Abstract
The Federal Communications Commission can improve the transparency of its decision process by establishing a set of guidelines for its review of mergers. Its current process is opaque, and parties cannot discern the weights that it places on “public interest” goals in some areas and how those balance against harms in other areas. The antitrust agencies’ Merger Guidelines provide an example of how the Commission could signal more clearly its analytical process and push for more efficient remedies and lessen political pressure to undertake inefficient trade-offs such as requiring build out to areas unaffected by changes in competition due to a merger.
I. Introduction
In late 2011, AT&T withdrew its attempt to purchase T-Mobile. That proposed transaction received significant and detailed public scrutiny from the Department of Justice (DOJ) and the Federal Communications Commission (FCC or Commission). The DOJ detailed some of its analysis in a complaint to enjoin the transaction, and the FCC issued a 150-page staff report discussing the likely competitive effects of the transaction. 1
Since that time, there have been at least six significant wireless transactions (and several other license transfers) that have generated DOJ review and Commission decisions: (1) Verizon–SpectrumCo; (2) T-Mobile–MetroPCS; (3) Sprint–SoftBank–Clearwire; (4) GCI–ACS (in Alaska); (5) AT&T–ATNI; and (6) AT&T–Leap. 2 The Commission and DOJ approved all of those transactions, some with divestitures of spectrum licenses. While none of them appear to present the same change in competition contemplated by the proposed AT&T–T-Mobile transaction, the Commission’s decisions are substantially less forthcoming about the details of its competition analysis undertaken in reviewing the transactions. The clear competition analysis in the Staff Analysis of proposed AT&T–T-Mobile transaction should lead the FCC to develop a public merger guideline framework akin to the DOJ/Federal Trade Commission (FTC) Merger Guidelines so that firms will have more certainty about which transactions will be judged to promote the public interest and which will not. Under the current decision-making rubric, justifications are vague, and transactions appear to be approved on a much more ad hoc basis.
A key trade-off for wireless mergers is the possibility of increased efficiencies that could reduce prices and increase output, compared to the possibility for a reduction in competition that could increase prices and reduce output. Cost savings in the wireless industry can come about from reductions in marginal costs resulting in price-reducing effects, 3 from a saving in marginal capital costs in an industry that is growing and needs to increase its capacity, and saving in traditional fixed costs (such as reductions in infrastructure costs, which are typically not credited in traditional merger analysis because they rarely result in lower prices or some other benefit to end users).
In the AT&T–T-Mobile transaction, a large part of AT&T’s justification for the merger was savings in incremental capital costs from combining the two networks and spectrum holdings. The Commission recognized that there would likely be some efficiencies from the combination but determined that the magnitude of those efficiencies was not as great as the parties claimed. After correcting the level of efficiencies, the Staff Analysis found that the merger simulation put forth by the parties would lead to higher prices with a more reasonable estimation of the efficiencies and recommended that the merger be designated for hearing. 4
In the subsequent transactions, the Commission’s decisions have not provided guidance on how it would think about a merger where the trade-off between efficiencies and a threat to competition is less clear. For example, in many of the subsequent transactions, the trade-off has been expected harm to competition in some areas and expected benefits in others. In these subsequent decisions, the Commission discusses its “sliding scale” approach, where it “employ[s] a balancing test” of public interest benefits and harms. Such a “balancing” would better be termed an assessment of the costs and benefits of the transaction, with explicit discussion and measurement where possible. Instead, none of the decisions provide a mechanism to understand the measurements or weights placed on the benefits and harms.
In this article, we propose that the Commission regularly incorporate a more transparent and analytical framework for merger analysis than its vague balancing and sliding scale approach. While the outcomes of specific transactions may be no different, the framework would provide a greater degree of objectivity and transparency for merger evaluation and more information for those contemplating future transactions. For the most part, the Commission’s competition analysis should follow more closely the approach taken by the antitrust agencies, both by laying out the parameters of the analysis and by taking a thorough and complete analysis.
However, the FCC argues that its purview is somewhat broader than the antitrust agencies and, as such, should have a different standard. 5 In part, the FCC’s “public interest” standard is vague so that articulation of a framework would provide more certainty to the marketplace. The FCC’s governing statue is different than the antitrust laws—under the antitrust law, the burden of proof falls to the government to show that a merger would harm the public interest; whereas under the FCC jurisdiction, the burden is on the parties to show that the transaction will be in the public interest. In addition, the Commission’s industry expertise can add great value to the standard competition analysis. However, the Commission should make clear how it will complement its competition analysis with additional public interest findings. For example, how will it trade off “benefits” in one area for “harms” in another?
Section II provides a background on the typical Commission transaction analysis. Section III looks at the Commission’s analysis in specific transactions. Using this information, Section IV proposes a framework for the Commission to apply to future mergers that will provide more guidance and a clear analytical framework to evaluate transactions in light of the Commission’s goals.
II. FCC Standard of Review and Process
Market Definition
The Commission has evaluated wireless mergers using a combined “mobile telephony/broadband services” product market. Essentially, the definition encompasses mobile voice and data service. No party challenged the product market definition in any of the proposed mergers. It is possible that sufficiently well-developed Wi-Fi service could in the future cause the market definition to incorporate such services, but at present, Wi-Fi does not appear to provide sufficient price pressure on commercial services.
The Commission evaluated each of the mergers on a local as well as a national basis. Local markets are justified on the basis that service in Washington, D.C., is not a substitute for a person living in Palo Alto, California. Thus, the relevant choice set for consumers is inherently local. A national product market definition is also used because of the nature of product marketing in the industry—to the extent wireless prices and packages are set on a national basis for the larger carriers, competitive forces in large localities can affect prices in other smaller localities. 6
Finally, the Commission looks at the “input market for spectrum” because spectrum is a key input to mobile wireless service. The FCC has adopted a “spectrum screen” to identify transactions that might lead to concern about the concentration of spectrum holdings and, in particular, holdings of spectrum suitable to provide mobile wireless service. The FCC uses its screen in general to identify areas where a transaction would lead to a company controlling more than one-third of the suitable spectrum. At the present time, the FCC has “determined cellular, PCS, Specialized Mobile Radio (‘SMR’), and 700 MHz band spectrum, as well as AWS-1 and Broadband Radio Service (‘BRS’) spectrum where available” are “suitable” for providing mobile telephony/broadband service. 7
The FCC initially used a hard spectrum cap—with two cellular licensees in the 1980s and early 1990s, no company was allowed to own both licenses in the same geographic area. With the introduction of personnel communications services (PCS) via auctions in 1994, the FCC implemented a 45-megahertz (MHz) cap on the amount of Commercial Mobile Radio Spectrum (CMRS) so the two incumbent cellular licensees in each area could not buy a large PCS license covering the same area, guaranteeing new entry. The Commission, under Chairman Powell, eliminated the spectrum cap in 2003. A hard cap would provide guidance for companies but reduce flexibility.
The advantages of a more well-defined rule for spectrum ownership is not limited to mergers. There is a strong argument for a clear rule for auctions, where many companies are competing for spectrum and there is the ability to pursue backup strategies with multiple licenses available at the same time.
Competitive Effects
In analyzing a proposed merger, the Commission generally begins its competition analysis using its spectrum screen. In areas where it determines that the aggregation of spectrum will be greater than one-third of the available spectrum, it looks at whether the transaction would increase “the likelihood that some competitors or potential entrants would be foreclosed from expanding capacity, deploying 4G technologies, or entering the market.” 8 In addition, the Commission examines whether the transaction would raise rivals’ costs by precluding their access to spectrum for capacity expansion or introduction of new services.
The Staff Analysis of the proposed AT&T–T-Mobile transaction moves from the spectrum screen to discuss the likely unilateral and coordinated competitive effects of the proposed transaction in detail and uses the competitive effects merger simulation model from the applicants to show price increases resulting from the merger.
The FCC also looks at the impact of mergers on roaming and interoperability. Roaming is a particular concern for carriers with smaller footprints. Without wide geographic coverage, these providers would likely be at a competitive disadvantage. However, incentives to invest in wide area coverage may be diminished if firms were required to provide roaming to their competitors at regulated prices below a market rate.
Efficiencies
Efficiencies are used by merging parties to justify most mergers and are frequently an important justification for the price paid to acquire the assets. In a public interest analysis, efficiencies are critical to understanding the overall impact of a merger. The FCC follows the DOJ/FTC guidelines in requiring that the claimed efficiencies be “verifiable, transaction-specific,” and then assesses whether they outweigh any identified competitive harms.
In its analysis, the Commission follows standard antitrust guidelines to examine whether efficiencies would occur without a merger and focuses primarily on marginal cost reductions, as those would be the type of efficiencies likely to lead to reductions in prices to consumers. In the AT&T–T-Mobile transaction, the FCC staff report examined carefully the quantitative estimates of claimed network efficiencies from combining spectrum and networks.
“Balancing”
Based on its findings regarding potential competitive harm and possible welfare increasing efficiencies, the FCC has developed a “sliding scale approach” to evaluating benefit claims. 9 In essence, the Commission finds that the more likely and substantial a competitive harm, the more likely and substantial the offsetting benefits should be. It is unclear why the FCC uses the term “sliding scale” rather than a comparison of expected costs and benefits. The use of expected values and confidence intervals is common in policy analysis and can be used to describe the evaluation exercise more clearly and put it into more of an analytical framework.
At the FCC, the applicants bear the burden of showing that the claimed efficiencies will outweigh any potential public interest harms. 10 The standard under the Communications Act and Commission precedent is different from the antitrust laws, where the DOJ would bear the burden to show that the likely competitive harms would outweigh the competitive benefits from merger-specific efficiencies if it were to challenge a merger. In contrast, the Commission may consider whether a transaction will enhance, rather than merely preserve, existing competition, and can take a more expansive view of potential and future competition in analyzing that issue. 11
In its balancing evaluation, the FCC generally looks at potential competitive harms in some areas and compares those to the efficiencies or other factors it considers public interest benefits. As a result, it is possible that a merger would harm some people and benefit others. 12 The potential for weighing harms to one group against benefits to another is the way the FCC merger review differs most from the antitrust agency review. 13 In contrast, at the FTC and DOJ, merger remedies are meant to remedy the specific anticompetitive problem raised by the merger so no group of end users is harmed by significant anticompetitive effects of the merger.
Other Commitments
In many transactions, the FCC looks at “voluntary commitments” by the parties to see if those will tip the balance toward a conclusion that the public interest benefits of a transaction outweigh the likely competitive harms. Two commitments are prevalent in the decisions reviewed below: build out and roaming. Companies have committed to a more rapid build out of advanced wireless service as a means of getting credit for increasing public interest benefits from a transaction. However, increasing build out may be an inefficient use of resources and may not do anything to counteract the competitive harm.
It is possible that additional build out may increase social welfare. For example, companies may not build out to an area if it is not privately profitable; they may not find it profitable to charge a higher price in a more costly area because of nationwide marketing plans. However, there may be consumers who would realize consumer surplus in excess of the build out costs. 14 The Commission has not made such an explicit argument or attempted to measure the costs and benefits of mandating build out conditions to see if the argument has any factual foundation. 15
Roaming commitments are a mechanism to alleviate concern for other wireless companies that rely on roaming. Typically, roaming allows smaller networks that cannot profitably build out their coverage to less dense areas to provide service in these areas so customers of these networks have coverage in more areas. However, agreements to provide roaming in one area may not necessarily address competitive concerns, especially in specific local markets in other areas.
The ad hoc nature of other public interest benefits was clear in the analysis of the AT&T–T-Mobile transaction, where the parties claimed they would not build out as rapidly absent merger approval and that they would, in turn, create a large number of new jobs if the transaction were approved. 16 There are two potential benefit claims here. First is the increase infrastructure deployment, which is clearly in the purview of the FCC. With respect to this benefit, the staff analysis found that the incentive to build out was supplied primarily by the need to compete with Verizon, which was also building out its Long-Term Evolution (LTE) network, rather than from merger efficiencies. Thus, this potential benefit was not merger specific.
The second potential benefit was the claimed increase in jobs. It is not at all clear that the FCC has any expertise in deciding to what degree a merger would change the national unemployment rate or how beneficial such a change would be—there are many ways the government can increase spending to increase employment. If it would not be profitable for a carrier to create jobs on its own, then it is very likely that it would be inefficient for society to create such jobs. It is not clear why this particular likely inefficient method of funding jobs would be better than some other mechanism the government could employ. In other words, why is it good public policy to tax the telecommunications sector to support a nationwide jobs initiative? In this case, the FCC did not have to opine on the benefits of increased jobs. Since AT&T’s claimed job increase was premised on increased build out, and increased build out itself was not a merger-specific benefit, the claimed increase in jobs, whether or not it would have been a net benefit to society, was not a merger-specific benefit.
As a result, the FCC should have a clear theory of what benefits come from “voluntary” commitments, how enforceable the commitments are, and the value of the benefits.
III. Specific Transactions
Verizon–SpectrumCo
In late 2011, Verizon and SpectrumCo. proposed a transaction in which Verizon would acquire wireless licenses from SpectrumCo., Cox Cable, and Leap Wireless, and transfer one license from Verizon to Leap. The wireless license transfers from SpectrumCo were part of a larger agreement whereby the two parties would also market wireline and wireless services jointly. In mid-2012, Verizon Wireless filed applications to transfer some spectrum licenses to T-Mobile.
Verizon Wireless was a joint venture between Verizon Communications Inc. and Vodaphone Group, Plc. Verizon is one of the two largest wireless carriers in the United States, measured by either subscribers or revenue, and provides near nationwide coverage. It was in the process of expanding the reach of its 4G LTE network across the country.
Leap Wireless was a low-price wireless carrier operating in many major areas across the country over its own network and more through other agreements.
SpectrumCo was a joint venture of cable companies Comcast, Time Warner, and Bright House (and originally Sprint Nextel, but it sold its interest to the cable companies) that bid on and won licenses in the FCC’s 2006 auction of Advanced Wireless Services (AWS) spectrum licenses. Cox Cable also owned some AWS license.
T-Mobile was the fourth-largest wireless carrier, with near nationwide coverage.
Competition overlaps and concerns
SpectrumCo and Cox purchased AWS licenses in 2006 but had not begun to provide commercial wireless service using the spectrum as of the time of the transaction. The cable companies had some Wi-Fi hotspots but no real commercial mobile service. As a result, there was no competitive overlap for existing service. However, some parties expressed concern about the elimination of potential competition. There was the possibility that the cable companies could enter with their spectrum and networks, but the FCC deemed that to be a low-likelihood event, and such entry, were it to occur, would not occur in the foreseeable future.
In addition to the spectrum transaction, Verizon Wireless and the cable companies agreed to sell each other’s services, even within the Verizon Communications landline territory. 17
T-Mobile was one of the parties initially objecting to the transaction, presumably because it had a set of AWS licenses and would have liked to acquire the SpectrumCo and Cox AWS licenses to complement its network. Subsequently, it withdrew its objections after Verizon Wireless agreed to transfer AWS spectrum licenses to T-Mobile. T-Mobile asserted that the transfer would alleviate its competitive concerns.
The FCC also looked at the concentration change for Verizon Wireless’s holding of AWS-1 spectrum. The FCC was concerned “that the AWS-1 spectrum at issue in these transactions is the lone large block of currently unencumbered near-nationwide spectrum with a well-developed ecosystem immediately available for the provision of mobile broadband service.” 18 The FCC concluded that the lack of greenfield AWS-1 spectrum would increase costs for certain rivals. 19 The FCC did not provide any mechanism for understanding the magnitude of such cost increases for rivals or the incentive to foreclose on this particular block of spectrum. 20 However, it did conclude that the acquisition of AWS-1 spectrum “in numerous local markets causes significant competitive concerns.” 21
The FCC concluded that in some geographic markets identified by the spectrum screen there was potential for competitive harm from the transaction. 22 However, the FCC concluded that the areas were small and few in number, so that any reduction in competition would not affect the nationwide price. However, it was possible that competition in those areas would be affected, as carriers not only compete on price but also on network quality. It would be possible for a carrier not to increase price but to invest less in network quality, effectively increasing the quality-adjusted price. In addition, the FCC asserted that the concentration of spectrum could affect enough local markets to cause a nationwide price increase without the divestiture of AWS-1 spectrum to T-Mobile.
The FCC concluded that the transfer of AWS-1 spectrum to Verizon without divestitures “would constitute a concrete potential harm to future competition.” 23 The order provides no additional explanation about the harms or why roaming would be different with a divestiture to an existing provider. It is possible that the FCC determined that without the AWS-1 spectrum, T-Mobile would not have provided wire-area LTE service, and that the divestiture to T-Mobile would create an additional potential roaming partner for small regional carriers. However, the FCC does not make its analysis explicit, so it is impossible to evaluate this argument.
Efficiencies
In each of the transactions, the parties asserted that the license transfer would lead to more efficient use of spectrum. The most concrete example was the move of spectrum from the SpectrumCo and Cox, who were not using it, to Verizon, who would deploy it immediately. In other transactions, the parties argue that the spectrum rationalization would lead to more rapid deployment of LTE technology. 24
The FCC determined that Verizon had not demonstrated that it would use all of the spectrum it acquired to provide LTE service in the near term. In fact, the FCC argued that “documents do not indicate that Verizon Wireless would need to deploy more than 40 megahertz of AWS-1 spectrum in any of these markets to meet capacity demands.” 25 Because the FCC found that Verizon would not need to use the spectrum, and in fact could implement other capacity-enhancing technical solutions, the benefits from the merger would not be realized. The FCC determined, however, that it could credit the benefits because Verizon “has committed to undertake an aggressive build-out schedule of the spectrum it is acquiring through these transactions.” 26
The FCC’s justification for crediting the efficiencies is suspect. For example, if the efficiency benefits were equal to the benefits from adopting small cell technology, or even slightly less good, a firm might commit to build out if indeed it had the ability to foreclose competition. More importantly, build out commitments are not efficiencies—they are committing to use a resource. In some instances the FCC seems to like using spectrum and technical efficiency, but those two measures are not efficiencies in the antitrust sense and do not offset any possible increase in price. The justification for counting the benefits in this case is not based on sound economics but seems to be a justification for the FCC to favor build out, which in turn may lead to inefficient spectrum use.
More specifically, one would expect competition to provide an efficient network build out time horizon. Spectrum Co and Cox had no plans to deploy the AWS-1 before the transfer. After the transfer, Verizon had plans to deploy some of this spectrum, and it was required to divest spectrum it might not have deployed immediately to T-Mobile, who was likely to deploy it rapidly. Thus, the speed of deployment posttransactions was greater than the speed of deployment in a market without competitive concerns. Thus, it is not clear that it is efficient for Verizon to accelerate its deployment schedule. Certainly the quality of service would be higher as a result of the accelerated deployment. But it is not clear that the benefits of the accelerated deployment outweigh the costs.
Outcome
The FCC found that there would be competitive harm absent the transfer of spectrum to T-Mobile. It found that absent the transfer, it had concerns of spectrum warehousing and foreclosure of competition. It concluded that the transfer of AWS-1 Spectrum to T-Mobile “mitigates our concerns of harms from spectrum concentration, and further this assignment in itself has significant public interest benefits.” 27
The FCC decision rests in part on Verizon’s “voluntary” build out commitment. From an economic perspective, build out requirements are not an efficient mechanism for promoting competition. In fact, in many instances they could frustrate efficient competition and waste resources.
T-Mobile–MetroPCS
In late 2012, T-Mobile agreed to purchase MetroPCS. T-Mobile at the time was the fourth-largest wireless provider and had acquired additional AWS licenses as a result of the Verizon-SpectrumCo transaction discussed above and as part of the breakup fee from the failed transaction with AT&T. MetroPCS was the fifth-largest provider of wireless service. MetroPCS’s network covered about one-third of the U.S. population. MetroPCS was focused on no contract, month-to-month, fixed-price service.
Competition overlaps and concerns
The FCC started its analysis with the same product and geographic markets: mobile wireless combined with local and national service territories. It used its spectrum screen to identify nineteen Cellular Market Areas (CMAs), covering approximately 36 million people, or 12 percent of the population. The nineteen CMAs included thirteen in the top one hundred markets. 28 The FCC focused its competition analysis in these nineteen areas based solely on the spectrum screen, not on subscriber shares.
In the competition analysis, the FCC repeats the two paragraphs describing unilateral and coordinated competitive effects.
29
Then it describes in vague detail its competitive analysis: Discussion: In undertaking a market-by-market analysis of the 19 local markets identified by our initial screen, we consider competitive variables that help to predict the incentive and ability of service providers to successfully restrict competition. These competitive variables include, but are not limited to, the total number of rival service providers; the number of rival firms that can offer competitive nationwide service plans; the coverage of the firms’ respective networks; the rival firms’ market shares; the combined entity’s post-transaction market share and how that share changes as a result of the transaction; the amount of spectrum suitable for the provision of mobile telephony/broadband services controlled by the combined entity; and the spectrum holdings of each of the rival service providers.
30
Efficiencies
The FCC concluded that the transaction would facilitate the deployment of LTE and accelerate the provision of mobile broadband services. The Commission found that existing customers would benefit from a more robust national network than would be possible were the two companies to operate separately.
While not discussed in the submissions or the decision, Bernstein analyst Craig Moffet described how the network efficiencies would likely arise from the combination. He discusses how new MetroPCS customers would be served with the T-Mobile network. Because of the high churn of MetroPCS customers, its network would shed customers rapidly and put new MetroPCS customers onto the T-Mobile network so that there would only be a small number of remaining customers after eighteen months to transition to the T-Mobile network. At that point, the MetroPCS spectrum would be clear and able to be refarmed and used for new LTE service.
The FCC found that because the combination would still be the fourth-largest competitor, it would retain the “maverick” tendencies the two companies had, and the two Florida markets with possible anticompetitive outcomes would not outweigh the public interest benefits from the efficiencies. However, neither the magnitude of the potential competitive harms nor the benefits from the efficiencies were discussed in the order. As a result, it is impossible to know if a merger with four or forty markets with potential competitive harm and the same efficiencies would be approved. 33
The Commission approved the merger in March 2013.
GCI–ACS
In June 2012, two of the three mobile wireless providers in Alaska proposed to merge. GCI wireless had about 140,000 subscribers in Alaska; operated local wireline telephone service covering about 35% of the homes; and was the primary cable television provider in Alaska, passing 90% of homes in the state. ACS Wireless had about 118,000 wireless customers in Alaska and also operate four local exchange companies and a long-distance provider. AT&T was the third wireless operator in Alaska. In addition, Verizon was entering the mobile data market in Alaska but did not provide mobile voice service, although it had plans to provide voice services upon the introduction of VoLTE, expected in early 2014.
The spectrum holding in Anchorage (with 41% of Alaska’s population) would be GCI/ACS, 125 MHz; AT&T, 135 MHz; and Verizon, 22 MHz. In addition, other entities had spectrum licenses, had not built networks, and were not using the spectrum: Sprint, 92 MHz; T-Mobile, 10 MHz; MTA Wireless, 40 MHz; and Triad 700, 12 MHz. 34
Competition overlaps and concerns
The parties proposed to combine their wireless holdings into a joint venture that operated a wholesale network while maintaining separate retail operations.
As a result, the Commission found that there would be the potential for unilateral competitive effects from the merger. However, because of the presence of AT&T and the entry of Verizon, and the extent of available spectrum, the Commission found that the potential for unilateral effect would be small.
For example, each firm offered an “Alaska-only” plan that was substantially less expensive than the nationwide plans offered by AT&T. Presumably the two firms competed with each other on these plans and the Commission found that there would likely be harm, but because “the Alaska-only plans appear to be held by a much smaller segment of the parties’ overall subscribers, any likely harm would not be as significant.” 35 The Commission does not provide the comparison for “as significant,” so we do not know how significant the competitive harm might be. The Commission required the parties to maintain their existing plans for two years (although this does not obviate the potential harm of prices not decreasing as fast as they otherwise might). In addition, it is possible that future customers might benefit from Alaska-only data plans but would have less competition as a result. One could imagine the data users would be willing to use an Alaska-only plan if they travel infrequently and could make use of Wi-Fi while traveling in other parts of the country or pay roaming fees.
The FCC makes much of the low population density of several CMAs. 36 While such areas provide fewer customers to which to sell service to cover the fixed costs of facilities, it is not made clear in what sense a merger provides an efficiency in these specific areas. For example, if both carriers have facilities in these areas premerger, in what sense will the merger add to service in these areas postmerger? If just one carrier has facilities in an area, why would the merger change the chosen level of coverage? If neither carrier found it profitable to serve an area premerger, why would the combined firm now find it profitable to do so?
The Commission finds that coordinated effects are less likely due to the nationwide pricing plans offered by AT&T and Verizon. However, it spends significant time discussing the retail competition between GCI and ACS using the joint wholesale Alaska Wireless Network (AWN). The focus on the retail competition and competitively sensitive information is at odds with Commission precedent—the FCC routinely rejects inclusion of mobile virtual network operators (MVNOs) in its competition analysis for mobile wireless transaction. In fact, in this transaction, the Commission states, “Accordingly, as in previous transactions we will consider only facilities-based entities providing mobile telephony/broadband services.” 37 As a result, the Commission should have examined this transaction (and in many ways did) as a merger rather than as a joint venture with ex post retail competition. It could also have examined the transaction as a merger at the wholesale level.
Essentially, this was a merger of the number two and three providers in Alaska, with a new entrant that might become a large provider in Alaska. However, because of the tie to the continental U.S. and pricing plans, the Commission dismissed coordinated effects. While the pricing plans are tied to the rest of the country, build out and other network features are aspects of competition that the Commission routinely examines but omitted in its explanation of the competitive effects.
Efficiencies
The Commission found that “the proposed transaction would generate savings from avoiding network duplication, increase coverage of the existing network, and facilitate the deployment of LTE and the provision of mobile broadband in an area of the country that faces unique service challenges.” 38
The network efficiencies appear to result from elimination of cell sites from the network (and a reduction of new cell sites), elimination of a switch, and other network efficiencies. 39 The Commission also found that there would be expanded coverage and improved service.
The elimination of cell sites is likely to reduce fixed costs primarily, which typically are not credited as cognizable efficiencies. In the AT&T–T-Mobile transaction, such efficiencies were not considered a benefit of the transaction. The Commission may want to articulate a consistent policy on treatment of fixed cost savings due to elimination of “duplicate” facilities, since one might expect to find such savings in any merger of wireless companies with overlapping facilities. For example would they be credited only for capital constrained companies who could not deploy modern technology without the merger? Should they be credited for failing firms?
Under antitrust law and Commission precedent (including that discussed in this order), efficiencies must be “transaction-specific.” 40 The order does not discuss the transaction-specific nature of the efficiencies. First, it is possible that cell site savings could be realized with arrangements short of a merger. Many cell sites are operated by third parties, or the two companies could have leased space to each other without a merger. Second, the Commission notes, “Further evidence in the record shows that, although GCI is financially stable, executives believed that the company lacked the spectrum required to deploy LTE.” 41 It is illogical to discuss the substantial unused spectrum that could be deployed to frustrate any anticompetitive price increase (see discussion above), but then assert that the increased access to spectrum is a transaction-specific efficiency. Since CGI did not have spectrum to allocate to LTE, ACS must have had the spectrum on which LTE would be deployed. Presumably, ACS could have deployed LTE on its own. Thus, it is not clear why the ability to deploy LTE is a merger-specific efficiency since, arguably, ACS had the additional spectrum and therefore the ability to deploy LTE premerger.
The Commission approved the merger in July 2013.
Sprint–SoftBank–Clearwire
In 2012, SoftBank proposed to buy Sprint and, subsequently, for Sprint to acquire the portion of Clearwire that it did not already own. Given that SoftBank did not operate any wireless service in the U.S., it was fairly straightforward from a competitive perspective for the Commission to conclude that there was no threat of anticompetitive consequences from SoftBank acquiring Sprint. In addition, because Sprint already owned more than 50% of Clearwire, there were minimal concerns about any change in behavior from the consolidation of ownership.
The Commission found that SoftBank’s commitment to investing in Sprint and Clearwire would “facilitate certain transaction-specific public interest benefits, the acceleration of advance mobile broadband services and enhanced competition in the mobile wireless market.” 42 Because the Commission found that there were unlikely to be competitive harms, it placed a lower burden on showing the likelihood of realization of these benefits.
AT&T–ATNI
In February 2013, AT&T and ANI proposed that AT&T would acquire wireless licenses from ANI that ANI had acquired as a result of FCC and DOJ–required divestitures as part of Verizon’s acquisition of AllTel in 2010. ANI operated these licenses for three years.
Competition overlaps and concerns
The Commission found that there would be competitive harm in several local areas but not nationwide, as these areas would only account for a small portion of the country and not influence nationwide prices. However, it found concern in areas in several states (Georgia, Idaho, North Carolina, South Carolina, and Illinois). It did not find competitive concern in Ohio, the final state where the parties systems overlapped.
In its analysis of competition, the FCC examined market share and coverage (including whether the coverage was 2G, 3G, or 4G). Generally, the FCC would lay out the state of market share and coverage, all redacted. It would then either point out that neither T-Mobile or Sprint has a significant market presence 43 or say that “while Sprint has a [Redacted] percent market share, it does not have significant 2G or 3G coverage, and while T-Mobile has significant total coverage, it does not have significant market share.” 44 The FCC concluded that the Sprint and T-Mobile would have limited ability to quickly and effectively respond to any anticompetitive behavior in all of these five states.
Concluding that a company with significant total coverage but low market share would be unable to respond quickly should merit further discussion. This might make sense if the network with coverage faced a capacity constraint and an inability to expand capacity quickly, but such issues were not discussed. In addition, it seems that the Commission should address how difficult it would be for a firm with a substantial market share to expand in response to a price increase. But the decision simply asserts its conclusions.
Efficiencies
The FCC evaluates the parties’ claimed benefits from the merger: increased network build out, expanded and improved service, cost savings, and benefits to AT&T’s customers. In general, the FCC is skeptical that the claimed benefits are verifiable, merger-specific or will reduce marginal costs lowering prices to consumers. 45
The FCC then examines AT&T’s “voluntary commitments” to see if those would lead to the transaction being in the public interest. It made a commitment “to undertake an aggressive build-out schedule for upgraded networks in the Allied service areas that it is acquiring through these transactions, as well as provided a roaming commitment that helps ensure a reasonable transition.… Further, AT&T has provided certain commitments with respect to customer transition and migration.” 46
As discussed above, it is not clear how pushing for aggressive build out will necessarily mitigate the competitive harms. While one could come up with a theory for why it might help, the FCC never articulates any reason why the build out commitment would alleviate competitive concerns or be in the overall public interest when market forces could justify less capital expenditures. In addition, the commitment for build out belies the Commission’s dismissal of AT&T’s build out promises in the AT&T–T-Mobile transaction. In that transaction, AT&T claimed that absent the merger, it would only upgrade its network to LTE to 80% of the country, but would build out to 97% if the transaction were approved. 47 However, the FCC pointed out that Verizon was planning an aggressive build out and that AT&T was likely to build out aggressively even absent the merger. Subsequently, AT&T has announced much more robust build out plans. The analysis in the AT&T–GNI order does not look at the competitive forces pushing for build out absent the merger.
Second, the FCC finds that AT&T’s voluntary commitment for roaming is in the public interest. 48 However, earlier in the order, it stated, “With regard to the arguments expressing concern about the availability of roaming, we find that the Commission’s general roaming policies and rules should ensure that entities can obtain roaming agreements on reasonable terms and conditions.” 49
Ultimately, the Commission concluded, “In considering the Applications, we find that the proposed transaction results in some probable competitive harm, and that under our sliding-scale approach, we cannot conclude that the potential public interest benefits will outweigh these public interest harms. However, when we consider AT&T’s voluntary commitments in the areas of roaming, network deployment, and customer transition in conjunction with certain public interest benefits, our competitive concerns will likely be mitigated.” 50
AT&T–Leap
In August 2013, AT&T agreed to purchase Leap Wireless (“Leap”). AT&T was the second-largest wireless provider with 110 million subscribers. Leap was the fifth-largest provider of wireless service (T-Mobile had consummated its acquisition of MetroPCS by that time so that Leap moved from sixth to fifth largest) with 4.6 million subscribers. Leap’s network covered about one-third of the U.S. population and was focused on no contract, month-to-month, fixed-price service.
Competition overlaps and concerns
The Commission used both its Herfindahl-Hirschman Index (HHI) screen and spectrum screen to identify markets it would examine more closely to understand potential anticompetitive effects. The Commission found that the loss of Leap as a competitor would still leave the major four wireless providers. However, it found that “Leap has been providing a meaningful alternative for value-conscious consumers through its facilities-based prepaid service offerings.” 51 Despite some innovative offerings, the Commission did not find that Leap had recently been a “maverick” at either the local or national level.
The Commission also examined “upward pricing pressure” that might arise as a consequence of the transaction. If found “that, although AT&T and Leap are not each other’s closest substitutes over Leap’s entire facilities-based service area, Leap has provided meaningful choices for certain consumers, particularly in specific local markets. For instance, the porting rate between AT&T and Leap, in certain geographic markets, is much higher than the average porting rate between AT&T and Leap calculated over all markets where Leap has a facilities-based presence.” 52 It follows this statement with an evaluation of the specifics of the applicants’ modeling, which are redacted.
In addition, the Commission looked at the impact on the “value” segment of the wireless marketplace. AT&T had both GoPhone and Aio Wireless offering for such consumers.
The FCC found that most local geographic markets would not suffer detrimental competitive effects. 53 However, it concluded that there would likely be anticompetitive effects in south Texas, Spokane, Reno, and Lake Charles. It concluded that AT&T’s spectrum holdings would hamper the ability of its competitors “to add capacity or offer new and innovative services.” 54
Efficiencies
The Commission placed limited weight on the idea that the transaction would cause AT&T to compete more vigorously for prepaid customers. In addition, it was “largely unable to verify AT&T’s claims that the proposed transaction would lead to the enhancement of its provision of LTE services to consumers, in particular Leap customers.” 55 In addition, the Commission did not place weight on the increased use of Leap spectrum for LTE build out. However, it did credit some additional spectral efficiency, but not to the degree claimed by the applicants. For example, it found that even with spectral efficiency, AT&T would only be able to add a single 5x5 MHz LTE channel, which AT&T had described as relatively inefficient. 56
As in other mergers the Commission found that the applicants “made several voluntary commitments, which as explained below allow us to find that the proposed transaction overall would be in the public interest.” 57 Some of these are spectrum divestitures that would directly address the competition concerns in those markets, some are commitments to deploy LTE more rapidly, and some targeted rate plans for value-conscious customers.
The Commission approved the AT&T–Leap transaction on March 13, 2014.
IV. A More Robust Framework
Discussion of these recent mergers does not illustrate that the FCC made any “wrong” decisions in approving the transactions. However, the details included in the specific transactions and the lack of detail in analyzing certain portions of the transactions show that a more explicit framework would eliminate some of the uncertainty and arbitrariness of the decision process. In many of the decisions, there is a patina of rigor, but many of the justifications do not present a fully detailed analysis.
For example, in some cases the Commission determined that there would likely be some degree of competitive harm, but that the efficiencies outweighed the harms. In the AT&T–T-Mobile transaction, the Commission found that there would likely be competitive harms and used the parties’ merger simulation model (which included both competitive effects and efficiencies) to show that the efficiencies would not outweigh the harms. In the subsequent mergers, there has been no attempt to quantify the relative harms and benefits—the Commission simply concludes that either the benefits outweigh the harms or, with additional safeguards and commitments, the benefits would outweigh the harms.
Clarity About What Will Lead to Anticompetitive Effects
In order to provide more transparency and clarity for firms considering transactions, the Commission should provide more detail on the nature, scope, and impact of potential anticompetitive harms. First, it should have a theory of unilateral effects tailored to the facts of the situation. For example, in the GCI–ACS merger, it could have used the diversion ratios that it found from the Numbering Resource Utilization/Forecast (NRUF) data and estimated the unilateral effects in each region. 58 The Commission has the tools to be more specific about how competition might respond as well in a coordinated effects setting. Then, with a magnitude of likely or potential competitive harm, the Commission can proceed to investigate the potential efficiencies.
Clarity About What Is Needed in Merger Efficiency
Merger efficiencies tend to be very vague in FCC orders, in part because much of the information is competitively sensitive and therefore submitted under protective orders. As a result, we do not know exactly how the FCC estimates the magnitude of the efficiency benefits.
The industry and ultimately the public would benefit if each merger review had a detailed section explaining how the efficiencies translate into consumer benefits. Simply asserting that LTE will be built out faster or fewer cell sites will be needed is not sufficient for an understanding of the transaction-specific efficiencies.
The Commission should determine which efficiencies require a merger and which could be accomplished in ways short of a merger that do not pose the same risks to competition. Those remaining efficiencies should be quantified in terms of marginal cost reductions and fixed cost reductions. Marginal cost reductions can be marginal operating cost reductions and also reductions of marginal capacity cost. In none of the mergers did the FCC provide analysis of the magnitude of the reductions or an analysis of the share of the efficiencies that would be marginal cost reductions. 59
Other Public Interest Benefits
In its evaluation of the wireless mergers, the Commission credits firms for commitments to build out to provider greater geographic coverage and to implement advanced services more rapidly. As discussed above, such commitments may not be socially beneficial if they require more resources than the benefits they engender. As a first step, the Commission should consider that such commitments do require the use of real resources and articulate why, if such facilities or services would not be provided by the market in the absence of the merger, such facilities or services would be efficient to supply under the commitments.
Second, the Commission should develop a framework for companies to understand how it would value increased build out or the availability of more advanced technology. In some cases, increased build out would provide the second or third or fourth service to an area; in other cases, it would provide service where none has existed before. Such different build out may lead to very different net social benefits. The same is true for mandating the implementation of more advanced services. Of course, in all cases, it is important to see if competition would force such investments without conditions and therefore not be a transaction-specific benefit.
Explicit Framework for Trade-Off
Once the Commission has a better set of estimates of the likely competitive harms and the likely transaction-specific efficiencies and other public interest benefits, then it can use these to come to overall welfare estimates of the effect of a proposed transaction. The Commission should develop an explicit framework, building on the analysis it has put forth in the recent mergers, to provide clearer guidance for parties to understand the rigor it will bring to evaluating proposed mergers. With a framework, it would be possible for the FCC to add explicit and quantifiable benefits (if any) from commitments such as additional build out (how many additional users would get incremental consumer surplus from new coverage) or more rapid deployment of advanced services (what the magnitude of consumer surplus from such services would be compared to what would happen absent the merger).
With such a framework, the FCC’s merger analysis could have the rigor of the DOJ/FTC merger review process, provide more predictability, and also incorporate the “public interest” considerations that the FCC uses in addition to competition analysis. As of now, such considerations can lead to approvals that involve “commitments” (or “voluntary commitments”) without any discussion of how the overall agreement provides a net benefit to society or any indication of how similar future mergers would be analyzed.
For example, there was significant uncertainty about the likelihood of approval of a rumored Sprint–T-Mobile merger. Craig Moffett of MoffettNathanson surveyed former Commission officials on their views of the chances of approval, and the range was from 0% to 85%! 60 While there is almost always some uncertainty about the chances of approval, the wide range in part could reflect the vague standards for evaluating the competition and public interest effects in Commission reviews.
Footnotes
Authors’ Notes
Patrick DeGraba is a Senior Economist at the Federal Trade Commission. The views expressed in this paper are those of the authors and do not necessarily represent those of the Commission. He was the chief economist of the FCC’s Wireless Telecommunications Bureau during the evaluation of the AT&T–T-Mobile transaction.
Gregory Rosston is a senior fellow and deputy director of the Stanford Institute for Economic Policy Research and director of the Public Policy Program at Stanford University. He served as senior economist for transactions for the Federal Communications Commission for its evaluation of the AT&T–T-Mobile transaction.
Declaration of Conflicting Interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship, and/or publication of this article: Gregory Rosston is currently providing auction advice to T-Mobile and provided analysis for Comcast in connection with the proposed merger with Time Warner, but did not represent them or any other party in connection with any of the transactions discussed in this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.