To review: As a baseline policy, I argue for allocating the digital TV band – 49 channels, 6 MHz each, or 294 MHz of VHF/UHF spectrum – to seven “overlay” licenses. Each of the 7 overlays would cede exclusive control over 42 MHz to the licensee, subject to encumbrances. Namely, existing DTV stations would be grandfathered, allowed to continue broadcasting. But they could be paid to accept interference, or to change channels. This would allow the stations to efficiently cluster, while providing the economic structure – bargaining with new overlay licensees – to actually get this crucial reallocation achieved.

Were broadcasting signals to bunch up on adjacent bands the “white space” would be made far broader in scope and more productive in accommodating new deployments. Simple moves, while protecting over-the-air stations, could unleash two-thirds or more of the DTV band for entirely new services. More radically, the existing stations could choose to vacate broadcasting for cable/satellite/broadband distribution, freeing up the entire band. While this is the policy that we should be pursuing – as I outlined in this 2001 paper published by the AEI-Brookings Joint Center on Regulatory Studies – it requires congressional legislation and is therefore a much more ambitious policy.

Stacey’s objects to the policy on the grounds that the new licenses would simply be purchased by the existing cellular carriers – AT&T, Verizon, Sprint, and T-Mobile – and the spectrum warehoused. This is demonstrably incorrect. First, a simple prophylactic policy would limit the licenses to one per customer, ensuring space for entrants. Second, were the assertion about warehousing correct, operators would long been engaging in the practice. They do not. The 190 MHz of licensed cellular spectrum available to U.S. carriers in 2005 was intensely utilized, not parked. Indeed, high-speed data networks were layered on top the bandwidth used to supply voice services via licenses issued in the 1980s and 1990s. The additional 90 MHz of licensed AWS bandwidth auctioned in 2006 found T-Mobile, the smallest of the four national carriers, the high bidder. (The largest incumbent, Cingular/AT&T, was a non-factor.) T-Mobile used their licenses to build a $2.7 billion 3G network – just made operational and now hosting Google’s Android platform. Third, were mobile carriers to profit by restricting supply, they would not be investing $20 billion per year, give or take, in new infrastructure. That defeats the whole purpose of the purported warehousing strategy. The capex creates new capacity via cell splitting and technology upgrades.

In fact, the profitable way to “warehouse” spectrum is not to buy it, but to have regulators restrict access to it for competitive products. Perversely, that is exactly what is happening with the FCC’s “white spaces” proceeding. Mobile carriers do not fear losing market share to local area unlicensed applications, which generate approximately zero service revenue and are rarely viewed by subscribers as cell-phone substitutes.

The unlicensed white spaces plan is a debacle in the making. This is already painfully apparent. The FCC is asking precisely the wrong questions – wholly focused on assuring that new low-power devices will not interfere with the existing configuration of broadcast TV stations – and ignoring the big picture: how to get past the TV Allocation Table of 1952. Only then can society enjoy the bountiful benefits of 294 MHz of hugely productive radio spectrum. The issuance of disparate, small, non-exclusive rights to use devices approved by regulators offers no hope of making this essential transition. The alternative mechanism, assigning broad ownership rights to competitive overlay licensees, does.