Posts Tagged Viber

Top Ten Lines in the FCC’s Staff Analysis and Findings

Geoff Manne’s blog on the FCC’s Staff Analysis and Findings (“Staff Report”) has inspired me to come up with a top ten list. The Staff Report relies heavily on concentration indices to make inferences about a carrier’s pricing power, even though direct evidence of pricing power is available (and points in the opposite direction). In this post, I have chosen ten lines from the Staff Report that reveal the weakness of the economic analysis and suggest a potential regulatory agenda. It is clear that the staff want T-Mobile’s spectrum to land in the hands of a suitor other than AT&T—the government apparently can allocate scare resources better than the market—and that the report’s authors define the public interest as locking AT&T’s spectrum holdings in place.

Top Ten Lines in the Staff Report

  1. “While there is more to establishing likely competitive harms than measuring market and spectrum concentration, these [concentration] metrics shed light on the scope and scale of the competition that would be eliminated by the proposed transaction.” Staff Report, para. 17. An important admission. The staff is signaling that the merger analysis cannot begin and end with a concentration analysis. The Staff Report fails to explain, however, what more is needed to establish anticompetitive effects. The answer is direct evidence that the merging firms significantly constrain each other’s ability to raise prices. And the Staff Report fails on this score.
  2. “Second, the proposed transaction would result in the elimination of a nationwide rival that has played the role of a disruptive competitive force in the marketplace.” Staff Report, para. 17. Setting aside the weakness of the claim that T-Mobile—the only major carrier to lose subscribers in 2010—is a disruptive force, the Staff Report fails to explain how T-Mobile’s supposed disruption has anything to do with the instant merger. Is the staff saying that T-Mobile is so disruptive and so irreplaceable that any merger eliminating T-Mobile would be anticompetitive? The Staff Report’s “disruptive” evidence, chronicled from paragraphs 21 through 28, could be regurgitated in a Sprint/T-Mobile merger review or in a Leap/T-Mobile review. Would those mergers be presumptively anticompetitive as well? Critically, the evidence of T-Mobile being a disruptive force does not speak to the issue of whether T-Mobile constrains the price of AT&T.
  3. “Market concentration statistics of the type generated by this transaction commonly indicate that buyers would have fewer viable choices, making both unilateral and coordinated competitive effects more likely.” Whether concentration statistics indicate anticompetitive effects in general or in a hypothetical U.S. market is beside the point. What matters here is whether concentration statistics are a good predictor of higher wireless prices. And the answer is a resounding no. As the former chief economist of the FCC has noted in a forthcoming paper, wireless concentration is negatively correlated with wireless prices. At a minimum, the FCC should note this finding—the abstract has been viewed nearly 1500 times and the full paper has been downloaded 250 times from SSRN—and provide the proper caveats to any concentration analysis they conduct regarding the wireless industry.
  4. “Although T-Mobile faces challenges as the industry develops and responds to the increasing data demands of consumers, the record does not support the bleak short-term outlook for T-Mobile that AT&T has portrayed in its submissions.” Staff Report, para. 22. To say that T-Mobile faces challenges is an understatement: T-Mobile is uniquely losing subscribers and its German owners want out of the U.S. market. How can the agency better predict the short-term outlook for T-Mobile? Is there a crystal ball in the FCC’s basement? If the short-term were as rosy as the agency suggests, then why would T-Mobile’s owners—who presumably have the best vantage on the firm’s future performance—seek a buyer right before a turnaround in performance?
  5. “These initiatives [announced by T-Mobile’s CEO before the transaction] might have strengthened T-Mobile’s disruptive role in the industry, for example by highlighting its unlimited data plans, and using them to define its brand and differentiate it from rival brands that have adopted tiered pricing.” Staff Report, para. 23. How can T-Mobile go from a disruptive force to an even stronger disruptive force? You can’t be half-pregnant, and you can’t be half-disruptive. It seems that the Staff Report is now saying that T-Mobile would have been disruptive but for the transaction, which caused T-Mobile to abandon these really stupendous plans. According to footnote 61, these initiatives were announced in a T-Mobile press release on January 20, 2011. But the agency doesn’t bother to see how the market reacted to these initiates. It is curious that the agency would stake its disruptive claims on something so speculative.
  6. “T-Mobile has also repeatedly acted as a pricing innovator over the past few years, introducing offers such as . . .  T-Mobile introduced a simple online tool that allows a subscriber to manage all services on a multi-line family plan, for example, setting and changing the limits for minutes, messages and downloads (e.g., games, ring tones) on a child’s line.” Staff Report, para. 24. According to the Staff Report, this “innovation” is among the seven most disruptive offerings from T-Mobile since 2007. Seriously? Is this impressive to anyone out there? Even assuming T-Mobile was the first to allow wireless users to adjust their settings online, how in the world did that constrain AT&T’s ability to set prices? The other innovations cited in the Staff Report are equally unimpressive. How well did that Unlimited Hotspot Calling or T-Mobile Hotspot @Home work out? If none of the major carriers embrace an offering like those, can we safely infer that they weren’t so innovative? If you want to make free Wi-Fi calls on your phone, download Viber. Yawn.
  7. “[O]ur analysis of the record reflects that T-Mobile charges lower prices than the other nationwide firms.” Staff Report, para. 25. Apparently, the staff doesn’t want you to know that T-Mobile had its legs cut out by regional carriers such as Leap and MetroPCS. Indeed, T-Mobile’s executives have admitted as much publicly, explaining how it was caught between the high-end service of AT&T and Verizon and the low-end, no-frills service of Leap and MetroPCS. And no firm wants to be caught in the middle of the road. Speaking of being caught, the staff should not offer such misleading statistics. To make things concrete, in Washington D.C., T-Mobile offers a $39.99 per month plan that includes 500 minutes of voice and no text messages. In comparison, Leap offers a $35 per month plan that has unlimited voice minutes and includes text messages. But the Staff Report wouldn’t count Leap’s offering because Leap is not a “national” carrier, despite Leap’s offering wireless service in 35 states covering 100 million people.
  8. “T-Mobile expressed interest [in selling wholesale access to Cablevision], had previously exhibited a willingness to sell wholesale mobile wireless capacity, and, in Cablevision’s view, was likely to continue to have excess capacity it could use to serve Cablevision’s customers in the future. Although the outcome of any negotiation is uncertain, a deal between Cablevision and T-Mobile appeared to be beneficial to both parties.” Staff Report, para. 28. The staff here wants us to believe that in addition to the proposed merger undercutting T-Mobile’s initiatives to revamp the firm, the proposed transaction would undercut a prospective deal with Cablevision that would ostensibly bring benefits to Cablevision’s customers in parts of New York, New Jersey, and Connecticut. Well now this all makes sense: Stop a merger that could generate benefits to AT&T’s and T-Mobile’s nationwide customers to preserve Cablevision’s option to offer a quad-play to its customers in three states. Cudos to the cable lobbyists for getting their client’s concerns front and center in the FCC’s merger analysis. Setting aside the uncertainty surrounding the actual wholesale discussion that Cablevision and T-Mobile may or may not have entertained, the Staff Report suggests incorrectly that Cablevision depends uniquely on T-Mobile for spectrum, and that Cablevision’s customers would benefit significantly from having a sixth or seventh wireless option. As further evidence of how out of touch the Staff Report is from market realities, Verizon just announced that it was purchasing all of the AWS spectrum held by several cable companies, a market reality inconsistent with the staff’s views that T-Mobile’s innovative future lay in partnerships with cable companies.
  9. “Combined, these five regional providers accounted for approximately six percent of the industry’s total subscribers and revenues at the end of 2010. None of these providers’ networks cover more than 34 percent of the U.S. population, and for most their more advanced broadband networks are smaller.” Staff Report, para. 38. Because these regional providers do not have the potential to serve 100 percent of the U.S. population, it makes no sense to denominate their size in terms of nationwide subscribers. Doing so necessarily understates their importance in the local markets they serve. By way of analogy, Comcast’s in-region share of video subscribers or “video penetration” is roughly 44 percent, whereas its share of nationwide video subscribers is roughly 25 percent. Of course, the latter statistic bears no relation to Comcast’s pricing power. Moreover, while Leap or MetroPCS alone do not cover a majority of the nation, their roaming agreement (and complementary footprints) allows each firm to provide nationwide coverage. Again, the Staff Report appears to be playing fast and loose with the data.
  10. “AT&T’s unilateral incentive to raise price in this case arises because providers sell differentiated products, and many of AT&T’s customers view T-Mobile as their second choice at current prices. . . . Local number porting data (data on where customers go when they switch wireless providers while keeping their phone number) indicate that each of them [the major carriers] has customers who view T-Mobile products as their second choice.” Staff Report, para. 50. What does “many” mean in this context? And what does it mean to have at least some customers who switched to T-Mobile? Could “many” AT&T customers mean five percent? Any share less than T-Mobile’s probability-adjusted market share of roughly 16 percent (equal to T-Mobile’s share divided by 100 less AT&T’s share) would not be evidence of significant cross-price elasticity between AT&T and Sprint. The Staff Report later defines “many” as “a non-trivial fraction of AT&T’s customers.” But why is the standard so low? Later the Staff Report claims that “a substantial fraction” of AT&T customers switched to T-Mobile, and did so “in response to changes in the relative price of T-Mobile products and the introduction of new T-Mobile products.” Setting aside its loose standard (from “many” to “non-trivial” to “substantial”) of economic significance, porting data cannot tell you why a customer switched from one carrier to another. To assess cross-price elasticity, one must estimate an econometric model using customer-level wireless bills, which the Staff Report does not do. Finally, to bolster its evidence of cross-price elasticity, the Staff Report cites a T-Mobile “Losing Your Shirt” advertising campaign targeting AT&T’s customers. That T-Mobile aspired to attract AT&T customers does not constitute evidence that T-Mobile actually disciplines AT&T’s prices. Many computer companies aspire to topple Apple, but that doesn’t make it so.

As you digest these criticisms, think of how an economic expert could defend these statements upon cross examination. Although the authors of the report will never be subjected to such an exam, it is a bit surprising that such bald and unsupported statements could survive the cutting-room floor.

, , , , , , , ,

Leave a comment

Debunking the New York Times Editorial on Wireless Competition

Yesterday, the editorial page of the New York Times opined that wireless consumers needed “more protection” than that afforded by voluntary agreements by the carriers and existing regulation. The essay pointed to the “troublesome pricing practices that have flourished” in the industry, including Verizon’s alleged billing errors, as the basis for stepped up enforcement. As evidence of a lack of wireless competition, the editorial cites several indicia, none of which is persuasive.

Let’s address each indicator in turn. (For readers interested in a more rigorous analysis of wireless competition, see the recent paper I co-authored with Gerry Faulhaber and Bob Hahn, forthcoming in the Federal Communications Law Journal.)

NYT: “Most customers, locked into their contracts by high early-termination penalties, have no easy way to switch providers.”

Although many wireless contracts contain early-termination fees (ETFs), the fees generally decline as the contract nears expiration. For example, AT&T’s fee begins at $325 and declines by $10 each month. One year into the contract and the fee becomes a fancy dinner date and a movie (not including the popcorn). The editorial fails to recognize that ETFs are a substitute for handset subsidies by the carriers: Limit ETFs and those subsidies will decline. Moreover, customers who are highly adverse to ETFs can either pay the full freight of the handset under a post-paid plan or enter into a pre-paid plan (and again pay full freight). I am happy to make a commitment—the prospect of an ETF makes my commitment credible—in exchange for a big discount on the latest iPhone.

NYT: “And two companies — Verizon and AT&T — now control 60 percent of the market nationwide.”

According to the FCC’s 14th Annual report (Chart 1), the combined shares of AT&T and Verizon are 55 percent, not 60 percent. Setting aside that rounding error, barring evidence of coordinating pricing by AT&T and Verizon, it makes no sense to add together their respective shares when assessing market power. To the extent that market shares have any meaning, Verizon’s market share alone (and not the sum of Verizon’s and AT&T’s share) informs Verizon’s market power. The same is true for AT&T. But as demonstrated by the Faulhaber paper, market shares have no predictive power of wireless prices (and hence market power), whether across time or across geography at a given point in time.

NYT: “Take cellphone text messaging. Companies typically charge from $5 for 250 texts a month (2 cents per message) to $20 for an unlimited package. Pay-as-you-go rates can be as high as 20 cents a message. But the cost of sending a text message is about a third of a penny, according to Congressional testimony from Srinivasan Keshav, a professor at the University of Waterloo. The markup is enormous.”

Like wireless voice services, the price of text messaging is falling. According to Nielsen (2011), the effective price per text message declined from six cents to a penny from 2005 to 2010. If one cent per message is not the right price, what is? Of course, the margins on any service in a network industry with steep fixed costs will be large. But they need to be large to induce the operator to take the risk of building the network in the first instance. Moreover, savvy wireless users with a data plan can download applications that replicate the functionality of text messaging for free. A search for “SMS text messaging for free” from Apple’s App Store yields myriad hits, beginning with textPLUS, Textfree, Fake-a-Message, Text-for-Free, Text Me!, Messagey, Fake Text Free, iText, and Texter. So long as carriers allow users to install these apps, we should expect text messaging prices to decline further.

NYT: “Even the most expensive monthly wireless data plans, costing about $15 for 250 megabytes or 6 cents per megabyte, are orders of magnitude cheaper than cellphone text pricing.”

For the reasons explained above, the days of surcharges for texting appear to be numbered. As these free-texting and other applications such as Viber (free Internet-based calls) increasingly cannibalize the carriers’ basic offerings, difficult policy issues emerge, like whether the carriers should be allowed to decide which applications can be used on their systems. Limiting a carrier’s ability to exclude such applications (or even charge for them) might put upward pressure on the price of the data plan.

NYT: “It is hugely profitable for companies to segregate voice, data and text into different plans and to force customers to buy a different plan for each device, like a phone or a tablet. But, on today’s networks, segregating services makes little sense technologically. This expensive segregation would be more difficult to maintain if the market were truly competitive and consumers could easily switch from one company to another that offers a better deal.”

Setting aside the inconvenient fact that voice, data and text are different services from the perspective of customers, why not look at switching data directly to see whether customers are trapped? Churn refers to the percentage of wireless customers who depart for other carriers. Assuming that most departing customers find a new wireless home, churn data are a reasonable proxy for switching data. In the first quarter of 2011, U.S. carriers reported monthly churn rates between 1.3 (for pre-paid customers) and 4.3 percent (for post-paid customers). Ignoring customer additions, a carrier that begins the year with 10 million customers and experiences a 2 percent monthly churn ends up losing 2.15 million of those original customers (roughly 21 percent) over the year. By this measure, wireless customers are hardly locked into place.

Finally, in addition to asking the FCC to curtail ETFs on phone contracts, the editorial requests that additional spectrum be made “available to more competitors.” I interpret “more” to mean “carriers that do not rhyme with KT&T or with Spurizon.” If only there was no such thing as spectrum exhaust, the social costs of excluding AT&T and Verizon from the next auction would be trivial. But alas, as acknowledged by the FCC’s National Broadband Plan and demonstrated in a new report by Peter Rysavy, existing carriers need access to at least 500 MHz of additional spectrum to support the precipitous increase in demand for bandwidth-intensive wireless applications. Yet another new study by the Global Information Industry Center at the University of California, San Diego shows wireless demand outstripping capacity shortly. When licensing any future spectrum, the FCC must carefully balance the need to ensure quality of service for existing subscribers against the speculative benefit of adding a sixth or seventh local carrier—90 percent of the population is served by at least five carriers according to the FCC’s 15th Wireless Report—to an already competitive wireless environment.

, , , , ,

Leave a comment