AT&T's acquisition of T-Mobile lets it build a truly national, robust network at the expense of competition: It's a little dirty but barely a secret in modern mobile cell world that AT&T doesn't really have national 2G coverage, much less 3G. AT&T leans on T-Mobile to roam customers in a large number of areas in which AT&T didn't spend money to build out service. This stems from an agreement years ago when AT&T Wireless consolidated on GSM service, and T-Mobile was building out its initial GSM service. In 2004, the companies dissolved a cooperative agreement (when Cingular bought what was then AT&T Wireless), but roaming never disappeared.
This lack of coverage is why AT&T didn't offer feature phone or smartphone service in large parts of the country outside urban areas. While these were mostly rural—such as Montana—you'd also find missing areas in adjacent cities in some markets. Because AT&T, like other carriers, only allows a fraction of one's usage to be on domestic roaming, you had a lot of peeved would-be customers who now own a Verizon iPhone
T-Mobile provided roaming 2G coverage in a lot of those areas, even though AT&T spent billions in 2009 to acquire licenses Verizon Wireless was obliged to sell to clear its deal for Alltel, the number five US carrier at the time. Still, AT&T will benefit from having consistent national service if the T-Mobile merger is approved by regulators. It's not a done deal.
AT&T also gets the depth of T-Mobile's spectrum portfolio in dense markets where AT&T clearly lacks the ability to deliver service to the level needed, such as New York City's boroughs and San Francisco. It won't be trivial to integrate the networks, but many carriers co-locate equipment with tower and building owners. And if they maintain the current deal and roaming is no longer a for-fee arrangement, AT&T can instantly get the benefit.
Both firms aligned across the same technology. Not just GSM, although they're the only two national GSM in the US. But they both chose to push short term on faster HSPA: HSPA 7.2, which challenges EVDO Rev. A by a factor of two or more, and HSPA+ in a 21 Mbps flavor, which can challenge the low-end of Verizon's 4G LTE rollout service—but nationally, not just in the one-third of the country to which Verizon expects to offer LTE by year's end.
However, T-Mobile's path was limited. While it extolled the virtues of HSPA+, which squeezes into 5 MHz channels, it had no real ability to acquire the additional spectrum needed for wider channels to exploit LTE. AT&T and Verizon collectively spent billions to lock down most of the sweet 700 MHz spectrum over which Verizon has already started its LTE deployment, and that AT&T will use starting mid-year for its own efforts.
On the Wi-Fi side, T-Mobile effectively exited the hotspot market in 2008, although most people didn't notice. The firm was able to sign a reciprocal five-year agreement with AT&T for access, which allows T-Mobile customers to use AT&T's network at no additional cost or fuss. That was more important when AT&T's network was largely paid or required hoops to get free service. AT&T's Wi-Fi network now comprises about 21,000 locations, of which about 20,000 are entirely free McDonald's and Starbucks stores. Barnes & Noble is in there somewhere, too. The rest are hotels and a few airports.
The convergence of AT&T and T-Mobile's interests are fairly obvious. Verizon Wireless and T-Mobile don't line up because Verizon already has thorough national coverage with 2G and 3G (provably the best 3G coverage), and uses an incompatible 2G/3G technology in CDMA. While Verizon has a path to GSM in its 4G flavor, it will be using CDMA for 2G and 3G for many years to come.
Sprint Nextel is engaged in pursuing three separate standards. iDEN, used by Nextel, is still in use, despite the firm's best efforts to migrate users to CDMA. Sprint's core 2G/3G network is CDMA. Its 4G plan was to get WiMax deployed early and extensively, which was furthered when it acquired Clearwire with its separate spectrum licenses and operations. That didn't pan out. WiMax needed a much faster deployment, and the money wasn't there to do it. WiMax is an also-ran technology cell mobile; it will have great niche uses and might be the most appropriate technology in some countries. But LTE will rule the Asian, European, and North American markets. Sprint Nextel has also not completed a multi-billion-dollar requirement to migrate public-safety networks to new frequencies in exchange for new spectrum. They are far overdue, and that ugly situation shows no sign of completion to my knowledge.
The real question is whether the Justice Department, FCC, and FTC will allow a merger to take place. There's no benefit to consumers from this merger, reducing competitors from four to three. Sprint Nextel arguably has no good plan for long-term viability, and a deal for Verizon to acquire it might be allowed to avoid bankruptcy, which wouldn't benefit the market (although Sprint could shed massive debt, union contracts, and likely federal obligations which would prove what everyone said when the public-safety spectrum swap was allowed years ago under FCC chair Kevin Martin.)
T-Mobile's plucky upstart nature has gained it over 30m customers, and allowed it to nip at the heels of the big three, likely saving customers billions of dollars a year collectively. The FCC and Congress never intended initially for a few carriers to win. Anti-regulatory and pro-incumbent fervor has led to a situation where there may be only two viable national carriers: AT&T and Verizon Wireless.