One might think from the press coverage that all spectrum auctions are multi-billion dollar affairs like the AWS-1 Auction in ’06 or the 700 MHz Auction in ’08. But these auctions are the exception rather than the rule. More typical are the steady stream of small auctions like Auction 78, which auctioned remaining licenses in the AWS-1 band.
Which brings us to the Wireless Bureau’s Public Notice of the Broadband Radio Service (BRS) auction. Some of us have followed the adventures of the 2.5 GHz band back when it was “Wireless Cable” and the non-commercial licensees used it to offer closed circuit television for what we now call distance learning. These days of course, we know this as the “Broadband Radio Service” (BRS) and the “Educational Broadband Radio Service” (EBRS), and we care about the 2.5 GHz band as the home of Clearwire and the great hope of WiMax.
You might think that the “WiMax” auction would be a big deal — but only if you don’t know the band, its history, and the inventory up for auction. If you know that, you know why this auction is likely to prove as boring but ploddingly necessary as a run for office supplies.
So why do I consider this worth blogging about, other than my sentimental fondness for the band and my general obsession about things spectrum? Because (a) my cause celebre, anonymous bidding, faces its first post-700 MHz challenge, and (b) 2.5 GHz is the home of the major WiMax plays, and what happens in the auction has the potential to shape the field going forward and influence whether deployment goes more smoothly or gets all bollixed up.
More detail below . . . .
As one can see from this list of licenses and reserve prices, this does not promise to be one of your billion dollar blockbuster auctions. Basically, these are the dribs and drabs, snouts and entrails, and other leftover bits after many years of of complicated restructuring and consolidation within the industry. It has value — particularly to the handful of companies trying to make a living in the band. But no one expects the likes of AT&T, Verizon, or MetroPCS to come slug it out to the death, which is what drives up auction revenues. While the reserve price total of approximately $17 million is more than I’ve got in loose change, it is hardly enough to excite comment at a time when “trillion is the new billion.”
Still, several things make this auction worth sitting up and taking notice. The first is that there actually is a WiMax industry now, with equipment vendors, service providers and everything. And while WiMax works in other bands (such as the 3.65 GHz band, another band for which I have a sentimental fondness), the big plays are unfolding in the 2.5 GHz band around Clearwire (particularly as AT&T and Verizon have committed to LTE in 700 MHz and T-Mobile has committed to LTE over in AWS-1). So those who care about the WiMax market need to at least pay attention. There are a number of small and midsize operators who could potentially make good use of these licenses licenses under the circumstances, and watching who shows up, who bids, and who walks away with what will have impact (although hardly groundbreaking impact) on the shape of deployment in the band. It would not surprise me, for example, to see some wireless ISPs (WISPs) currently operating on unlicensed spectrum show up to try to get themselves some licensed spectrum for their coverage area.
More important from my perspective, this represents the first spectrum auction since the 700 MHz auction to put the question of whether or not to use “anonymous bidding” on public notice. For those unfamiliar with anonymous bidding and the fight around it, start here and proceed here then here with big finish here.
Briefly, until the 700 MHz auction, spectrum auctions were “open.” That meant everyone saw who was bidding on what and how much “eligibility” each bidder had (think of eligibility as needed poker chips to bid). At the time, this looked seemed like an efficient way to do things. Over time evidence began to mount up that incumbents used the open nature of the auction to target rivals and silently collude to divide spectrum licenses with a minimum of fuss. Nothing as bad as outright collusion or even explicit bidding signals. It just works out as a matter of math. For example, if I know VZ needs to plug holes in its network, I might target those licenses to drive up the price to VZ. Or, if I want to avoid a fight, I might not try to bid on them. British Economist and auction expert Paul Klemperer provides an excellent anecdote illustrating this in What Really Matters About auction Design. Before the 1995 PCS auction, PacBell made it clear it would fight fiercely for the license in its service area in Los Angeles, where it also enjoyed significant brand recognition and other advantages. The result was that few bidders chose to fight PacBell for the license and it went for a fairly modest fee.
Anonymous bidding eliminates the ability to target or avoid bidders. In anonymous bidding, the FCC reports the value of the winning bid, and how many entities bid on the license. That’s it. (The FCC publishes the full bidding report after the auction, which is utterly critical to ensuring transparency and figuring out what went right or wrong in any auction.)
As one might expect, the entire existing wireless industry (with the exception of VZ, which had been targeted in some auctions) hated anonymous bidding, and fought like hell against it. Kevin Martin first tried to introduce it in the 2006 AWS auction on advice of his then-Chief Economist Leslie Marx. Unfortunately, with the evidence resting on complex mathematical analysis and only yr hmble obdn’t and Greg Rose supporting anonymous bidding in the face of opposition from the entire wireless industry — from big companies to little rural systems — Martin failed to convince a majority of the Commissioners to vote for it. Then came the AWS Auction, with results that exactly matched what Greg had predicted. Armed with the evidence, and backed up with new allies such as Google who wanted to maximize the chance of new entrants winning, we were able to carry the day and get anonymous bidding in place for the 700 MHz band.
In the 700 MHz AT&T and Verizon won a good chunk of the licenses, but had to pay through the nose to get them. Also Echostar, previously blocked in the AWS auction, won a national footprint. For us, this proved two things: (a) anonymous bidding worked as predicted in preventing either targeting or collusion to divide up licenses cheaply; and (b) incumbents would spend money like water to get the licenses they wanted.
So now we roll into the PN for the 2.5 GHz band auction. With Kevin Martin gone, it’s unclear whether anonymous bidding will survive. While some folks in the Wireless Bureau love it and think it is an important rule for making auctions more efficient, I know from experience that other folks in the Wireless Bureau hate it. With the 8th floor operating with only 3 Commissioners, one of whom is waiting to move on, and everyone focused on the upcoming DTV Transition on June 12, this item is unlikely to attract a lot of attention. Even if the final decision happens after Genechowski and newly nominated 3rd Democrat Mingon Clyburne come on board, along with whatever Republican they finally settle on, the majority of Commissioners will be coming to a very complex issue cold.
In this environment, it is not at all unlikely that the wireless industry will be able to persuade folks that anonymous bidding was just another whacky idea of the despised Kevin Martin, best swept out the door and forgotten. OTOH, it is unclear if the wireless industry will show up to make such a push, particularly given the rather low stakes and minimal interest in this auction. If Clearwire ends up dominating the auction, a likely outcome no matter what rules are used, the incumbent’s best argument against anonymous bidding evaporates with little gain to them.
If the Bureau does end up eliminating anonymous bidding, we can expect Clearwire to get targeted by rivals to drive up their costs — as Verizon was in the past. In this auction, that won’t have a particularly dramatic result unless the added cost impacts Clearwire’s cost of deployment. Other forms of mischief might also emerge (I will leave it to Greg to do the more sophisticated analysis).
I’m not sure if these reasons convince anyone else to keep an eye on things. But to my fellow spectrum junkies I say —
Stay tuned . . . .
Harold, good piece.