Genachowski's Secret $15 bn Piggy Bank, or T-mobile Triumphs Over M2Z.

I’ve been rather pressed for time, hence have not had much chance to blog on the FCC’s recent spectrum policy announcements for D-Block and the broadcast migration offer. Combine these two speeches with Genachowski’s recent statement in an interview that the NBP will finance the $25 billion via existing programs and it is clear that the FCC will adopt the T-Mobile’s “asymmetric auction” proposal for the AWS-2 and AWS-3 band, leaving M2Z high and dry. The only question is whether or not there will be spectrum caps to keep AT&T and Verizon from snarfing the good stuff, but do not expect the NBP to touch something as “controversial” as spectrum caps even by veiled implication the way the DoJ did in its comments.

Mind, this is another example of the “spectrum auctions are the crack cocaine of public policy” problem. The thirst for revenue pushes all other considerations out the window. I’m not convinced the T-Mobile approach is wrong (especially if subject to spectrum caps), and I think the D-Block finesse was extremely clever. But when revenue sits in the driver’s seat, policy invariably takes a wrong turn somewhere along the road. But it is difficult to imagine how Genachowski could resist a $15 bn secret cash cow to fend off accusations that Democrats are once again writing checks against our children’s future blah blah blah.

I unpack all this below. . . .

Given how long and complicated this is, a little background is in order.


You can find my much longer version of “what is M2Z” from back when it first came up in public policy in 2007 here. Back a ways ago, Congess by law cleared our a bunch of spectrum for “advanced wireless services” (AWS). There were three bands, so of course they were named “AWS-1,” “AWS-2,” and “AWS-3.” The FCC auctioned the most valuable band, the AWS-1 band, back in August 2006. The AWS-2 band, a measely 10 MHz divided into two 5 MHz paired bands, was to be auctioned as soon as anyone got around to it. But the AWS-3 band, a whopping 20 MHz of spectrum, presented a problem.

Why? Because cellular systems rely on using “paired” bands of spectrum and not one big band. This goes back to the limits of 1970s and 1980s technology, when cellular got developed. The old walkie talkies could only let one person talk and another receive, or the two people trying to talk/transmit with each other would interfere with each other. (This is why people had to say “over” on those old tv shows showing people using old walkie talkies or CBs. You couldn’t talk until the other person stop pressing the “talk” button, and you only knew that if the person told you “I’m done talking.”)

Cellular solved this by having two bands, an up band for the cell phone to speak to the tower, and a down band for the tower to speak to the cell phone. These bands needed a certain amount of “space” between them. So the FCC would allocate spectrum for cellular service in pairs. Although cell systems and wireless technology have evolvede enormously, they still have the same “paired” spectrum architecture built into their underlying structure. So all auctions intended for cellular service, until relatively recently, auctioned paired spectrum. AWS-3 was a lot of spectrum, but it was not easy to pair it with anything. It could have been opened to unlicensed, but Congress had mandated that the FCC auction spectrum. So the AWS-3 band just sat there, forlorn, waiting for someone to come along and put it to its highest best use.

Enter M2Z, a Silicon Valley start up. They said “give us AWS-3 for free and we will give you a free nation-wide wireless broadband service, supported by adds and a pay tier.” M2Z, structured as a wireless internet access provider rather than a cellular provider, would take advantage of better technologies that allow people to send and receive on the same band. So M2Z had the powerful argument that rather than let 20 MHz of usable spectrum lie fallow because it didn’t fit the architecture of the cellular carriers, the FCC should give it (later changed to ‘auction it’) to folks with better technology/network architecture who could put it to use.

I won’t get into everything that made M2Z attractive in some ways and controversial in others. Suffice it to say for this exercise that the FCC rejected M2Z’s proposal in 2007 and started a rulemaking to set auction rules. A year of lobbying later, and M2Z persuaded then FCC Chair Kevin Martin that M2Z offered the best way to get universal broadband (for the then definition of broadband of 768 kbps), so he issued a public notice on a revised M2Z plan. That provoked strong resistance from the wireless carriers, most passionately T-Mobile.

Leaving aside the “of course T-Mobile and all greedy evil wireless incumbents oppose free broadband,” T-Mobile (and others, but primarily T-Mobile) had two big concerns. First, T-Mobile held spectrum right next to the AWS-3 band. The model phones they used (based on the European band plan) would likely experience interference if M2Z went live — which M2Z pointed out was really not their problem (T-Mobile countered they would experience interference anyway, even if they built phones designed for the American spectrum band plan). But in addition, T-Mobile desperately needs more spectrum. It made huge investments in the AWS-1 band, so it has a network that could absorb the AWS-2 and AWS-3 bands fairly easily. It has had covetous eyes on the AWS band for some time, and was not going to let it go without a fight. On the policy side, T-Mobile (rightly) pointed out that they are one of only two other wireless companies in a position to compete with AT&T and Verizon on a national basis, so if the FCC really cared about cellular competition they would sooo give the spectrum to T-Mobile (by auctioning it) and not some unknown start up.

But T-Mobile still had the “pairing problem.” As long as the choice was “M2Z or let it sit there,” M2Z had a fairly strong case. So T-Mobile proposed something new in wireless, asymmetric band auctions. Rather than trying to auction spectrum in symmetric bands, based on the idea that voice uses the same bandwidth in both directions, T-Mobile proposed auctioning the AWS-2 and AWS-3 bands as a single, asymmetric pair with a mucking big 25 MHz block (the other half of the AWS-2 band and the 20 MHz AWS-3 band) for downloads and a tiny 5 MHz block (what’s left of the AWS-2) for uploads. Given the growing importance of wireless broadband access, and the near-universal (and wrong in my opinion) assumption that users want much more download than upload, going to asymmetric bands for cellular auction became attractive.


So after Kevin Martin left, the AWS-2/AWS-3 band languished as we shifted into “all National Broadband Plan all the time” mode. And, being a relatively obscure proceeding, it remains languishing although M2Z and T-Mobile continue to snipe about it in all relevant dockets.

Now look at the two major items Genachowski has rolled out on spectrum. First was Genachowski’s mobile broadband speech at New America Foundation. When I first heard it, I was pretty goddamn horrified. It sounded like something left in Michael Powell’s desk drawer from 2001 before he got interested in unlicensed. “Blah blah auctions wonderful, blah blah puts spectrum in the hands of those who use it most efficiently blah blah, ALL HAIL THE GLORIOUS GODS OF THE MARKETPLACE — WHO DELIVER UNTO US THE BOUNTIES OF THE EARTH IF WE ONLY HEED THEIR WORDS AND USE MARKET MECHANISMS FOR EVERYTHING. And, oh yeah, we’ll do some opportunistic sharing stuff like white spaces.”

This wasn’t just at odds with where we all expected things to go (the speech was at New America Foundation for God’s sake! They are the biggest boosters of unlicensed/opportunistic sharing in our community. This would be like giving your roll out of the “public option” at the Chamber of Commerce.) It was utterly at odds with everything that had gone before from the entire Obama Administration. Not that Genachowski contradicted anything anyone has said. But until now the Obama people have put opportunistic sharing (aka “unlicensed spectrum”) on the same level as getting out more licensed spectrum; they’ve projected that they understand each is a valuable tool for promoting wireless access. Genachowski was a throwback to the Bush years which treated auctions as the big deal and unlicensed as a footnote to be addressed later, after the important stuff got done.

Another thing that tripped my “WTF” meter was the notion of auctioning the broadcast licenses as part of the voluntary buy-out scheme. Under the plan, broadcasters will voluntarily surrender licenses, the feds will auction them, and the broadcasters will get some percentage. That seems rather cumbersome if the object is to get spectrum to wireless carriers. And it raises some substantial legal problems, since federal law requires that all revenues from spectrum auctions go into the Treasury. So why not just allow broadcast licensees to sell licenses to wireless carriers directly? Nothing stops the FCC from allowing conversion from a “broadcast service” license to a “broadband service” license as part of approving the transaction. So why go through the rigmarole of a legally questionable auction when you could authorize broadcasters to sell to wireless companies directly? Answer: Because auctions generate federal revenue, whereas allowing broadcasters to cash out to wireless carriers directly does not.

The next piece of the puzzle was Genachowski’s public safety speech. No surprise, what grabbed attention was the decision to auction the “D Block” and request $12B for public safety build out. To briefly sum up: “D Block” was the effort to take 10 MHz of the extremely valuable 700 MHz spectrum and create a public/private partnership to build out the public safety network. It failed, leaving the FCC a very unpalatable choice among bad alternatives. While I think Genachowski has come up with a very clever way to finesse the “D Block Trap” by requiring all 700 MHz licensees to give public safety access to the entire commercial band at need, it did not escape my attention that the FCC proposed to offset the public safety network build out cost with revenues from the D Block auction.

Then came the story from Amy Schatz that the total cost will be $25 billion, with Genachowski’s assurance that the FCC will pay for it out of existing programs. That’s nice to say, but as others have noted, even if you assume you can divert some Universal Service Funds (USF) and make some generous assumptions on what the voluntary broadcast buyout program will bring in, the numbers don’t add up. So where can Genachowski find the remaining money?

Which brings us back, full circle, to the AWS-2/AWS-3 T-Mobile v. M2Z fight. The AWS-1 auction raised about $13 billion, and there is good reason to believe that the price paid in that auction were artificially depressed by some rule defects that we got changed in the 700 MHz auction. I expect that the FCC could claim with fairly strong support that an AWS-2/AWS-3 asymmetric auction could raise between $15-20 billion.

Which is why, once I saw the Amy Schatz piece, I guessed T-Mobile will win.

So Is This A Good Thing Or A Bad thing.

I’ve always been ambivalent about M2Z. There is a lot I like, but a number of concerns — including whether they would make good on their commitments or if they would pull the usual “promise anything to get a license then explain why you can’t really do that” scam. So while I generally don’t like distributing spectrum by auction, and liked a lot of things M2Z promised, I’m not heartbroken to see them lose out. And, on the flip side, I think there is value in strengthening the competitive position of carriers who are not AT&T or Verizon (no offense guys, but you are dominant and all). Dumping another 10 MHz of 700 MHz spectrum and 25/5 MHz of AWS spectrum into the CMRS market could do some good in this regard — unless AT&T and Verizon snarf it.

Which brings us to the next question. Is the FCC actually going to do anything to promote competition in its auction rules. In the past a bunch of us in the public interest community have pushed for “spectrum caps.” For the uninitiated, “spectrum caps” are a limit to how much spectrum an entity can hold in a given license area. The Republicans eliminated these a bit less than ten years ago because there was so much gosh darn competition spectrum caps seemed silly, whereupon the industry promptly collapsed in on itself — an outcome entirely predictable to anyone who studies actual economics as opposed to that religious pseudo-economics they peddle at U of C. As a result, we have a “wildly competitive” market of two dominant vertically integrated firms, two also rans, some struggling small fry, and a bunch of rural guys in areas with low rates of return. Spectrum caps could (modestly) help some of the also rans (I am not so wildly optimistic to believe that a genuine new entrant will show up at this late date) and provide needed capacity for some of the non-cellular users of spectrum (who never get considered in this, because U.S. spectrum policy is now entirely driven by cellular — we only notice WISPs and WiMAX when we want to pretend things are more competitive).

But Democrats are, and I say this with the deepest respect, “averse to conflict.” Furthermore, I have seen traditional Jewish Weddings that took less time, effort, and angst to prepare than the National Broadband Plan. This is Julius Genachowski’s and Blair Levin’s Most Very Special Day. And they are not going to have their Most Very Special Day tarnished by putting in all manner of controversial things like spectrum caps. This is going to be about accentuating the positive and promoting competition by making more spectrum available to everyone and now is NOT the time to be marring this Very Special Day by bringing up the nasty fight about whether spectrum caps decrease or increase revenue.

(But check back a week later when we move to public notice on the auction rules.)

Stay tuned . . . .

One Comment

  1. Harold, I enjoyed your discussion of the various spectrum issues raised by the National Broadband plan. However, I disagree with a couple of your assertions – particularly your conclusion that the abolition of spectrum caps has driven current market results. I’ve explored this in more depth in a response on my blog: http://www.bitsonbroadband…. At minimum I owe you an opportunity to respond, and as always, would appreciate your thoughts.

Comments are closed