spectrum for wireless broadband. Competing with itself in-region would not be attractive; and BOCs have yet to engage in meaningful competition out of region.

Given this scenario, competitors can be expected to argue that the FCC and the Antitrust Division should require BellSouth and AT&T to divest at least their in-region 2.3GHz and 2.5GHz spectrum (together with any existing deployments). Divestiture benefits consumers and businesses if the spectrum is used by an entity or entities that will compete in the broadband, backhaul and special access markets.

It may even be argued that AT&T and BellSouth should be required also to divest all their out-of-region spectrum in order to enable nationwide competition.

Preventing a BOC from acquiring assets that could be used to compete with its core wireline business--and in this case also with its growing wireless business--is not a novel idea. In the Primestar case a few years ago, for example, the Antitrust Division prevented cable operators from buying DBS satellite assets, for fear they would warehouse the satellites in order to prevent new entrants from using them to compete with traditional cable TV.

Similar precedents involve the divestiture of manufacturing plants, drug compounds in the FDA approval pipeline, and other competitive assets in non-telecom markets.

The FCC ought to be particularly receptive to such arguments, given the importance that former Chairman Michael Powell and current Chairman Kevin Martin have placed on the development of facilities-based intermodal competition in their decisions and public statements these last few years. Intermodal competition cannot develop if two of the three modes--wireline and wireless--are in the hands of a single, gigantic company.

Interestingly, the fact that the FCC has in the past permitted BOC ownership of this spectrum and never found it to be a problem is irrelevant. The public interest standard that the FCC applies in merger approval proceedings is based partly on antitrust analysis, but also explicitly permits it to "consider technological and market changes, and the nature, complexity and speed of change of--as well as trends within--the communications industry."

Thus, the FCC could take into account the recent technological and market changes that have transformed this spectrum into a valuable competitive alternative to wireline broadband services. The FCC could determine that the merger, by effectively shielding so much spectrum from the hands of potential competitors, would not be in the public interest absent substantial spectrum divestiture.

The Antitrust Division may also act. Provided it found that the spectrum controlled by the postmerger company represents a source of substantial potential competition, it would be fairly straightforward for it to conclude that without divestiture of the spectrum, the effect of the merger "may be to substantially lessen competition"--the standard it must follow in reviewing mergers under the Clayton Act.

Whether the FCC or the Antitrust Division in this administration will take competition-preserving action is an open question. But for those who seek competition in U.S. broadband markets, and for proponents of network neutrality for whom a third pipe could be a godsend, it certainly looks like a fight worth fighting.