May 2010 Archives

May 28, 2010 4:30 AM

Wireless Competition Report: The "Ugly"

Before I talk about some else's "ugly"--I'll 'fess up and say that last post was dry as dirt.  You know what's worse?  "The Ugly" isn't going to be any less ugly! 

So, let's just move through it.  The "ugly" part of the Wireless Competition Report is the adjacent market analyses--the "downstream" and "upstream" markets information.  While some of this data--meticulously compiled as it was--is . . . well . . . interesting, but its relevance to the state of competition in the mobile wireless services market is questionable.

Simply put, the adjacent market analyses were superfluous, almost by definition. Why?  Because, strictly speaking, the statutory language that mandates the preparation of the Report requires the FCC to report on the state of the mobile services market and the service providers that comprise that market--period.  This means that the market for which Congress seeks information on is the market to be isolated.  Congress did not require the Commission to Report on "downstream" markets, or any other ancillary markets.  

"Downstream" Markets

Strictly speaking, the "downstream" markets for devices, operating systems, and mobile applications seem almost too integrated with wireless service to be considered "downstream" in the production/distribution chain.  This is because it is not possible to receive any wireless mobile service without some sort of device.  Semantics aside, though, the Commission's data seems consistent with healthy competition in the wireless market--that is to say that output of handsets is increasing by number of manufacturers (which have doubled from 8 to 16 in the past 3 years), Report, ¶300, and the prices of handsets and smart phones have decreased dramatically in the past 3 years, Report ¶310.  

But, the chicken-egg problem still exists, with respect to being able to make any inferences regarding the state of mobile wireless service competition based on the "downstream" market data.  The largest carriers offer consumers the greatest choice of handset and smart phone models.  Report, ¶308, Chart 43.  However, one cannot say what this information means vis-à-vis the competition for mobile wireless services.  Do the largest carriers have the most customers because they offer the most handsets?  Or, can the carriers with the largest number of customers offer the greatest choice in handsets because they have diverse enough customer bases to support the most diverse number of handset models?  Is there another, more anticompetitive, theory to explain the relationship between "downstream" market devices and mobile wireless service competition?  We don't know.  No correlation between handsets and service competition is ever offered.  This is the "ugly."  Informative?  Sure.  Probative?  Of what?  Not clear.

"Upstream" Markets
The "upstream" market section of the Report suffers from the same problem--the difficulty of drawing a correlation between mobile wireless service competition and the "input" markets of spectrum, access to tower sites, network equipment costs, and backhaul costs.  As noted in one of the previous posts, I did "kind of" like the effort to look at inputs. . . at first . . . until I thought about it more.  On deeper consideration, it seems unlikely--given the economies of scale and scope that characterize wireless mobile networks--that this exercise is ever likely to produce any information that would not be potentially misleading.  Rather, it would seem that the large economies of scale and scope in the wireless network services market would simply indicate that (other things being equal) the carriers that serve the largest number of users (either directly, or through MVNOs) will have the lowest costs per user.

Also, the Report failed to describe to what degree, if any, these costs were competitively significant.  In other words, to some degree or another, all carriers must deal with the "role of spectrum", obtaining tower space, network equipment, and backhaul from their cell sites or other points of aggregation.  It was unclear from the Report whether any carriers face input costs that cannot be overcome by superior competitive performance.

Role of Spectrum.  Here the Commission posits that some firms (the first movers on cellular spectrum in the 1980's, and the winners of the 700 MHz auction) have better spectrum than others, in that it is cheaper to build out--due to its superior propagation properties.  However, the Commission also notes that this "lower build-out cost" spectrum costs more to buy than the more expensive to build-out, higher frequencies. Report, ¶271.  This seems like the classic "operating leverage" concern confronting all firms in all industries. 

A higher initial fixed cost, can frequently yield very high profits--past a given output level.  Conversely, a lower initial cost of entry is often associated with higher variable costs.  Neither is necessarily best.  For example, many MVNOs probably outperformed many facilities-based networks over the past few years.  Will MVNOs, with their low cost of entry and high variable costs, always outperform the same facilities-based carriers?  Who knows?

Other Infrastructure Costs.  Costs associated with tower site acquisition, and associated network equipment, seem to favor the firms with the largest established customer bases.  This is because it costs less to incrementally add sites than it is to build a network starting with a low customer base.  Again, while this suggests that the wireless mobile market is characterized by high fixed costs, it does not explain whether "newer" entrants--with less investment in legacy network design, equipment, and reliance on traditional forms of backhaul (copper DS1s and DS3s)--are able to compensate for these higher costs of initial entry with greater capacity, lower operating costs for a given number of subscribers, and greater revenue opportunities (for example, through offering higher-speed broadband services through newer network design). 

Backhaul.  It is, likewise, unclear to what degree backhaul costs effect competition in the mobile wireless industry, or the degree to which wireless demand affects competition in the backhaul market.  The Commission notes that "traditional" copper backhaul is quickly losing ground to fiber-based backhaul.  Report, ¶294.  However, the Commission also suggests that "unaffiliated" (with an incumbent LEC) wireless carriers may be at a disadvantage to "affiliated" wireless carriers, due to the costs of special access backhaul (traditionally provided by the incumbent LEC).  Report, ¶¶ 295-296.  On the other hand, given that the largest two wireless networks are also the largest purchasers of "other" incumbent LEC backhaul, it would be helpful to know whether these two carriers have a greater incentive/propensity to differentiate their costs by devoting more "spend" to out-of-region competitive fiber providers--thus promoting backhaul competition. 

Another problem with the "backhaul" section is that it never attempts to quantify backhaul costs in absolute terms, or as a percentage of annual costs or revenues.  The only reference it makes is to a study in Verizon Wireless' NOI Comments, stating that backhaul was expected to increase from a $3 billion market now (2008 total mobile wireless revenues-$150 billion (Report, ¶201)) to an $8-10 billion market in the next 5 years).  Report, ¶296, n.785.  Assuming an industry cost of backhaul at $3 billion, this would put backhaul costs at slightly less than advertising costs, at around $3.4 billion for the most recent year.  Report, ¶128. 

It might be the case that backhaul costs, and the other "upstream" input costs discussed in the Report, are a competitive concern, but the Commission didn't support this rhetoric with data.  Again, misleading, and, therefore a little "ugly" . . . . 
May 26, 2010 2:50 PM

Wireless Competition Report, Part 2: The Bad

In all fairness to the Commission, the mobile wireless market is undisputedly complex, and the Commission notes the difficulty in analyzing this market.  Yet, this is precisely the "Bad" part of the Commission's Wireless Competition Report: the information provided, to the extant it is relevant to competition in the mobile services market, is frequently presented in an incorrect, or otherwise misleading, manner.  Here are some areas of potential bias that deserve an "on the other hand."

Concentration.  Both the Commission, and the Department of Justice, have stated that the HHI is just a starting point for a competitive analysis.  The Commission makes a point that the "weighted average" HHI (relying on data not available for public inspection) is over the FCC's self-imposed "highly concentrated" threshold of 2800 for the first time since 2003.  Report, ¶51.  The FCC also notes that, in a merger review, it would have to see an HHI of 2800 with a delta of 100 before becoming concerned about concentration.  The current "weighted average" HHI is 2848 . . . phew (a delta of less than 100)!  While some of the concentration has occurred "organically", with the networks with the highest CapEx investment getting the most customers, much has undoubtedly come from mergers that the Commission found to be in the public interest.

The HHI the Commission used was based on a "weighted average" of market shares, based on past market share as defined by dollar sales revenues of the market participants.  However, the Merger Guidelines allow for HHI calculation by unit sales, relative share of capacity, or even a "bid model", which means that "[w]here all firms have, on a forward-looking basis, an equal likelihood of securing sales, the Agency will assign equal market shares." Guidelines § 1.41, n. 15. 

It seems like the mobile wireless service market is a market where all competitors would have, on a forward-looking basis, an equal likelihood of securing sales.  For example, while the FCC notes early in its Executive Summary, p.6, that "net adds" for the top two firms vs. "the next two" are asymmetrically high, a closer inspection reveals that only one of "the next two" top national firms experienced net losses over the last 2 years, p.9, and that firm's losses were only severe in 2008.  Moreover, absent data-skewing due to a merger, three of the four national firms had positive net adds over the last two years, and a much less dramatic disparity in percentage increases. 

Thus, considering that all firms seem to have an equal chance of gaining new sales, it would probably be safe to include the 5 firms listed in Table 4 of the Report (which cover around 75% of the population) in this "1/n" market share assignment, because "nationwide" firms do not appear to geographically price discriminate. Report, Table 10.  Whether one uses 5 firms, or 4, the HHI is appreciably less than 2800 (either 2000 with 5 firms, or 2500 with 4 firms).  Similarly, if one were to take capacity as being the defining factor for competitive advantage, the HHI would still come out less than 2800 (top 5 firms have 80% of nationwide spectrum).

HHI Does Not Indicate Diminished Market Performance At An Index of 2800.
Appendix C, Table C-3 of the Report lists 14 markets with the highest penetration rates in the country (from 98% to105%).  These are the markets characterized by the highest "output" levels (high output is a performance characteristic of competitive markets).  Of these, only 6 were below 2800, and only 2 were below 2500.  Moreover, the number of wireless-only households continues to climb (now over 20%).  Report, Chart 46.

Exclusion of MVNOs. If the Commission is going to use sales of facilities based mobile wireless providers in its HHI calculation, it is only logical to include "nationwide" MVNO sales as well.  Why?  Because these carriers are getting sales and satisfying consumer demand.  The success of the MVNOs indicates competitive market performance, because these sales represent excess capacity that is being put into the market, and not being withheld from the market in order to increase prices. 

The Commission concedes that Tracfone would be the 5th largest nationwide provider, and that Tracfone's "Straight Talk" offer is carried on the Verizon Wireless network. Report ¶101.  The Commission further notes that Verizon Wireless is the leader in industry performance.  Report, ¶223, n.634.  If Verizon Wireless is enabling an MVNO to come out with a $30 prepaid "all you can eat" voice product, and a $45 data product, and MVNO sales accounted for almost half of Verizon's net adds in the 4th quarter '09, then why would the FCC want to misleadingly use these MVNO sales "against" Verizon?  Sprint has a CDMA network, too.  And, with their 4G rollout, they should have just as much chance as Verizon of picking up Tracfone's contract if Verizon fails to provide superlative service.

Profitability. In order to justify "profitability" analyses as evidence of (or even relvant to) "market power", the FCC uses a "perfect competition" (i.e., price should equal marginal cost) definition of market power in ¶12 of the Report ("the ability to profitably charge prices above cost for a sustained period of time"), but later, buried in ¶55, the FCC uses the correct definition of market power for a firm competing in a "real world" less-than-perfectly-competitive market ("the ability to charge prices above the competitive level for a sustained period of time"), which is also the definition used by the DoJ and FTC. 

After explaining that profitability is not the correct method of defining "market power", the Commission concedes that we have no accurate measures of profitability for this industry.  Report, ¶215. But, does the inability to correctly correlate admittedly-inaccurate measures of profitability with market power (higher than "competitive" prices) stop the FCC from devoting 7 pages of the Report (pp.223-229) to exactly this purpose? Regretably not.

The facts, though, are that retail prices have been declining every year for the past 11 years, both in absolute terms and in relation to the overall CPI.  Thus, it would seem that "market power" defined as "the ability to keep prices above competitive levels for a sustained period of time" has not existed for a while in the mobile wireless industry.  The existence of "higher-than-cost" profits are, by definition, not probative of "market power."   Given their irrelevance, the "profits" in this section (no matter how the are defined) are misleading--more likely to be the effect of operating leverage in a growing, high fixed cost industry, than an indicator of "market power."

CapEx: Industry Size/Investment.  The Commission clearly thinks that industry, and firm-specific, CapEx investment over the years seems to have some relationship with the level of competition in the industry.  While it is not surprising that the two largest firms also have the highest levels of CapEx, it is unclear whether their investment gave them a better product with which to get more customers, or whether their large customer bases have forced greater CapEx in order to maintain service levels. Report, Chart 33.  However, it is never clear from the Commission's discussion what, if anything, these figures mean.  For example, if you are a carrier that is not investing in CapEx are you at over-capacity, or, are you losing customers due to poor customer experience?  If you are such a firm, would increasing CapEx get you a better network/level of service? 

Bottom Line:  There are significant areas of the Report that could be interpreted as implying that a reader could reasonably conclude that the wireless mobile services market is not competitive, and is becoming less so.  If the Commission explicitly stated that this was its assessment in its Report to Congress, then people could argue about the conclusions.  However, the Commission's decision to not include certain information, or to include concededly-irrelevant information, and to present this information in a manner that implies only one conclusion, can only be regarded as troubling.  

May 25, 2010 12:59 PM

Wireless "Competition" Report, Part 1: "The Good"

Last Thursday, at its May Open Meeting, the FCC adopted its annual Wireless Competition Report.  The adoption of the Report caused something of a stir, because the Commission departed from the statutory directive requiring the Report to:

include an identification of the number of competitors in various commercial mobile services, an analysis of whether or not there is effective competition, an analysis of whether any of such competitors have a dominant share of the market for such services, and a statement of whether additional providers or classes of providers in those services would be likely to enhance competition.
47 U.S.C. § 332 (c)(1)(C) (emphasis added).  The report, notably, reached no conclusions about anything required by Congress, except identifying the number of competitors in various mobile services markets.

At the meeting, Wireless Telecommunications Bureau Chief, Ruth Milkman, by the account in Friday's Communications Daily, explained,

'What we set out to do when we started drafting this report was to collect the facts and to analyze them and to collect facts about a broader expanse of the mobile wireless ecosystem,' she said. 'We were focusing on the data and the analysis rather than conclusions. We thought we would just lay out for the commissioners all the data and the analysis and stop there and that's what we did.'
So far, so good (I guess)--except for the cringe-inducing "ecosystem" malapropism (which seems to be pervasive over at the Commission, when referring to service supply chains).  I mean, I guess what the Bureau did was OK--laying out facts for the Commission to draw the conclusions required by law.  But, no one on the 8th floor ever bothered to render these conclusions.  Nonetheless, sloppy language breeds sloppy analysis.  In this case, if there was a logical analytic framework for the Report, it wasn't stated explicitly, nor was it possible to discern intuitively. In the next few posts, I'll provide my take on the Report overall--the Good, the Bad, and the Ugly.  

The Good: lots of information . . . a whole lot--maybe even too much information on too many "markets", but not enough information on the relevant market--the market for mobile wireless services. For example, the Commission seems eager to divide the supply chain into multiple submarkets--either on the input side, or with respect to adjacent, or "complementary" markets (which, except possibly for mobile applications, the Commission conflates with the term "downstream" markets).  The report may have been more accurate if the Commission divided the "mobile services market" into "upstream" and "downstream" components (an exercise that would have enabled the Commission to accurately describe the vigorous wholesale market, through which unaffiliated "MVNOs", purchase service to be packaged with unique handsets and offered to the consumer at retail).

Nonetheless, one of the things I really kind of liked was the attempt to provide information on the input markets facing the mobile service providers.  Unfortunately, this effort was probably too ambitious and, on deeper thought, it would have been impossible to execute on a meaningful basis.  Also, with respect to this portion of the Report, the Commission may have been better off leaving the sections out.  It is better to exclude incomplete data, than to include partial data, and then create a misleading narrative based on that data.  

Another interesting, and novel, feature of the Report is the Commission's effort to disaggregate service revenue, consumption, and growth trends by the differentiated service--like voice, texting, and data--notwithstanding whether the services were sold in the same way (i.e., text only, etc.).  This information was interesting, but not presented in a competitively meaningful way.  While the Report helpfully explains changes in mobile wireless service consumption trends, it only suggests, but does not attempt to demonstrate or quantify, probable positive cross-elasticity of demand (substitutability) between voice and text messaging, and possibly data.  

The Report theorizes that average voice consumption per user may have declined in the past year as a result of "per text" prices precipitously declining from 2 ½ cents per text to around a penny per text over the same period.  If voice and texting are indeed "close substitutes", then the Commission may want to refine its analysis in the next Report to include voice/texting as a single product market category.  This would suggest a more competitive market, as voice plan prices and per text prices continue to fall, pari passu.

On the data side, the Report notes that data revenue--as a percent of total revenue per customer--is expanding at an accelerating rate.  It is difficult to draw inferences regarding the competitive nature of the wireless mobile data market, because it appears to still be in its incipiency.

Finally, on the "Good" side of the Report, one cannot deny the work that went into collecting the information.  The professional staff of the Commission is truly to be commended.  One might disagree with the information the Commission chose to present, how the information was presented, and the conclusions the Commission would have the reader infer based on the information--but this is a Commission policy concern, and does not diminish my respect for the work that went into this Report.  

Telecomsense Presents Real Staff of Genius
[italics are to be sung in your cheesiest '80's band voice . . . think "Survivor"]

So, here's to you, Mr. and Ms. Wireless Competition Report Writers . . . for the past year, you've spent about 360, 000 minutes each to tell us that we've all been talking on the wireless phone for about about 8,500 minutes and sent about 4500 text messages--each, on average.

Here's to hoping--for your sake--the Commission can reel in its appetite for the
minutiae of minutes, messages, and handsets in next years' Report. 

Hungry, hungry, beaureaucrats . . . can we get a list of "exclusive"co-lors? Pink's my favorite!

 And take heart . . .  At least this year, they didn't require you to compile numbers on the "sexting" segment of the mobile data market . . . naughty, naughty sexters . . .

So, here's to you, oh collectors and crunchers of the numbers . . . it's time to turn off those pretty little smart phones  . . . this one has a Disco Inferno Radio Alarm Clock App! . . .,
 Close down the Word "Table Wizard" and crack open a cold Bud Light, because without your massive minutes of dedication, we wouldn't know how many minutes "unlimited" truly meant.

Mr. and Ms. Wireless Competition Report Writer . . .