Spectrum: The Capital that Drives Wireless Competition

Wireless competition depends on spectrum: We can only have as much effective mobile broadband competition as have spectrum to support. This fact doesn’t get as much attention as it should in the general discourse on network competition or the specific controversy currently raging over the proposed ATT-T-Mobile merger. While Washington is clearly aware of the need for more spectrum, the link to competition has been lacking.  The National Broadband Plan called for a massive release of spectrum to enable mobile networking:

Recommendation 5.8: The FCC should make 500 megahertz newly available for broadband use within the next 10 years, of which 300 megahertz between 225 MHz and 3.7 GHz should be made newly available for mobile use within five years.

In the year since the Plan was published, there’s been a lot of talk in Washington about this recommendation: Some partial inventories have been released, some candidate bands have been identified, a number of bills have circulated regarding spectrum auctions and additional inventories, but there’s been no concrete action to implement the recommendation.

The spectrum crunch is real: Without spectrum, we don’t get the benefits of LTE, the next-generation mobile broadband system that speeds up mobile apps and potentially increases broadband competition across the board. The essential element of LTE, after all, is its ability to translate larger chunks of spectrum into faster network interactions, and that depends on having access to lots of radio spectrum in the first place. LTE is not only about efficiency, it’s also about faster networks that consume more bandwidth overall to provide a more satisfactory service to a broader range of applications. The candidate spectrum for LTE was firmly in the clutches of local TV broadcasters and government agencies a year ago, and it remains in the same hands today.

Against this background, two of our national mobile network operators have struck a deal for short-term relief from the spectrum crunch, the proposed takeover of T-Mobile by AT&T Wireless. Prior to the 700 MHz auction, AT&T held licenses to 91 MHz in the largest 100 markets (it’s up to 110 or so now,) so the acquisition of T-Mobile’s 54 MHz would increase their spectrum holdings by roughly 50%, putting them on par with Clearwire but in better frequencies. Here’s an oldish slide that tells the tale:

Spectrumtable

It’s no secret that Deutsche Telekom had been looking to sell T-Mobile for some time prior to Sunday’s bombshell announcement. While most mobile networks are gaining subscribers – especially the small carriers such as MetroPCS and Leap – TMo has been losing them.

While most national carriers are planning for their LTE transitions, spectrum-starved TMo has had to content itself with polishing up its 3G service with higher-efficiency and faster HSPA+. TMo was late to the spectrum auction party, and made a poor bet on smartphones, partnering with HTC for the ill-fated G1 instead of going with more solid plays such as a 3G Blackberry with AWS support and better Android platforms from Samsung, LG, and Motorola, and isn’t well positioned in terms of overall spectrum holdings.

Deutsche Telekom is a thoroughly European company that does well in the minimally competitive continental markets dominated by relatively uninspiring Nokia phones and uniform GSM technology. To succeed in the rough-and-tumble American cellular markets, established networks invest savagely in spectrum auctions and equipment to expand data services, and upstarts such as Clearwire and LightSquared offer different value propositions and technologies than established players. In the U.S., apps and platforms are as important as towers and backhaul are in Europe, and the market moves at Internet speed. Even in England, Europe’s most competitive mobile market, DT has chosen to partner with France Telecom in the “Everything Everywhere” joint venture rather than going it alone.

It’s been a foregone conclusion that DT was leaving the U.S. market as a stand-alone competitor, and the only question was who they would partner TMo with. Before Sunday, the smart money was on Sprint, but a Sprint-TMo tie-up never made sense due to the incompatibility of the two firms’ technologies. Sprint, like Verizon and most American carriers, is a CDMA company, while TMo uses the same GSM/HSPA+ technology that AT&T employs. TMo was also planning an eventual transition to LTE, like AT&T, Verizon and most others, while Sprint is still officially banking on Wi-Max to provide its upgrade path to the future.  And while Sprint isn’t spectrum-rich in its own right, as the lead partner in Clearwire it has access to ample spectrum; for Sprint, the value proposition in a T-Mobile acquisition was customers rather than spectrum.

Sprint has already undergone merger pains with the incompatible Nextel network it acquired in 2005 and knew well enough to discount its offer for TMo to a level that would include transition costs. AT&T didn’t have to discount its offer, and was therefore able to out-bid Sprint for a network that it can more or less immediately use, subject to the vagaries of some of the oddball AWS frequencies in the TMo spectrum portfolio.

The merger highlights the fact that technological analysis and economic analysis of the same phenomenon can often lead to different answers about the good and the bad. The merger of the nation’s second and fourth largest networks clearly reduces competition: Instead of a market with four national networks, we now only have three; instead of the market with three equal sized competitors that would have come about from a Sprint-TMo merger, for example, we have two major players with a combined market share of 70%, one minor national player, and a host of small regional operators; and instead of three national LTE networks sometime in the future, we’ll now only have two (unless Sprint dramatically changes course and chooses LTE over Wi-Max.)

Approval of an AT&T-T-Mobile merger means that regulators would be bound by consistency to approve a future Verizon-Sprint merger and usher in an official duopoly in mobile networking. A duopoly American cellular market would resemble a continental European market, where the common scenario features two dominant carriers with a combined 75+ percent share and a small group of upstarts nipping at their heels.

Marketshare

Duopoly markets must be more heavily regulated than more competitive ones, and government regulators don’t make pro-innovation choices nearly as well as technical competitors do; this is why the U.S. has pursued pro-competitive broadband policy goals since the 1996 Telecommunications Act and why we lead the world in innovative mobile devices and applications. But as the National Broadband Plan reminds us, effective competition requires capital, and in the mobile space some of that capital takes the form of spectrum. Network operators are all retailers in a sense, buying capacity (spectrum and backhaul) in bulk and reselling it to individual consumers. To successfully operate an LTE network that enables more people to use more bandwidth, more spectrum licenses are a precondition.

DT’s desire to leave the U.S. market suggests that we don’t have enough available spectrum to sustain four national 4G networks, the same message the National Broadband Plan delivered. When market behavior reinforces policy predictions like this, it’s wise to pay heed. So when are the Washington policy makers who have the power to re-purpose inefficiently used and underused spectrum going to take action?

Enabling the mobile competition policies that worked so well in the 2G voice era to carry over to LTE requires quick and effective action. Make no mistake about it, competition depends on the resources that mobile networks have at their disposal. 150 MHz of spectrum is enough to sustain five or ten national voice networks, but only enough for a single national LTE network; the more competitors we want, the more spectrum we need. The U. K. has taken this message to heart, adjusting the results of its LTE spectrum auction to ensure that four networks will emerge from the process with the ability to provide national coverage.

This may seem counter-intuitive, because total spectrum requirements should be dictated more by the total number of users (and what they’re doing) than by the number of operators. This is true except for two big qualifiers: Multiple operators can’t use the same spectrum as efficiently as a single operator can, because they have to allow for guard bands between their holdings (like the TV White Spaces between broadcast channels do) as well as statistical variations on what they can share via roaming agreements and the like. Operator A can’t make a roaming agreement with Operator B that compromises his ability to serve his own customers. And more importantly, the existence of competitive operators stimulates innovation, which gives rise to more consumption. So we don’t just tolerate spectrum inefficiency in the name of competition, we actively promote it. Spectrum enables wireless competition, and we will only have as many healthy mobile networks as we have spectrum to power.

Policy makers need to make a grown-up choice: Do they want competition among a good number of healthy, wireless competitors, or do they fear the political clout of the current spectrum occupants so much that they’re willing to compromise American competitiveness for the sake of preserving legacy uses such as high-definition over-the-air TV spread out over a wide range of frequencies? This sounds like an easy choice to those of us who don’t have to win elections to keep our jobs, but the reality is that broadcasters have more First Amendment power than the rest of us, and they are in fact fully willing and capable of making life very difficult for elected officials unless there’s a united front on both sides of the aisle against their spectrum-squatting ways. Likewise, it’s time to get government at all levels to stop wasting valuable spectrum on legacy applications that can be relocated and updated to use spectrum more efficiently.

Finally, public policy makers need to carefully consider the effects that the naïve and misguided “net neutrality” fixation has on network competition. MetroPCS, the kind of scrappy, upstart network operator that should be encouraged by lovers of innovation, is the first operator to find itself in the crosshairs of the FCC’s Open Internet rules. MetroPCS offers a limited Internet plan that allows users to view YouTube videos but not others without an additional fee.

Neutralists are aghast that anyone would offer access to part of the Internet instead of the whole enchilada, but their objection is more ideological than practical. If people want to buy a “YouTube Plan” instead of a “Whole Internet Plan,” why shouldn’t they be allowed to do so? No good argument has been offered to explain why such partial plans should be disallowed, and those arguments that have been offered sound like pleas for eating broccoli three times a day: “Trust me, it’s good for you.” We might very well all be better off on vegetarian diets, but such things are personal choices, not public policy decisions.

Many smart people in Washington seem to think that the approval of the AT&T-T-Mobile merger is a foregone conclusion. They believe regulators will find a way to address the current constraints for spectrum, special access, the ability to shoulder arbitrary net neutrality rules and related issues such as handset licensing, and will approve the deal with necessary conditions. They could be right, but the fact remains that regardless of the outcome of this particular merger review, effective wireless broadband competition depends on the long-term availability of adequate spectrum.

The message couldn’t be more clear: A $39 billion dollar bid for an ailing network whose main value is spectrum is a clarion call to Washington to implement the National Broadband Plan’s recommendation to free up 500 MHz for mobile broadband; AT&T’s desire to merge with T-Mobile is like buying a car to get a set of tires. The transaction places a value of 750 million dollars per megahertz on T-Mobile’s spectrum portfolio. When we translate that figure into the economic gains that come from the actual deployment of 500 MHz into productive use, we’re looking a number that’s large enough to make a dent in the national debt.

Most importantly, we have to realize that the continuation of America’s successful pro-competition and pro-innovation broadband policy requires a much larger pool of spectrum. This proposed merger is telling policy makers that spectrum is not just the oxygen that enables innovation to breathe, it’s the capital that drives competition.

Print Friendly

About the author

Richard Bennett is an ITIF Senior Research Fellow specializing in broadband networking and Internet policy. He has a 30 year background in network engineering and standards. He was vice-chair of the IEEE 802.3 task group that devised the original Ethernet over Twisted Pair standard, and has contributed to Wi-Fi standards for fifteen years. He was active in OSI, the instigator of RFC 1001, and founder, along with Bob Metcalfe, of the Open Token Foundation, the first network industry alliance to operate an interoperability lab. He has worked for leading applied research labs, where portions of his work were underwritten by DARPA. Richard is also the inventor of four networking patents and a member of the BITAG Technical Working Group.