I have been doing some analysis of Verizon’s latest move in the VZ/SpectrumCo transaction, the announcement VZ will engage in a series of AWS spectrum swaps with T-Mobile. Between this and Verizon’s commitment to sell off its Lower 700 MHz A&B block licenses, I am almost happy — except for that whole cartel thing with the major cable companies. But if we ignored the cartel thing, this deal now becomes the rare bird that actually enhances both Verizon’s position in the market and that of a prospective competitor. This is not quite a triumph for Coasian market efficiency, since it took the threat of agency action to nudge Verizon in the ‘right’ direction. I also need to point out that when we start out with a fairly dismal market structure, it does not take much to improve things. Giving spectrum to T-Mo is good, but it does not address all the competition problems created by our unfortunate means of distributing spectrum, which still ends up concentrating it in the hands of a very small number (i.e. 2) of companies. So ‘happy’ is a relative term.
As a result, the transaction needs a few minor conditions to make it complete: a data roaming condition to keep competition afloat (and in case the data roaming rule does not hold up in court) and accelerated build out/use or share to ensure rural communities see a 4G network before the end of the decade, but otherwise this looks pretty good (even with AT&T likely to snarf all the Lower 700 MHz B block licenses). Mind you, it reenforces the need to get interoperability and special access reform done if we want to see real competition — but we have rulemaking proceedings on those.
Unfortunately, there is that whole “cartel” thing with the major cable operators. And despite all the positive aspects of this transaction as now configured, they cannot outweigh the negative of creating an anti-competitive cartel at the center of our communications infrastructure. But let me set that aside for the moment to focus on the spectrum side of things.
More below . . . .
Allow me to start by saying that Verizon clearly took very careful notes of everything that went wrong for AT&T/T-Mobile. VZ has handled the unexpected hiccup of concern on spectrum concentration masterfully, addressing concerns before they could harden into serious resistance while simultaneouslyimproving its overall condition. I suspect this owes much to the fact that VZ’s Chief Executive, Lowell McAdam, understands that a strong wireless network depends not merely on the amount of spectrum but also underlying network architecture.
The Verizon/T-Mobile Spectrum Swap.
The next thing to note is that this is not a sale of spectrum from Verizon to T-Mo. It is a series of complicated swaps, which includes a set of “partitions” of existing licenses where the coverage area of the license does not match the territories where the parties would like to swap spectrum. In a bunch of regions, notably in the west, T-Mobile is actually giving spectrum to Verizon. In other regions, T-Mobile comes out with more spectrum. And in a bunch of places, both companies end up with the same amount of aggregate spectrum, but it is reconfigured so that they have contiguous blocks that work better with their national networks.
This is what makes it so hard for people used to thinking of spectrum as a “thing” with an intrinsic value to understand what is going on here and how both parties can come out ahead. Spectrum is not a “thing,” and licenses don’t represent something real. Licenses are about an exclusive right to emit energy on specific frequencies in a particular way. The value of that right depends on a lot of factors. While some of those factors are intrinsic to the physics of the frequency (e.g., the fact that 700 MHz spectrum acts in a particular way makes it more valuable for mobile broadband), others are not. They have more to do with things like how well it works with your existing network architecture or who in the world builds equipment for this so you can achieve economies of scale.
This is a prime example of what Coase actually meant (as opposed to the pseudo-religious garbage spouted by too many neo-Coasians in Policyland) when he talked about how market forces, over time, given the right circumstances, can improve efficiency of use. T-Mobile and Verizon Wireless have each improved their overall efficiency through a bunch of swaps. This result could not have been achieved at the time of the auction, because the technology that makes this configuration more valuable and efficient overall did not exist at the time of the auction. So the auction did not place spectrum “in the hands of those who would put it to its highest best use,” as the neo-Coasians always chant when we talk about spectrum policy. To the contrary, the allocation by auction was astoundingly inefficient, as demonstrated by SpectrumCo’s utter failure to develop it and Verizon, T-Mo and others who could use it not getting it. But, over time, parties positioned to make better use of the spectrum — given proper market conditions — acquire it and put it to more productive uses. Which is what Coase actually said would happen, on average, under the right conditions (neo-Coasians also leave off these rather critical caveats).
OK, But How Does All This Serve The Public Interest?
As I’ve noted before, the FCC has a legal obligation to ensure that these transactions serve the public interest, which means more than just enhancing efficiency overall. The FCC also needs to ensure there is enough competition in the market to force the remaining carriers to pass on these efficiencies to their customers. So let us consider the outcome of these swaps, and Verizon’s commitment to sell off its Lower 700 MHz A&B licenses, and see whether this actually benefits the public rather than just the participants.
What Verizon Gets. Assuming this gets approved (which is the high percentage bet at the moment, although I think some additional conditions are still reasonably likely), Verizon Wireless will be positioned to have, hands down, the best 4G wireless network in the United States, and to be able to deploy it well ahead of competitors. It combines the national Upper 700 MHz C Block as a huge data backbone, with large paired AWS spectrum blocks ideally suited to LTE.
Even better, this spectrum is essentially undeveloped greenfield. Whereas everyone else must go through an expensive and complicated “refarming” process of shifting spectrum from current 2G and 3G networks, Verizon can immediately deploy 4G on the new spectrum, then refarm its existing spectrum at leisure, and with the benefit of learning from everyone else’s experience. In addition, as I noted recently, everyone else is going to need to cobble together multiple bands to meet the need for sufficient spectrum (although the VZ/T-Mobile deal points to a possible solution to this, as I explain below). That requires either new multiband technologies or more expensive (and heavier) phones with more chip slots and more shielding. Either way, the need to have multiband devices imposes additional delay and expense. But Verizon does not need multiband. It has sufficiently large block sizes to meet the need (at least in the short term). As an added bonus, the AWS bands Verizon is using are globally harmonized for AWS, so it gets the benefit of signifcant economies of scale and already developed technologies.
As a result of all this, Verizon can deploy its LTE network much more cheaply, and in a much shorter time, than any of its rivals — including AT&T. Obviously a huge win.
But didn’t Verizon lose from needing to make these sales? Not really. As I pointed out in my extensive analysis of Verizon’s announcement it would sell its Lower 700 MHz A&B licenses, those licenses don’t fit Verizon’s network architecture. In fact, it was already in the process of selling them off and swapping them for AWS licenses before it announced the SpectrumCo deal. As noted above, the deal with T-Mobile is not a sale/divestiture but a series of swaps that actually makes Verizon’s overall national architecture stronger.
Keep in mind that when weighing all this, we want companies to deploy really good networks. That’s what makes it possible for us to generate economic activity and do cool things on mobile networks (unless they blow your data cap, but that’s the subject of another post). So the fact that all this allows Verizon to deploy a stronger LTE network faster is, in fact, a public interest positive. But the analysis doesn’t stop there. Remember, we need competition to make sure Verizon actually passes these efficiencies on to customers, rather than simply keeps them in the form of monopoly rents. So given my previous concern about Verizon getting so far ahead no one can compete effectively, what
What competitors get. Last year, during the AT&T/T-Mobile deal, all the talk was about how T-Mobile was doomed to eventual obsolescence because it had no path to LTE. Now it has a pretty clear one, although it requires refarming some of its existing spectrum.
T-Mo’s LTE network to be is neither as powerful or as easy to deploy as Verizon’s. Unlike Verizon, which did not need to tap into its AWS spectrum until recently, T-Mo desperately needed to get that stuff online as soon as it could. Much of the spectrum it has is therefore already in use for customers using a mix of 3G and HSPA+ devices. T-Mo therefore needs to refarm this spectrum to move out of 3G/GSM and onto LTE. T-Mo also lacks the kind of supercharged wireless backbone that Verizon has with the Upper 700 MHz C Block. On the plus side, however, it now has a nice set of large AWS paired blocks that will allow it to offer LTE without resorting to multibanding. Ultimately, T-Mo will have a mixed LTE/HSPA+ network that will allow it to provide excellent service. On the downside, if it began to experience serious growth (rather than just stopping churn and perhaps some modest growth), it would begin to see serious spectrum constraints again. So while it is a stronger competitor, there is still a limit to how much it can challenge Verizon for dominance.
If I were an analyst, I would also point out that T-Mo needs to deal with its stingy, grumpy parent Deutsche Telekom (DT). Will DT agree to actually invest the money T-Mo needs for this kind of build out? I think yes. True last year DT kept insisting it wanted to exit the U.S. market. But life is very different for DT now than it was a year ago. Last year, the U.S. looked like a declining market whereas Europe and Asia seemed like opportunities for growth. Now, T-Mo has a future, and the U.S. economy, while not growing great guns, is looking a heck of a lot more attractive than Europe.
So while this transaction does not give T-Mo the capacity to go toe-to-toe with Verizon, it does put T-Mo back in the game. Given T-Mo’s “challenger” strategy, that certainly helps put competitive pressure on the big two. Not enough to solve the evils of the world, perhaps, and maybe not enough to overcome the huge advantage Verizon derives. But getting T-Mo back as a serious competitor in the “Big 4” certainly helps mitigate the concern about Verizon Wireless getting too far ahead.
As for the Lower 700 MHz A&B blocks, assuming the FCC makes Verizon’s commitment on this mandatory, and does nothing to prevent AT&T from acquiring the licenses, we can expect AT&T to go after Verizon’s Lower 700 MHz B licenses. We can expect U.S. Cellular to make a big play for the Lower 700 MHz A licenses — despite their interference problems from broadcast Channel 51. (U.S. Cellular has recently been trying to cobble together a larger footprint using Lower 700 MHz A.) This transaction with Cox, where AT&T is buying the Lower 700 MHz B licenses and U.S. Cellular is buying the Lower 700 MHZ A licenses, is probably a pretty good preview of the outcome. Mind you, Verizon claims it has 36 potential bidders for the Lower 700 MHz licenses already lined up, so others might pick up some of the licenses as well.
I’ve written a fair amount about the Lower 700 MHz licenses here, here and here. On their own, they don’t do much to move the needle on national competition — especially if AT&T gets to firm up its already substantial Lower 700 MHz B holding. There is also the problem that there isn’t a lot of equipment out there for Lower 700 MHz A. Nevertheless, when weighing up the potential public interest benefits, the fact that Verizon is going to make a bunch of Lower 700 MHz licenses available to others rather than sitting on them counts as a benefit.
So How Does It All Balance Out On The Public Interest Scale?
To stress again, none of these benefits even begins to offset the harm to the public interest of allowing the Joint Marketing Agreements (JMAs) and the Joint Operating Entity (JOE) to go forward. Since 99% of the reason I say that is covered by the Super Secret Confidential Order, I can’t explain why very well. And given how long this post is already, I will need to try to explain that in a separate post.
So does the lift to T-Mo and other competitors keep Verizon from jumping so far ahead that competition is basically dead in a year when Verizon deploys? IMO, not quite — but much of that has to do with general competitive problems outside the scope of the transaction. That doesn’t mean you ignore them — the world is what it is and you have to judge a transaction based on ‘what will the reality be like after the deals close.’ In this case, the fact that the lack of interoperability in the Lower 700 MHz band and the high cost of special access are realities that are going to hobble competitors (other than AT&T) as they scramble to keep up with Verizon. But the solution to those issues mostly rests in other proceedings. The only relevant condition here has to do with the JMAs and the JOE and how they impact the incentive/ability of the cable operators to offer competing WiFi offload, WiFi roaming, and back haul services. The other two need to be resolved through pending proceedings (although it is entirely possible that, as part of the negotiation around a final order, the Chairman’s office might commit to resolving both the interoperability proceeding and the special access proceeding sometime soon).
There are basically 3 conditions that I would say are still needed here (outside of conditions related to the JMAs/JOE), and they are the same three we asked for on the spectrum side back in the Public Knowledge Petition to Deny. First, we need a data roaming condition to ensure that everyone else can survive while they are refarming and otherwise trying to build out an LTE network. Opponents usually piss and moan about how this “punishes success” and so forth. Nonsense. Companies that roam pay handsomely for the privilege. But for companies like Verizon and AT&T, which are fairly self-sufficient on spectrum (or can turn to each other if they must), the value of collecting money from their rivals is offset against enhancing churn on their rival’s system by keeping their rivals spectrum starved. As rational actors, VZ and AT&T prefer customers to get fed up with Sprint and switch to VZ or AT&T rather than charge Sprint a pretty penny to have access to the spectrum it needs to keep customers. But from a competition perspective, we would rather see Sprint remain a real competitor. Hence the whole “public interest” thing.
True, there is a data roaming rule. It is on appeal to the D.C. Circuit. Given the way the D.C. Circuit works, having a roaming condition is still necessary. If the rule survives, then no harm done. If not, the advantages conferred on Verizon (and to some degree T-Mobile) by this set of transactions makes data roaming both extremely urgent and merger specific.
Second, the FCC needs to accelerate the build out deadlines for the AWS licenses. Right now, the licensees only need to provide “substantial service” within the majority of the license territory by ten years after the AWS spectrum was finally cleared of federal users — which in some cases means even this pathetic build out requirement doesn’t need to be met until 2021. We cannot have this justified on the basis of Verizon building a super-cool 4G LTE network while leaving less profitable rural areas behind. Well, we can, but we shouldn’t.
As always, the question is: “what happens if they don’t meet the build out obligations?” We have pushed for something called “use or share,” which means that any unbuilt portions become available to TV white spaces devices. When the licensee actually builds out, they tell the database manager and the spectrum comes out of the database. This has a lot of advantages in terms of stimulating smart radio/multiband technology and making unused spectrum available to those who want to use it. And if the licensee ever decides to build out, they can do so with no additional cost imposed by the condition.
So What Happens Next?
The FCC is taking comment on both the VZ/T-Mo applications and on how these proposed transfers impact the VZ/SpectrumCo transaction. While treating these as separate transactions, expect the FCC to resolve them at the same time.
At this point, the FCC is looking to finish before the end of the 180-day voluntary shot clock that Bill Kennard invented to get Republicans in Congress off his back when the Baby Bells were reconsolidating. This would put a final order around mid-August. This might take less time if the FCC decides to ignore the agreements completely, or it might take more time depending on whether either the FCC or DoJ presses for conditions on the agreements. Arguably, the FCC might do the license transfer while the DoJ holds open an investigation into the agreement. Or the FCC might wait on DoJ.
All of this is rather murky, and I suspect it will remain so until we get closer to the 180-day mid-August deadline. Institutionally, FCC and DoJ have worked closely together and have been unwilling to move at cross-purposes with each other. As a practical matter, it is much more difficult for the DoJ to bring a case if the FCC has given it the seal of approval by transferring the licenses — even if the FCC loudly and ostensibly declines to make any ruling on the contracts (which would present its own problems — but as I say, I am reserving that for another post). The FCC would, I suspect, be loath to undercut DoJ in such a fashion. On the other hand, the FCC (and Genachowski personally) have a strong interest in showing their Congressional critics that they can get things done on time. So the FCC would no doubt like DoJ to hurry things along, and may well decide that being able to issue an Order on Day 180 outweighs outweighs other considerations.
So despite rumors of all sorts flying around, I would not expect any sort of official decision on whether the FCC will approve the transfer while the DoJ is still considering the transfer and the agreements for another couple of weeks. Given that the FCC received comments on the impact of VZ/T-Mo on VZ/SpectrumCo on July 10, I would not expect an Order to reach the Chairman’s office for another week after that at least. Then the Chairman would need to circulate it to the full Commission, which would expect to take several weeks to consider it (this is, after all, a very large and complex transaction and the Commissioners can’t really have solid opinions until they see actual language circulated).
Gonna be a hot summer.
Stay tuned . . . .