Results tagged “AT&T”

December 23, 2015 5:25 PM

The Tortoise of Super-Fast Broadband

Two weeks ago, AT&T announced plans to bring its "GigaPower," very high-speed (300Mbps and up), broadband Internet service to 38 new markets in 2016--on top of the 18 markets that AT&T already has up-and-running on the service.  A day later, Google Fiber announced on the company's blog that it was exploring expanding its service (available in 3 cities/markets, but under construction in 6 more) to Chicago and Los Angeles.  Both announcements were widely reported by the news media, which has favored a "fiber race" narrative ever since both AT&T  and Google  announced--on the same day--their respective plans to deliver gigabit speed Internet service to Austin, TX.  

Thus, analyst Jeff Kagan compares the strides of "the two heavy hitters in ultra-fast, ultra high speed, gigabit Internet services."  In the Washington Post, reporter Brian Fung observes that, "AT&T is benefiting tremendously from a chain reaction that Google initially began," though he concludes that, "Google's early lead in the fiber race [is] being eaten away by AT&T's traditional advantage in building networks."

But, while the "fiber race" narrative may add an element of human drama to the otherwise impersonal dynamic of broadband competition, Google's fiber-to-the-premise ("FTTP") network is not the first that AT&T would be compared with in the media.  It is, however, the first time the media has favorably compared AT&T to a FTTP-based service provider--and this is the more interesting aspect of the story.    

AT&T Starts the "Fiber Race"

As AT&T's service name--"U-Verse┬« with GigaPowerSM"--suggests, AT&T started building its gigabit speed network long before Google Fiber.   In 2004, AT&T observed that, by using a fiber to the node ("FTTN") architecture (which deploys fiber to the last traffic aggregation point prior to distribution to the customer's premise (the "node")), it could quickly provide better-than-DSL speeds (i.e., 6Mbps vs. 3Mbps) to the maximum number of customers, and position AT&T to be able to progressively replace copper with fiber as bandwidth demand moved from the network core to the edge (residential consumers).  In 2005, AT&T decided it would call its IP network "U-Verse" and service was launched in 2006.  See this 2006 timeline/summary from AT&T. 

To illustrate how FTTN is designed to evolve, consider the tremendous surge in demand for wireless data over the last 10 years.  To expand capacity, AT&T has created more cell sites, and has steadily added fiber to replace the copper lines that "backhaul" traffic from the cell sites to its backbone network.  This means that, in some areas, AT&T can use new fiber in order to "groom" existing U-Verse neighborhoods onto new broadband distribution nodes closer to the customer--thus reducing the copper loop length, and enabling faster DSL transmission speeds.  

It's No FiOS

In 2005, Verizon began deploying its FTTP network, FiOS, and--although Verizon's FTTP would take longer to deploy (to reach a similar percentage of customers) the network itself was/is considered the gold standard.  Thus, among "experts," in the media, and, by self-described "wonks," its early years, AT&T's U-Verse network was always being compared--unfavorably--to FiOS.  U-Verse was the "Jan Brady" of broadband.

An industry newsletter, from 2007, reports that (at the FTTH (Fiber to the Home) Council meeting in late 2006), "AT&T, with its FTTN deployment, showed that it was thinking along the same lines as Verizon . . . [b]ut many in the audience were skeptical about whether AT&T could even deliver HDTV with its fiber-plus-VDSL plant." (emphasis added)  A year later, when AT&T increased its broadband Internet speed from 6Mbps to 10Mbps, one tech news site reported, "AT&T Bumps U-Verse Top Speed to 10Mbps, Verizon Chuckles."  Months later, at the end of 2008, AT&T almost doubled its top speed --to 18Mbps . . . and was still ridiculed.     

As recently as 4 years ago, Susan Crawford--who framed the President's views on telecommunications policy--had already counted AT&T out of the "broadband" market.  In an essay in the New York Times, Crawford argued for the regressive application of Title II regulations for broadband services (which the FCC adopted this year) on the basis that cable was a monopoly, unlikely to be challenged by U-Verse, which "cannot provide comparable speeds because, while it uses fiber optic cable to reach neighborhoods, the signal switches to slower copper lines to connect to houses."  

Perceptions Are Not Reality
Fortunately, for AT&T, its consumers (the people that pay for the network) disagreed with the critics.  In fact, almost a year before Prof. Crawford had discounted U-Verse as a competitor to cable, consumers were telling Consumer Reports that U-Verse was among the best choices (with, of course, FiOS) for bundled broadband, TV, and phone service.  Only a month after Crawford's essay, AT&T verified the Consumer Reports survey, reporting that consumer U-Verse revenues increased by an impressive 44% in 2011.  

By the end of 2013, despite the early skepticism about "whether AT&T could even deliver HDTV with its fiber-plus-VDSL plant," AT&T's U-Verse passed Verizon's FiOS in numbers of video customers served.  AT&T's network and top Internet speeds have consistently improved every year, with its top U-Verse speed increasing to 75Mbps a year ago.  Not surprisingly, consumer adoption of U-Verse broadband Internet service has also steadily improved--growing at 30% annually over the last 5 years.

Finally, while the eye of the media has been on FTTP deployments like Google Fiber, it's what has been happening in copper that has almost-certainly put AT&T in position to provide super-fast Internet access more quickly--and in more places--than any other ISP.  Over the last 6 years, advances in DSL technology have allowed for faster transmission speeds--very close to those supported by fiber--over legacy facilities.

Overnight Success

Having been "counted out"--or never counted "in"--by the media has some advantages. One of these benefits is all the positive publicity AT&T is now getting from publications that may have never expected something they associate with an "innovative" "edge" company--like super-fast broadband--to be done better by a "monopoly" "ISP."   

But, it's difficult to overcome perceptions--particularly when these perceptions have been fed by the FCC.  The Washington Post article, cited above, looks for an explanation for how AT&T was able to overcome "Google's early lead in the fiber race."  Given the perception of Google's early lead, it would be hard for AT&T to convince anyone that it started the race before Google.  Instead, the article quotes AT&T's Jim Cicconi as saying, "[w]e're pretty good at this, and we've had a lot of years to get good at it."  That's as good an explanation as any.  

April 8, 2014 3:06 PM

The Comcast-Time Warner Cable Merger and TV Quality Broadband Deployment

In Comcast's public positioning of its proposed acquisition of Time Warner Cable, executives of both companies have chosen to characterize the merger more by what it's not than by what it is.  So, we know that the merger is not going to result in any significant efficiencies, because it's not going to reduce consumer prices for cable (even an unconstrained monopoly reduces prices when costs decline).

We also know that the merger is not between two competitors, because--as the companies make it a point to tell us--they don't compete.  TWC's CEO says, "[w]hether you're talking about broadband or video, we don't compete with one another."  Comcast's CFO goes as far to state, "[w]e don't compete in one single zip code."  

Doesn't it kind of seem like they're trying just a little too hard to sell the notion that the combined service territory of Comcast and TWC is not relevant (because, you know, they don't compete)?

Product Market Definition

The last time the DoJ's Antitrust Division ("Government" or "DoJ") looked at a Comcast acquisition, it determined--based on documents from Comcast--that Comcast's "joint venture" (as it was structured at the time) with NBC-Universal would reduce competition in the "video programming distribution" market. See Comp. Impact Stmt. (CES).  The Government seemed especially concerned at the ability of post-merger Comcast to destroy nascent competition from online video distributors. CES at C and D.

Based upon the Government's concerns in the previous Comcast acquisition, and DoJ's focus on cross-elasticity of demand in defining a relevant product market, let's focus on some recent information from the Leichtman Research Group to get some valuable insights into how the Government might define a relevant product market.

Consider that, among multi-channel video providers, cable companies lost 1.7 million customers in 2013.  But, AT&T and Verizon added 1.5 million MPVD subscribers last year.  The Leichtman numbers show that customers are not so much "cutting the cord" (only 105k customers stopped buying from an MPVD in 2013) as they are switching MVPDs--but customers are choosing MVPDs that are also broadband providers.  Very high percentages (according to AT&T, well over 90%) of both cable and telco MPVD subscribers are also broadband customers.  The Leichtman data confirm this for Comcast and TWC, as well.

Purchasing video service from another broadband provider, allows the customer to purchase services they want from the MPVD, but also purchase services directly from an online vendor, like Netflix.   In its earlier analysis of the significant competitive effect of online video distributors, the Government referred to this practice as "cord-shaving." CES, at C.2(b).

Given consumer behavior, it seems likely that the Government will focus on a broadband market--of a sufficient speed to facilitate a competitive MPVD service--as the primary relevant product market.  Because it is this market in which the traditional "hypothetical monopolist" test would yield the greatest supply substitution responses.  For all practical purposes, we should consider broadband providers offering service at 10-15Mbps as participants in the "MVPD-bandwidth" market.

Geographic Market Definition

If one's primary concern was to look at the area over which the post-merger firm might be able to reduce competition, then that territory would be (at least) the total number of MVPD-bandwidth broadband customers in each geographic market served by Comcast or Time Warner Cable.  Within this total subset of homes passed will also include the majority of the customers capable of being served by AT&T and Verizon.

What is difficult to figure out from publicly available data is what percentage of MVPD-bandwidth homes will be served within that area by Comcast, Time Warner Cable, AT&T, and Verizon.  For our purposes, just to get a ballpark idea of the type of numbers we would be looking at, we are going to use a datapoint from the Leichtman 1Q 2014 Research Notes that the number of FiOS and U-Verse addressable homes stands at 41 million, giving the companies a video market penetration rate of 26%.  

Let's further assume--and this is a generous assumption toward Comcast--that AT&T and Verizon compete with Comcast and TWC in 70% of their combined service territory, but that all of AT&T and Verizon's customers were won in this territory.  This would give us a total denominator of about 59 million homes passed (that could receive MVPD quality broadband).

Market Shares

To get useful MPVD-broadband numbers, we are going to work with the Leichtman numbers we used earlier, but, because it is impossible to tell from the telco broadband numbers how many AT&T and Verizon broadband customers are actually U-verse and FiOS customers, we are going to use MPVD customers as a proxy, in order to allow us to get some ballpark market share numbers.

merger table_smaller.pngSo, we can see that the result of this merger, for anyone that has to depend on getting content, carriage, or online video distribution to these 60 million households will be looking at a market that goes from "moderately concentrated" to "highly concentrated" under the DoJ Horizontal Merger Guidelines at Section 5.3.

Competitive Effects

The competitive effects on both MPVD rivals like AT&T, RCN, and Verizon, as well as online video distributors like Netflix, are likely to be significant in terms of their ability to get competitive programming.  Add to this the fact that Comcast will also control 12 major regional sports networks, and it is easy to see how the post-merger firm could restrict output of the most inelastic, and  "linear," of linear programming to broadband and online video competitors.

Comcast RSN Map w caption.pngThis last effect is, potentially, disastrous for the future deployment of more MVPD-bandwidth broadband in the area that would be served by the combined Comcast-TWC, because it eliminates what is potentially the biggest source of pent-up consumer demand for MVPD-quality broadband as a substitute for traditional MVPD bundled service--online access to regional sports programming.  

How do we know the significance of real-time sports programming to the value of the broadband Internet?  Because the first truly linear, all HD, over-the-top channel--the WWE Network--has attracted almost 700,000 customers paying $10/month, in only 6 weeks

If the DoJ and the FCC value the availability of MVPD-bandwidth broadband throughout the Comcast-TWC territory, then Comcast might have a reason to worry.  But, commenters on the political left and right have conceded Comcast's powerful influence over the government; so, Comcast probably does have a decent chance of moving forward with this acquisition.  Unfortunately, it just postpones the day when consumers can choose to buy only the video content they want from the vendors they want.

October 31, 2013 2:14 PM

Stuff [Free Press] Doesn't Want You to Know

[The craziest thing happened to me a couple months ago.  You've heard about this NSA spying thing, right?  It has outraged consumers, governments, businesses across the globe, and yearns for us all to consider installing a pretty serious VPN/encryption lockdown.  Anyway, based on what they've seen of my wildly eclectic interests/borderline insanity, the NSA claims that--from what they can tell about my disordered personality--I share online behavioral patterns that they have only seen with some of the world's most unstable/dangerous criminal minds and if they knew more about me they could get a better handle on where the bad guys are (online, anyway).  So they made me a deal: I come clean about the full extent of my online interests/browsing behavior, and they give me the stuff they pull on all of you. "Stuff [They] Don't Want You to Know*" is an occasional blog series where I share some of this information with you.]

So, anyway--to answer your questions--I now know a lot of messed up things about a lot more people these days.  Given the date, and the overall spookiness in the atmosphere today, I'm sure the more inquisitive of you want to know whether I know all about the AT&T-UFO connection. Well, as a matter of fact, I do.  But trust me when I say this: you can't handle the truth.  For now, that vault's just going to have to remain closed.

So let's pull back the curtain on something you can handle.  I recently came into possession of a number of texts/emails between our regulation-happy friends at Free Press and all their regulation-happy friends in the policy world, like Susan Crawford and Tim Wu.  Some of these were texts that seem to have been traded during hearings to take their minds off the inevitable IP transition, and others appear to be just how they were passing the time on long plane or car trips.  And, yeah, you heard right.  Apparently, they do use their phones on planes; and, no, they haven't taken the "it can wait" pledge not to text and drive.

Despite their private contempt for public safety on the highways, and the regulations of the Federal Aviation Administration, one thing's for sure:  these guys do love themselves some regulation (as long as it doesn't, you know, cramp their style).  How much do they dig regulation, you might ask?  Well, they have a little game that they like to play, and I'm going to tell you about it, because it's actually been kind of funny to watch the results.

Here's the game: you take any movie and substitute the words "regulation," "regulator," or "regulate" for any one word in the title of the movie.  So here's how it works, from some actual Free Press private communications:

Regulate Me to Hell (Drag Me to Hell)
The Good, the Bad, and the Regulators
The Silence of the Regulators
12 Angry Regulators
Inherit the Regulation
Regulations of Steel
Fifty Shades of Regulation (DQ'd b/c not a movie yet)
It Takes a Regulation
Texas Chainsaw Regulation
Planet of the Regulators
Gone With the Regulation
Regulators on a Plane
T3: Rise of the Regulations
The Regulations of Narnia
Birth of a Regulation
Triumph of the Regulation

I know; the last two surprised me, too--but, you have to admit, the titles work as films glorifying regulation. Pretty funny, huh?  You could have knocked me over with a feather when I saw this stuff--I didn't even know those guys had a sense of humor. 

Who knows?  Maybe this game will catch on and we'll see Randy May and Scott Cleland trading their own regulation-game quips at the next Free State Foundation Conference . . .

*"Stuff [They] Don't Want You to Know" is a series of occasional blog posts that are entirely fictional and intended to poke gentle fun at figures within the telecom policy world.  Nothing in this series should be mistaken for the truth.

May 1, 2013 11:12 AM

What's the [Low] Frequency, Kenneth? The Government's Uniquely "Consumerless" Concept of Competition

Has anyone else noticed how nutty the news stories have become about the FCC and DoJ fight to promote wireless competition?  Here are some examples: this and this, but I'll summarize for you.  First we have the DoJ "letter" to the FCC; a letter which I think the FCC probably sent the DoJ along with a self-addressed, stamped envelope a few months ago.

I mean, seriously, how could two separate agencies--both independently, and within six months of each other--come up with the same notion that the next available spectrum to be auctioned would be put to its best use by Sprint and T-Mobile (who had not even bid on spectrum the last time it was available) because of its radio frequency characteristics?  That last part was highlighted because it's like the peanuts on top of the walnuts on top of the almonds in this all-nut sundae of a theory.

Like most tin foil hat theories, this one has a small kernel of logic.  For a smaller carrier, especially a new entrant, low-frequency spectrum provides a lot more value per cell site--and requires a lot less cell sites--for a carrier to achieve adequate coverage. But do the FCC and DoJ want to promote smaller carriers or new entrants?  Of course not; that might provide consumers with some value.  And since the FCC/DOJ believe that only national firms count toward improving competition in the marketplace - new entrants as envisioned by these agencies would fail to meet that goal.

The DoJ and the FCC didn't have this theory of theirs until they also seemed to arrive at the conclusion--as near as I can tell, sometime during their analysis of the proposed AT&T/T-Mobile merger--that mobile wireless competition is best measured by market share on a national level.  And, with a market artificially defined as "national", despite the fact that consumers make choices locally, a "market" could only be truly competitive if each firm's share (of customers, of spectrum, of cool new handsets, and crunchy nut confections) is roughly equal.

Does anybody recognize the problem with this raison d'etre?  Does the conclusion at the end of the last paragraph sound a little like the description of a commodity market?  Yeah, it kind've does, doesn't it?  Are wireless services a commodity market?  Well, the AT&T iPhone crowd from 2007 didn't seem to think so; nor did the Verizon Droid evangelizers from 2009.  So, let's just say no; wireless is not a commodity market.  Like with cars, people seem to take a certain personal pride in their selected combination of network and handset.  

Why would anyone expect that differentiated product markets would result in competitors having a roughly equal share of sales?  After all, some people like (and can afford) fancy overpriced compact cars, while others need pimped-out, baller SUVs because . . . that's just how they roll.  So isn't it nice that we have BMWs and Escalades?  Do they have the same market share?  Yeah, probably, but that's beside the point.

The problem with the government's idea of what competition should look like is that it starts from a lot of flawed premises--all of which come from the same flawed premise: consumer preferences don't count.  The relevant geographic market is national, not because this is the way consumers actually purchase wireless service, but because this is the way the government likes to look at it.

To the government, market shares are only unequal because firms have unequal amounts of low frequency spectrum, and not the other way around.  They don't seem to understand that AT&T and Verizon have customers that, for the most part, have chosen not to buy service from at least 3 other firms.  Now that's competition.  

Why doesn't the government just reconcile itself to the reality of consumer driven competition and "wreckanize" that the consequences of choice can produce distinct winners and losers?  Yogi Berra told us a long time ago:  "If the people don't want to come out to the ballpark, nobody's going to stop them."  Why do the DoJ and the FCC keep trying?

May 17, 2012 3:23 PM

U.S. Mobile Data: More Bricks, Less Straw

As punishment for requesting their freedom, the Egyptian Pharaoh told the Israelite slaves that they had to maintain their quota of bricks, but with less of an essential input: straw.  The burden of meeting demand with less resources applies as well in today's mobile services marketplace.  Wireless carriers face demands for greater bandwidth to support growing mobile data services but, for the intermediate term, cannot expect additional spectrum capacity--it's essential input--on either a firm-specific or industry-wide basis.  

It's unanimous: no matter who you talk to about wireless data, everyone agrees that "more bricks, less straw" is the unavoidable policy.  Thus, as wireless data demand continues to show no sign of abating, wireless service providers will simply have to make do with less than optimal spectrum capacity. So if we're stuck in a "more bandwidth demand, less capacity supply" world, how do we solve the problem of how to ration capacity?  

Who Needs Spectrum When You Can Upgrade Your Capacity?

So what's a wireless operator to do?  Well, for starters, you upgrade existing capacity like crazy by constantly deploying the most efficient technology.  But this isn't cheap.  Since wireless data exploded in 2007 with the iPhone, AT&T alone has gone through a 3G upgrade, an HSPA upgrade, an HSPA+ upgrade, and, more recently, is in the midst of an upgrade to LTE.  

Other companies have accelerated their own pace of upgrades as well.  From December '06 (right before the iPhone launch) through December '11 (when most firms still have a long way to go to realize full LTE deployment) industry capex has increased by almost 50%, according to CTIA (the actual numbers are in a report that I can't afford, so take my word for it).  But, even these improvements won't keep up with surging demand.

With No Spectrum Relief In Sight, Do You Play The Price Card?

Given the limited options for rationing capacity another, though unpopular, move is to raise prices.  Over the last several months we've seen AT&T raise data prices, after realizing that the government was not--anytime soon--going to allow AT&T to efficiently augment its own capacity.  Verizon quickly followed suit.  For now, Sprint appears to be content to let its shareholders shoulder the costs of increased wireless data demand.  But to be sure, increased demand without increased supply does create network strain--regardless of who pays.

Last week, at the CTIA Conference, Chairman Genachowski maintained/reasoned/disputed that the failure of the AT&T and T-Mobile merger last year had anything to do with AT&T's decision to raise prices.  Yet, the Chairman knows better, as he has been a leading prophet of the spectrum shortage.    

How to Recover Costs of Spectrum-less Capacity Expansion?

Given the costs of constantly upgrading capacity, how does a carrier manage excess data demand?  As I indicated above, raising prices sounds like a simple solution, but must account for the fact that big data users are contract customers.  That's how smartphones, and data plans become affordable, and predictable.  

You see, the problem with raising prices for wireless data is that you can really only raise prices to the marginal customer (i.e., the person who's not your customer yet).  Crazy, right?  "Raising prices" is a statement of frustration and designed to curb consumption.  Carriers are telling prospective customers that the network is nearing capacity and use of the remaining capacity will cost you. This is a horrible situation--who wants to be the (unpopular Redskins owner) Dan Snyder of wireless data?

If Sophisticated Buyers Want to Subsidize Consumers, Let Them!

Carriers know that raising prices for mobile data, or throttling data speeds to the largest users of mobile data, is no way to treat your biggest fans.  But with the popularity of mobile device applications, which constantly stream information to and from the customer's phone, customers can unintentionally (and unnecessarily) stress capacity.  Applications can distort data consumption in a way that even the most conscientious web surfers cannot offset.

So, earlier this year, at a conference in Barcelona, an AT&T executive suggested that maybe some applications providers would want to buy capacity in bulk in order to assure their customers that using the desired app wouldn't cause the customer to exceed their usage cap, or become subject to throttling.  Not a bad idea, right?  I mean the applications developer knows how much bandwidth their customers use, and they have a lot more buying power than the consumer.  

Given the public's embrace of mobile data, and the cost of continually augmenting capacity, especially for firms with sub-optimal spectrum allocations, one would think the "public interest" would support options that allow customers to still enjoy wireless data, but at a lower cost/consumption threshold.  One would think . . . .

But Don't Tell Public Knowledge!

The AT&T suggestion seemed harmless enough, but the reaction from the self-proclaimed public interest group Public Knowledge was alarmingly critical.  Then again, this is the same group that published a paper arguing that all wireless carriers should provide flat-rated mobile data service.  The irony, of course, is that flat-rated price structures cannot be profitable unless the majority of users pay for more data than they consume.

The notion of "more bricks, less straw" is, for regulators and service providers, an unfortunate and dystopic reality.  Uniquely, Public Knowledge seems to relish the "more megabytes, less capacity" future with a fondness that can't help but be compared with how the ancient Egyptian brick consumers' lobby must have felt . . . right before the brick supply crashed.

August 23, 2011 9:54 AM

Sprint's Gambit

Virtually from the announcement of the AT&T/T-Mobile merger, both Sprint's advocacy and Sprint's ultimate goal in its opposition to the merger (blocking the merger outright) have been a puzzle to me.  Specifically, Sprint's categorical opposition to the merger makes me wonder: "why the opposition?" and "what are they really looking for here?"
Sprint's superficial, substantive arguments against the merger are a fairly generic, typical "Guidelines" style analysis from a consumer welfare (rather than Sprint-specific) perspective.  Coming from a competitor, these arguments invite suspicion. 
In fact, the Supreme Court has been extremely skeptical of competitors seeking to block mergers on the grounds that "competition" will suffer as a result of the merger.  See, e.g., Brunswick Corp. v. Pueblo Bowl-O-Mat, Inc., 429 U.S. 477 (1977)(rejecting antitrust standing of competitors seeking to block an acquisition on the basis that it would concentrate the market, and leave the post-merger firm with lower costs, creating a potential threat to plaintiff competitors).  

Regardless, it makes little difference whether Sprint's opposition is motivated more by self-interest than the public interest.  However, Sprint has been quite clear in its advocacy before the FCC and the Congress that the objective of its advocacy is to see the proposed transaction blocked.

The Puzzling Perplexity of Sprint's Advocacy Goals

If Sprint's preferred outcome is for the proposed transaction to be completely blocked, then we have to ask, how does this outcome maximize Sprint's self-interest?  This is the question that is so vexing to me.  After all, if a market becomes more concentrated, and leads to price increases and less innovation, one would expect that the last party to complain would be a competitor remaining in the post-merger market.

It seems obvious that Sprint will only benefit from higher retail prices.  Less innovation, as well, would have its rewards in terms of reduced capex pressure, and reduced pressure to constantly roll out new bandwidth-gobbling devices. These are real benefits, so why does Sprint want the merger blocked (vs. "conditioned")? 

The Gambit

A true "gambit" requires that a player sacrifice something of value to gain a greater strategic advantage.  In this gambit, there are at least two very obvious sacrifices that Sprint is willing to accept from the outset.

The first sacrifice is that Sprint will not seek merger conditions, because (according to Sprint) there are no conditions that could mitigate the consumer harms created by this merger.  This is a "real" sacrifice, because Sprint could have reasonably expected to extract some valuable concessions.

The other sacrifice is much more significant, but has gone completely ignored by the industry insiders and press.  The fact is that Sprint, through its advocacy, has disqualified itself from acquiring T-Mobile.  So Sprint is not seeking to disqualify AT&T from acquiring T-Mobile, so that it may subsequently acquire T-Mobile for a lower price.

This sacrifice is an unequivocal and inseparable part of its economic analysis of the merger.  As Sprint explains, the current proposed merger allegedly increases the HHI by 700 points, based on market share numbers listed.  On the other hand, a Sprint/T-Mobile merger would move the same "highly concentrated" baseline HHI up by 500 points.  See Table 4 ("postpaid only" column) in the economics declaration in Sprint's Petition to Deny (p. 257 of 377).

It is, therefore, quite clear that Sprint's analysis--if accepted by the government  as a basis to block the merger--would also disqualify Sprint and Verizon as subsequent bidders for T-Mobile, thus "cementing" the "national wireless market" (if you buy this definition) as a 3 firm market. Why do I say 3 carriers when we are left still with 4 "national" providers?  Because in that world, T-Mobile is fundamentally crippled from a competitive standpoint and that, in essence, is the big win.

The Gambit Payoff

Consider the advantages to Sprint of a capacity-constrained 3 firm oligopoly market structure.

1)  Customer Share/Acquisition. Constructively, T-Mobile will be an island, cut off from ready access to capital from its parent, Deutsche Telecom.  Without a "true" 4G network, T-Mobile will gradually lose customers to other carriers in the market.  Many would argue that Sprint is the next best substitute for T-Mobile, so Sprint may gain disproportionately from T-Mobile customer defections. 

2)  Accelerated Growth in Data Services.  Let's assume that Sprint (rationally) believes the fact that AT&T is very data capacity constrained.  If the government successfully blocks AT&T's proposed acquisition of T-Mobile, the remaining three 4G carriers (remember, under the Sprint theory we virtually ignore other competitors like Metro PCS who may have 4G networks, too) have no real means to acquire additional spectrum capacity. 

The number one provider of wireless data service--AT&T--will be supply inelastic in most cities for the foreseeable future (no capability to acquire sufficient spectrum). This puts AT&T in a "shortage" situation, where it must set prices not to maximize profits, vis-Ó-vis costs, but to increase prices, and reduce output, in an effort to ration service consumption.
Once AT&T is forced to implement "congestion pricing", it is logical to expect that Verizon, the number two wireless data provider, will adjust its own prices in order to preserve its network capacity in a spectrum-constrained market (it can't acquire spectrum either).  The only party that "wins" in this scenario is the firm with the largest excess capacity--Sprint--which through its and Clearwire's holdings has more spectrum, and more capacity than anyone else in the market, including AT&T or Verizon. See FCC's 15th Wireless Competition Report, at Table 28 (Sprint/Clearwire has a weighted avg. of 184.4 MHz of spectrum vs. a combined weighted avg. of 173 MHz for AT&T and Verizon Wireless combined)
3)  Peace of Mind.  Can you ever really put a price on it?  With a tight 3 firm oligopoly, characterized by high barriers to entry, Sprint no longer has to "watch its back" as the undisputed "national" value brand.

The Beautiful Genius of Sprint's Gambit

Sprint's analysis virtually ensures a 3 firm oligopoly going forward, because every potential bidder with the experience and financial capacity to acquire T-Mobile is eliminated.  T-Mobile lacks the resources to build out a 4G network, and no other carriers have the kind of money they would need to buy T-Mobile, at their present size, and then upgrade T-Mobile's network to being 4G capable.  Finally, if T-Mobile has to continue as an independent carrier, their only realistic alternative for offering "national" 4G services is through . . . Sprint's majority-owned wholesale carrier, Clearwire.  This is a brilliant gambit.