Spectrum crisis? What spectrum crisis?
The Times' editorial board expressed skepticism this week about the AT&T's proposed purchase of T-Mobile, noting that it "would eliminate one of the four largest U.S. mobile phone networks and leave just two companies — AT&T and Verizon Wireless — in control of more than 70% of the market." But we didn't tell regulators to put the kibosh on the deal, noting that it might lead to a better-performing network than AT&T and T-Mobile could produce on its own (at least in the foreseeable future).
One of the main argument in favor of the purchase is that the demand for mobile bandwidth is skyrocketing, rising faster than carriers can accommodate with the spectrum at their disposal. Federal Communications Commission Chairman Julius Genachowski has warned repeatedly about a looming "spectrum crisis," but even under the best-case scenario, new airwaves won't be available for wireless broadband for many years. For AT&T, buying T-Mobile and its frequencies is a shortcut to more spectrum.
Not everyone buys the idea of a debilitating spectrum crunch, however. Analyst Dave Burstein has argued that AT&T already has ample airwaves, particularly in the bands it's just starting to use for 4G services. In his DSL Prime newsletter Friday, Burstein offered another data point to undermine AT&T's case. A new projection by networking equipment kingpin Cisco predicts that demand for mobile bandwidth will increase at a slower and slower rate in the coming years, as the penetration of smartphones slows.
That makes sense, right? Much of the growth in the demand for bandwidth has come from two parallel forces: a new type of smartphone (epitomized by the iPhone) encourages people to make more use of the mobile Web, and more people are switching from conventional mobile phones to these new smartphones. Once everyone has an iPhone, an Android phone or the equivalent, much of the growth goes away.
Cisco estimates that annual rate of growth in the demand for mobile bandwidth demand will increase from 120% in 2010 to 140% this year, then decrease to 46% in 2015. Extrapolating from Cisco's numbers, Burstein projects that mobile bandwidth demand will grow only 30% to 40% in 2016 and 2017.
That's still double-digit growth, which is nothing to sneeze at, but Burstein thinks it's manageable. Wired broadband networks "have become less congested despite eight years of 30% to 50% growth," he wrote, because improvements in chip technology (driven by Moore's Law) have expanded the networks' capacity at a similar pace.
Moore's Law applies to wireless networking gear too, but not to the cost of building towers and leasing spectrum. The latter are the big unknown in this equation. As Burstein put it, "How much capital [those] costs will demand is a hard question."
-- Jon Healey
Photo: Attendees at a mobile phone trade show in Barcelona experience a capacity crunch at an Android smartphone display. Credit: Reuters / Gustau Nacarino