Federal Spectrum Opportunities

At last Wednesday’s Senate Commerce, Science and Transportation hearing on wireless spectrum, senators and witnesses alike expressed a general desire to “free up” more spectrum for wireless broadband. Sen. Bill Nelson (D-FL) said, “Spectrum legislation is not only necessary, but it has traditionally been bipartisan.” Along the same line, Sen. Brian Shatz (D-HI) said, “there is a real opportunity for bipartisan consensus” on spectrum legislation. It’s true, spectrum policy, although often difficult and complex, is rarely mired in partisan disagreement. Growing demand for additional wireless capacity for streaming video, Internet of Things (IoT), and machine-to-machine communications makes clear that relatively low-cost opportunities to repurpose spectrum are no-brainers we can all get behind.

However, a bit further down in the weeds there was a point of disagreement among the witnesses that is worth teasing apart. It has to do with mechanisms to seek out inefficient uses of spectrum by the federal government. There was general agreement on the panel that federal users will be a significant source of spectrum in the future, but not exactly consensus on the particular mechanisms to repurpose that spectrum.

There has been a lot of talk about federal spectrum because it is a clear source of bandwidth for mobile broadband and IoT applications. The 3.5 GHz band that the FCC is working to open up to small cell use is a good illustration. The 3.5 GHz band has been allocated to a number of services, but one of the largest incumbents are Navy radars. These radars are only used intermittently by ships on the coast. The thinking goes, why claim an entire band, for the entire country, all of the time, when the spectrum is only used some of the time on the coasts—let’s allow users to share that spectrum where it wouldn’t interfere with the radars. Whether such complex sharing schemes as those proposed for the 3.5 GHz band will work is subject for another day (though there is some indication they don't quite get the economics right), the point is simply that there is plenty of federal spectrum that lies fallow much of the time. Even more spectrum may be freed up by upgrading outdated systems or transitioning purpose-built government tools using dedicated spectrum to more general use architectures using commercial spectrum.

The key question is how we discover what can be repurposed to mobile broadband and at what cost. The current top-down model, whereby particularly obvious opportunities are identified, and pressure from the president applied until agencies might, reluctantly, work with industry and the National Telecommunications and Information Administration (NTIA), is not up to the task of efficient, cost-effective reallocation of spectrum to new uses that are more valuable to society. It is time for more fundamental reforms to ease this process, and there are a wide variety of tools being discussed.

The witnesses at the hearing expressed general support for introduction of market-based incentives for federal spectrum use. Tom Lenard (and ITIF) has long supported a sort of General Services Administration (GSA) for federal spectrum, whereby government users would pay fees that approximate the opportunity cost of the spectrum to a central agency, similar to how agencies pay rent to the GSA for use of its buildings—others agree this is worth adding to the toolkit. Pierre de Vries of the Silicon Flatirons Center presented a set of reforms to help decentralize spectrum management functions in the face of the need to pack diverse radio services ever closer together, reforms that would assist in clarifying rights in sharing scenarios.

But there was a slight difference among the panelists on the viability of utilizing incentive auction mechanisms to encourage repurposing of federal spectrum. The basic idea here is an incentive auction for the federal government, allowing agencies to share in the proceeds of an auction beyond just their relocation costs. FCC Commissioner Rosenworcel expressed support for the idea, saying that to build “a robust and reliable spectrum pipeline, we need to make sure that federal authorities see gain—and not just loss—when their airwaves are reallocated for new mobile broadband use.... We could begin by expanding incentive auctions to federal spectrum users.” Commissioner Rosenworcel tempered her recommendation, noting that annual budgets and lack of market signals make federal incentive mechanisms difficult. But this idea has been getting some traction, with matching bills in both House and Senate (though the proposals face challenges due to the way CBO scores auction revenue).

Blair Levin took a different view, saying that an incentive auction for the federal government faces several barriers. Information asymmetries and principal-agent problems are difficult for a constrained NTIA to overcome, especially given that government workers tend to eschew risk. Furthermore, no single federal agency has exclusive access to the bands allocated for its use. The incentive auction slated for next year for TV spectrum is complex, but at least we are dealing with a limited number of users and architectures in any given market. Levin also worries that creating property rights in a strategic resource would result in spectrum hoarding. On top of all this is the obvious concern that any incentive payment could be offset though future budgetary processes.

Here Levin’s principal interlocutor is his Brookings colleague, Dorothy Robyn. Robyn is returning to spectrum policy after serving as Commissioner at the Public Buildings Service of the GSA. Her experience with federal real property leads her to argue for NTIA auction authority, stating that “a major lesson for spectrum policy is that the ability to retain proceeds from the disposal of property is a key motivator for federal agencies if the incentive is properly structured.” This point, “if the incentive is properly structured,” is key for Robyn. How the incentives are structured and who exactly sees the gain matters.

CTIA’s Tom Power has suggested that what we need is not a full-blown incentive auction, but instead an auction for the right for private parties to negotiate with federal spectrum users. This approach would have several advantages: it narrows the number of people discussing often sensitive federal systems, allowing the parties to get into the gritty engineering details of possible relocation, sharing, or simply upgrading old government systems. Levin characterized this approach as “private sector bounty hunters.”

It is not immediately clear which path is the right one—are full-scale federal incentive auctions unworkable, or just in need of the right design? Would we be better off with smaller, specific negotiations, perhaps enhanced by clarified spectrum use rights? Would the relatively straightforward spectrum fees model be sufficient?

For practical purposes the needed congressional actions for both federal incentive auctions and narrower, right-to-negotiate auctions would begin in the same place. Either of these mechanisms would require reform to the Miscellaneous Receipts Act and the Antideficiency Act—a pair of laws that prevent payments directly to government agencies. The GSA-style spectrum fees, while still requiring congressional action, may be more straightforward to implement.

In the short term, the classic solution—identifying specific bands and mandating an auction—remains imminently viable to keep the spectrum pipeline flowing.