I have been pounding the table here for some time on the idea that there is no spectrum shortage.
Much of my thinking on the question comes from David Isenberg, a former AT&T (T) engineer who wrote "The Rise of the Stupid Network" in 1997. Isenberg's thesis was that spectrum is an ocean, that it's not a series of railroad tracks, and that its use is highly expandable when the right technology is applied to it.
That technology is edge technology, like the WiFi router sitting next to your desk. Just as on the Internet itself, the technology used to move bits should be defined at the edge, not by some central gatekeeper. Thus, unlicensed frequencies that once ran data at 1 megabit per second, in the late 1990s, now can run 100 megabits. By regulating equipment rather than selling the resource, spectrum can be sectorized and cellularized again-and-again-and-again, making its ultimate carrying capacity practically infinite.
Go to nearly any hospital and you'll find that they're replacing wires with wireless links, managing the unlicensed spectrum with equipment that uses low-power and ever-smaller cells to expand capacity.
Proof lies in the fact that Google (GOOG) is now offering its stake in Clearwire (CLWR) for $47 million, just one-tenth of what was paid for it. The $1.60/share price being asked severely undercuts CLWR's recent per-share price of $2.27. And I haven't seen a rush to buy, even at that price.
Why is the lie still told? Why does it have such a hold on the imagination of policymakers and investors?
It's because, since the mid-1990s, government has been in on the scam, and profiting from the fiction. By selling spectrum rights and claiming immense value for spectrum, government acts like it's doing something when people claim about price hikes or bandwidth caps. In fact it's just making a bad situation worse.
In opposing spectrum transfers like the one recently negotiated between Verizon and cable companies the FCC claims it's trying to uphold a "free market," in which there are multiple spectrum owners competing for customers.
But for a decade now the only real buyers of spectrum have been AT&T and Verizon (VZ). Their aim is not a free market. Their aim is a duopoly, one in which both sides can take monopoly rents because consumers have no alternative if they want service. It's very much what railroads themselves were doing in the late 19th century, and it was one of the issues that led to the rise of the progressive and populist movements in the 1890s.
The problem is that neither rate regulation, the liberal impulse of that era, nor the creation of "competition," the liberal impulse of the 1990s, is going to solve the problem. The solution is to acknowledge the reality of abundance, to create more unlicensed spectrum (rather than less), and move toward buying back what's been sold and regulating it as the Internet is regulated. Privately, through peering agreements, technical committees, and equipment vendors.
Let AT&T, let Verizon, let Clearwire, Google and Apple create services defined by equipment, regulate based on the capabilities of equipment, and let present gear be made compatible so that, over time, spectrum becomes a utility open to all competitors. Let those who want to offer services "peer" with networks from which they require carriage, as is done on the Internet, so we build infrastructure once and so there's an incentive to build infrastructure when and where it's needed, not everywhere because government demands "universal service."
Meanwhile, investors need to just avoid wireless stocks. Don't buy stock in AT&T, or Verizon. Don't buy Sprint (S) and don't buy Clearwire. Let carriers that can't get a return on their capital collapse, and then deal with the wreckage of a failed regulatory regime.
Disclosure: I am long GOOG.