The full FCC agrees with the Wireless Bureau that FiberTower’s failure to construct resulted from its own business decisions.

FiberTower loses again. The full FCC has backed the Wireless Telecommunications Bureau’s decision to cancel 698 licenses held by the company in the 24 GHz and 39 GHz auctioned fixed microwave bands, for failure to construct sufficient facilities.

As we explained last December in our sister publication FHH Telecom Law, FiberTower is not alone in its difficulties. Nearly all of the area-wide licensees in the four auctioned bands used to communicate between fixed points – at 24, 28, 31, and 39 GHz – have had difficulty in meeting their renewal obligations. In part the problems trace to a shortage of suitable equipment, and in part to markets that did not develop as expected. FiberTower, ironically, was one of the more commercially successful licensees, so the FCC action against it seems particularly harsh.

The usual duration for microwave licenses of all kinds is ten years. When an area-wide licensee applies for renewal after that period, it must show it is providing “substantial service.” The FCC rules define this, unhelpfully, as “a service which is sound, favorable, and substantially above a level of mediocre service which just might minimally warrant renewal during its past license term.” (Confusingly, this says the level of service required for renewal is substantially above the level of service required for renewal.)

Thanks to a “safe harbor” policy, a licensee is deemed to be providing “substantial service” if it demonstrates that it has constructed four links per million population in its service area.

The vast majority of licenses in these bands have been unable, via the safe harbor or otherwise, to make their “substantial service” showings by their expiration dates. Most have requested extensions of time. The FCC has routinely granted such extensions, typically for up to four years; it has then cancelled the licenses of systems that remain unbuilt at the end of the extended period.

FiberTower had received extensions on various of its licenses. As those were running out, FiberTower entered bankruptcy proceedings. In early November, the Bureau canceled 94 of FiberTower’s 24 GHz licenses, and 595 of its 39 GHz licenses.

FiberTower asked the full Commission to review the Bureau’s decision, alleging a number of errors. The Commission responded by striking three of FiberTower’s pleadings and disagreeing with the rest. FiberTower’s failure to build out the 698 licenses in question was not due to factors beyond FiberTower’s control, the Commission ruled, but rather resulted from FiberTower’s own business decisions.

The Commission’s decision puzzles us, as did the Bureau’s earlier. True, the FCC cannot tolerate wholesale flouting of its rules. But when rote enforcement runs counter to the public interest, which we think is the case here, the FCC would better serve the public by looking for ways to be lenient.

One purpose of the auction mechanism (apart from raising funds for the Treasury) is to give licensees skin in the game. Having put up money to buy their spectrum, the theory goes, licensees will have financial incentive to put that spectrum to use in developing a stream of revenue. But sometimes the incentives don’t work as planned. The low demand at 24 GHz may result in part from the availability of “free” point-to-point spectrum in the nearby 23 GHz band. In any event, canceling FiberTower’s licenses now will likely have the unwanted effect of driving down bids for this category of spectrum at any future auctions.

The “four links per million” safe harbor may also operate perversely, by requiring a licensee seeking renewal to construct links that have no market to support them. A licensee is understandably reluctant to build expensive “links to nowhere.” These are likely to be in the wrong locations when business does develop, and the licensee will have to abandon them if it loses the license. Moreover, FiberTower’s pleadings explained that the necessary costs before the construction of links dwarf those of construction itself, yet the licensee receives no credit in the renewal evaluation for having made those expenditures. The Commission brushed that argument aside, holding it to be an untimely effort to change its renewal policy.

What is the purpose of that renewal policy? The last thing anybody wants is for scarce spectrum to go unused. In case the financial incentive of having paid for the spectrum is not enough, the “substantial service” requirement directly pressures licensees into putting their spectrum to work, rather than just sitting on it – the “use it or lose it” principle. It is fair to ask, then, what happens to spectrum, like FiberTower’s, for which the Commission has terminated licenses. The answer: nothing. The spectrum will be unused, and under the Commission’s rules, unusable, until the Commission either gets around to re-auctioning it or changes the rules to allow some other form of licensing. Either will take years – even longer in this case, because the bankruptcy court has enjoined the FCC from transferring the licenses for the time being.

The whole concept of auctioning fixed spectrum has turned out to have inherent flaws. Given the facts on the ground, it seems to us the public interest would have been better served by leaving the licenses with FiberTower. That would at least have offered some promise that the company’s business might expand into the now-underused spectrum. The FCC’s action has now closed out this possibility, and instead brought about the result the FCC most wanted to avoid: spectrum that neither FiberTower nor anyone else can put into service.