United States' Reply Submission in Response to the Court's Minute Order of July 25, 2006

Thursday, September 21, 2006
Document Type: 
Briefs - Miscellaneous
This document is available in two formats: this web page (for browsing content) and PDF (comparable to original document formatting). To view the PDF you will need Acrobat Reader, which may be downloaded from the Adobe site. For an official signed copy, please contact the Antitrust Documents Group.


United States of America,    



SBC Communications, Inc. and
AT&T Corp.,


United States of America,    



Verizon Communications Inc. and   
MCI, Inc.,



Civil Action No.: 1:05CV02102 (EGS)

Filed: September 21, 2006

Judge Emmet G. Sullivan



Civil Action No.: 1:05CV02103 (EGS)


On October 27, 2005, the United States filed complaints alleging antitrust violations involving the provision of local private lines ("LPL") at more than 700 buildings where the mergers will harm competition by reducing the number of facilities-based competitors from two to one. The proposed Final Judgments filed the same day provide a straightforward and comprehensive remedy for the violations alleged by requiring a divestiture of facilities at each of those buildings, thereby replacing the competition lost at each of the buildings. Despite filing more than 2,100 pages of materials, amici have not seriously challenged the sufficiency of the proposed Final Judgments to remedy the violations alleged or explained why their entry would otherwise not be in the public interest.

Instead, amici once again attempt to direct the discussion to issues not before the Court. Amici urge the Court to address whether the mergers themselves are in the public interest – a question not before this Court and a question different than the legal standard under which the United States brought these cases. Amici allege that the United States should have brought a broader case and even suggest that the United States through its merger review, or this Court pursuant to the Tunney Act, should address a litany of grievances that are not even arguably caused by the mergers. In addition, amici repeatedly argue that the proposed Final Judgments should not be entered because the United States did not "prove" various elements of the Complaints in its submission to the Court. Nothing in the Tunney Act requires the United States to prove the underlying case, as if this proceeding were a trial on the merits, before the Court can approve the settlements.

The United States acknowledges that the merging parties are large corporations that provide diverse telecommunications services. This is why the United States devoted so much time and effort to investigating these transactions. But in the end, the United States concluded there was not sufficient evidence to support broader allegations of harm, including those championed by amici. Federal and state telecommunications agencies reached consistent conclusions. Amici now ask the Court to second-guess the manner in which the United States exercised its prosecutorial discretion. The Department of Justice was under no obligation to file any complaint. It filed the Complaints because its investigation revealed that the mergers would be likely to cause real and demonstrable harm to competition for LPL at the buildings identified in the proposed Final Judgments. A finding that the proposed Final Judgments are not in the public interest could leave this harm unaddressed.

In the remainder of this reply submission, the United States describes the harm alleged in the Complaints and the remedies proposed, with reference to issues and misunderstandings raised by amici. The United States then addresses the claims of the amici regarding issues beyond the scope of the violations alleged in the Complaints and rebuts the arguments of amici regarding the burden on the United States and the role of the Court under the Tunney Act.

    1. The United States' Product Market Analysis Is Straightforward, Reasonable, and Largely Undisputed

In his reply declaration, Dr. Majure explains the economic theory behind market definition and describes the generally accepted methodology set forth in the Horizontal Merger Guidelines ("Merger Guidelines").(1) The Merger Guidelines define the relevant product market as the smallest possible group of products over which a hypothetical monopolist likely would profitably impose a small, but significant, non-transitory price increase.(2)

In defining a product market here, the United States began its analysis with the LPL products of the merging parties.(3) As Dr. Majure explains, this is an appropriate place to start because LPLs are a fundamental input for any telecommunications product serving a large customer with a significant volume of traffic, and are routinely bought and sold commercially. LPL has a price and it is possible to consider what would happen if a hypothetical monopolist raised that price. The next closest substitute would be an undedicated or switched circuit. The only way to avoid directly or indirectly purchasing LPL would be to sacrifice the functionality of a dedicated connection. A customer with the volume and other requirements that justify paying for a private line would not be likely to make that sacrifice. The United States, therefore, concluded that there are no practical alternatives to LPL for most customers and that it constitutes a relevant product market.(4)

In mergers in which close substitutes exist for a particular product, market definition can be a contentious issue. In this instance, however, there are no significant close substitutes for LPL. Moreover, many of the amici have participated in regulatory proceedings related to special access, the term for LPLs provided by Regional Bell Operating Companies ("RBOCs ") pursuant to certain regulations, that consider these services as a product market for purposes of competitive analysis.(5) It is not surprising, therefore, that amici have not challenged the United States' alleged LPL market other than to suggest that the United States should have alleged additional product markets as well.

    1. Geographic Markets Consisting of Individual Buildings Are Consistent with Well-Established Antitrust Principles

Several amici criticize the United States for alleging that individual building locations can constitute an appropriate relevant geographic markets for LPL.(6) This market definition, however, is consistent with well-established antitrust principles as well as the prior practice of the Department of Justice.(7) Moreover, the Federal Communications Commission ("FCC") has repeatedly stated that a building-by-building approach is the most accurate way to evaluate wholesale competition for LPL.(8)

As with product markets, the Merger Guidelines define geographic markets by evaluating demand substitution responses, i.e., how purchasers would respond to a price increase. A relevant geographic market is the smallest geographic region in which a hypothetical monopolist could profitably impose a price increase.(9) Stated differently, it includes the set of sellers to which a buyer could practicably turn for the product at existing or slightly higher prices.(10) In making this evaluation, it is important to consider the "commercial realities" faced by consumers.(11)

From the perspective of a buyer of LPL, competitive options are limited to suppliers able to provide LPL service to a given building. Theoretically, a buyer might obtain LPL service from a competitor with facilities at a nearby location by moving its business to that building and thereby escape the hypothetical monopolist's price increase. The cost and disruption of relocating, however, will almost always outweigh the benefits of avoiding a slightly higher price.(12) A hypothetical monopolist owning all of the LPLs to a building would, therefore, be able to raise prices profitably.

NASUCA's consultant, Dr. Selwyn,(13) applied similar reasoning to reach the same geographic market definition in a recent FCC proceeding. In a declaration submitted on behalf of AT&T, Dr. Selwyn argued that competition for local telecommunications services, including special access facilities, should be evaluated on a building-by-building, rather than metro-wide, basis.(14) Dr. Selwyn recognized that a "physical facilities-based presence at a particular customer premises affords the CLEC access only to that specific premises and to no others."(15) He further argued that the "commercial reality" is that customers will not relocate to obtain a competing telephone service.(16) Citing to both the Merger Guidelines and Brown Shoe, Dr. Selwyn concluded that an appropriate demand-based definition would "necessarily define the 'relevant geographic market' as consisting, in each case, of one individual customer premises."(17)

COMPTEL argues that a building-specific market definition is not reasonable because the evidence does not support the "notion that customers, either retail or wholesale, or competitors, view a building as a market."(18) Yet COMPTEL, which purports to represent both customers and competitors in the market for LPL, argued that the FCC should use a building-by-building approach to evaluate competitive conditions for LPL.(19) Various CLECs touted the same building-specific approach in other FCC proceedings addressing LPL.(20) Even the submission by amicus Sprint Nextel ("Sprint") in this proceeding illustrates that customers evaluate purchasing options for LPL on a building-by-building basis:

Sprint Nextel maintains a database that shows whether a particular building is 'on-net' for one or more AAVs [alternative access vendors]. Sprint Nextel's enterprise services sales force accesses that database to determine available options (if any) for obtaining non-ILEC special access circuits to customer premises.(21)

Consistent with this commercial reality, the United States found that each commercial building could constitute a separate geographic market.

Although many amici have advocated this very approach in the past, they are now united in opposition to the use of individual buildings as an appropriate market. However, they cannot agree whether this is too broad or too narrow. The market definitions proposed by amici range from the very narrow (individual floors of each building)(22) to the very broad (the entire United States).(23)

As discussed below, the United States concluded, and therefore alleged, that the anticompetitive effects of the transactions will be limited to buildings where (a) the mergers will reduce the number of competitors from two to one and (b) where entry is not likely.(24) Treating individual buildings as separate geographic markets follows well-established antitrust principles and accurately captures the harm to competition the United States alleged in these cases. Because there are also some facts that suggest broader markets, the United States' Complaints acknowledge that the geographic market may be as broad as the metropolitan area. Nevertheless, if the market is as broad as the metropolitan area, then the market is highly differentiated, with different carriers able to reach very different sets of locations and buildings within the area. Regardless of whether the appropriate geographic market here is as narrow as the individual building or as broad as the metropolitan area, the competitive harm likely to result from the proposed merger is limited to a set of 2-to-1 buildings, as alleged in the Complaints. Because the Department had entered into settlements that remedied all the harm it had identified from the mergers, it was not necessary to determine whether a market definition that was broader than individual buildings was more appropriate.(25)

    1. The Competitive Harm Alleged in the Complaints Is Consistent with the Record and Well-Established Antitrust Principles

The United States alleges a straightforward theory of competitive effects. After the merger, SBC(26) or Verizon would be the only remaining supplier of LPL to certain buildings where they previously competed. The mergers therefore are likely to result in higher prices or lower quality (e.g., less responsiveness to service outages or requests to provide new circuits) for LPL and services sold across LPL to customers in these buildings.(27)

      1. HHI Calculations Would Add Little to the Competitive Analysis of the Harm Alleged by the United States

Several amici fault the United States' competitive effects analysis for not calculating HHIs or market concentration measures.(28) Theoretically, it would be possible to calculate building-specific market shares for LPL using either revenues or capacity. In the markets where harm is alleged in the Complaints, however, the first approach is inherently unreliable, while the second adds little to a competitive effects analysis. Dr. Majure's reply declaration explains why using revenues may lead one to draw incorrect conclusions in these markets.(29) Moreover, using HHIs and market concentration based on capacity would add little to competitive effects analysis here. In each of the buildings where the United States has alleged competitive harm, the number of competitors would go from two to one. To say that HHI figures would increase from 5,000 to 10,000 would add nothing useful to the United States' explanation of harm. The Complaints make it clear that customers in these buildings would be harmed by losing the benefits of the competition provided by either AT&T or MCI.(30)

In their responses, amici also have overstated the role of HHIs and concentration figures in the United States' enforcement decisions. Several amici have implied that HHI figures compel a finding of broader competitive harm from these mergers.(31) To the contrary, as a recent joint agency commentary on the Merger Guidelines stated clearly, though low market shares and concentration are a sufficient basis for not challenging a merger, large market shares and high concentration by themselves are an insufficient basis for challenging a merger.(32) It is not surprising, therefore, that data on recent merger challenges shows that the United States often does not challenge mergers involving market shares and concentration above the thresholds described in the Guidelines.(33)

      1. The United States Reasonably Concluded that Entry Was Not Likely to Prevent Harm in Certain Buildings

The Merger Guidelines explains that mergers that would otherwise appear problematic may not harm consumers if new firms likely would enter in response to a price increase by the hypothetical monopolist. The Guidelines explain that in order to counter the impacts of the merger, such entry must be timely, likely, and sufficient.(34) Moreover, the U.S. Court of Appeals for the D.C. Circuit has appropriately recognized that where barriers to entry are sufficiently low, even the threat of outside entry can deter anticompetitive behavior.(35)

Therefore, the United States investigated whether entry would be likely to deter competitive effects in any of the buildings where it otherwise anticipated harm to competition. It found that entry does occur – CLECs have indisputably constructed laterals to thousands of buildings in order to compete for LPL sales. On the other hand, the investigation revealed that there are thousands of buildings where competitive entry has not occurred. To determine why entry happens at some buildings but not others, the United States investigated the criteria CLECs consider in deciding whether to enter a building.(36) Though there is some variation among CLECs, the United States found that the most significant – the most often determinative – factors governing whether a CLEC will build into a particular building are the revenue opportunity in the building (as reflected by the capacity demand) and the cost of building a lateral to the building (which typically depends heavily on the distance of the building from the carrier's network). The United States also found that in limited cases, building-specific barriers may also impact the decision. Based on its analysis of all these factors, the United States concluded that entry would be likely in many buildings but that the conditions for entry would be unlikely to be met in hundreds of other buildings where competition would be harmed by the mergers.

The United States crafted entry screens that it could apply to the buildings where, absent entry, the merger would likely harm competition. These entry screens take into account the factors that the United States found to be the most important – the demand for LPL services at the building and the distance of the building from competitive fiber.(37) Notably, no amici challenge the specific distance/demand elements of the United States' entry screens.(38) Indeed, some amici materials – such as those submitted by Sprint – tend to support it.(39)

Some amici, however, contend that the United States' entry screens are faulty because they do not explicitly evaluate all potential barriers to entry mentioned in the United States' Complaints.(40) But the barriers identified by amici either are already implicitly incorporated in the demand-distance criteria used by the United States or are relatively minor ones that would have an impact only in atypical cases.(41) Recognizing that it would be impracticable to separately analyze any building-specific access costs or potential physical barriers for each of the more than one thousand two-to-one buildings, that the plaintiff ultimately has the burden of proving competitive harm in every market it alleges, and that the United States' entry methodology is based on the most important factors governing CLEC entry decisions, the entry screens devised by the United States constitute a reasonable and practical approach that predicts entry as closely as feasible.(42)

Several amici claim that the United States' methodology is flawed because CLECs would not enter a building in the absence of "committed revenue" in the form of customer contracts.(43) The United States, however, has not suggested that CLECs randomly undertake the construction of laterals regardless of the potential to win business from customers in the building. Rather, the screens utilized by the United States evaluate the entry response of CLECs when harm would otherwise be likely – i.e. when customer contracts are up for bid and only one carrier owns all the lateral connections to a building. The question is whether one or more CLECs would be likely to bid for that business, even if winning would require the CLEC to build fiber into the building in question. If the demand and distance criteria set forth in the United States screens are satisfied, the answer is likely to be yes. Entry (where the CLEC wins the bid), or the threat of entry (where the incumbent wins the bid), is likely to prevent any anticompetitive effects from the mergers in those buildings.(44)

Several other amici have argued, in essence, that entry is too difficult to predict with confidence and therefore should be ignored or discounted. NYAG's consultant, for example, suggests that judgments regarding the likelihood of entry are largely speculative and therefore should not be the basis for analysis or predictions.(45) NASUCA's consultant takes a similar approach, suggesting that because of the uncertainties involved in predicting entry, "the Department could have erred on the side of overinclusiveness in the selection of divestiture assets."(46) The fact that entry analysis is difficult or forward-looking does not relieve the United States of its obligation to undertake it. Merger analysis is predictive in nature. If the United States is not sufficiently confident that a harm can be predicted and proven, it does not allege one.(47) Remedies obtained in a consent decree are the product of adversarial negotiations, in which the United States must often convince the merging parties that it could successfully sue to obtain the same relief. The fact that it would be difficult to prove that entry is unlikely (particularly given the large amount of deployed fiber) is one of the factors that was considered in deciding whether to enter into these settlements.

    1. The Proposed Final Judgments Offer a Straightforward and Comprehensive Remedy for the Harm Alleged in the Complaints

As reflected in the Antitrust Division Policy Guide to Merger Remedies, any settlement must "fit[] the violation and flow[] from the theory of competitive harm."(48) Given the theory of competitive harm the United States alleged, the proposed Final Judgments provide a comprehensive remedy. Each requires the divestiture of lateral connections into the buildings where the merger will reduce the number of competitors from two to one and where entry is not likely to prevent harm to competition. The buyer of the divested assets would provide customers for LPL in those buildings with an alternative to SBC or Verizon. All customers – the tenants in the building as well as the carriers that need to buy a connection in order to sell their services to tenants – will have a choice of two facilities-based providers, just as they did before the mergers.(49) The divestitures will thereby remedy the harm in each of these markets for LPL and protect the enterprise customers in those buildings that purchase telecommunications services sold over those lines as well.(50)

      1. The United States Will Ensure that the Divested Assets Are Sold Only to Effective Competitors

ACTel suggests that the buyer of the divestiture assets may be unable or unwilling to serve wholesale customers for LPL.(51) Under the proposed Final Judgments, however, the United States has approval rights over each buyer,(52) and is obligated to review, among other factors, each potential purchaser's "ability to be a viable competitor" for wholesale LPL.(53) To date, all of the buyers that the United States has conditionally approved for the divestiture assets are active wholesale providers of LPL.(54) Thus, there is no reasonable basis to object to the decrees on the grounds that the buyers will not provide service to wholesale customers.

ACTel also argues that at least one of the proposed divestiture buyers –[REDACTED TEXT] – is small compared to AT&T and is "losing money."(55) Financial stability is one factor that the United States always considers when evaluating the fitness of a proposed purchaser of divestiture assets.(56) But there is little or no reason to doubt [REDACTED] financial viability or its ability to adequately – indeed, aggressively – provide LPL to the buildings in question.(57) In its most recent earnings announcement, [REDACTED TEXT] (58) Thus, amici cannot persuasively argue that [REDACTED TEXT]– or any other buyer the United States would approve – would be able to adequately replace the competition in the buildings addressed by the decrees.

NYAG's consultant, Dr. Economides, also complains that the divestiture of lines to several hundred buildings would not allow a competitor to "replace AT&T or MCI."(59) However, the purpose of the remedy is not to "replace" AT&T or MCI in its entirety, but rather to replace competition for LPL services to those specific buildings where the merger is likely to lead to consumer harm.(60) The proposed remedy accomplishes that by allowing another CLEC to use the divested assets to provide the competition that AT&T or MCI would have provided in each of these buildings.

      1. The Divested Assets Will Replace the Competition Lost from the Merger

Several amici argue that the divested assets are insufficient to remedy the harm to competition alleged in the Complaints. NYAG's consultant, Dr. Economides, complains that the fiber being divested is unused or "dark" fiber and, therefore, it "could be useful only if existing customers in a particular building needed additional bandwidth that Verizon or SBC could not supply . . . or if the customers chose to switch providers."(61) But this does not undermine the effectiveness of the remedy. The whole point of the remedy is to ensure that when customers do consider switching providers or buying additional capacity, a competitive alternative to the merged firm is available; a divestiture of dark fiber accomplishes that.(62)

Dr. Economides also complains that, because the divestiture is in the form of an IRU rather than full ownership, the divestiture buyer will have to rely on the merged firm to maintain the fiber, and the merged firm may not do an adequate job of maintenance.(63) If the terms in the IRU agreements were not sufficient to ensure that the divested fiber would be adequately maintained it could, theoretically, impair the effectiveness of the remedy. However, the United States has approval authority over the terms of the IRU agreements,(64) and will exercise that authority to ensure that they are adequate.(65) For instance, the divestiture agreements with all three buyers that have been conditionally approved by the United States, [REDACTED TEXT] (66) [REDACTED TEXT](67) Accordingly, the IRU form of the divestiture will not impair the effectiveness of the proposed remedy.

Finally, NASUCA's declarant, Dr. Selwyn, argues that the price paid by the purchasers of the divested assets indicates that they cannot be used effectively to provide competition.(68) Contrary to Dr. Selwyn's suggestion, there is no reason to expect that these assets would fetch prices equivalent to the prices paid when one firm acquires another firm outside of a divestiture context.(69) Prices here should be lower because the buyers of the divested assets are not acquiring a current revenue stream. Instead they are acquiring assets that provide the ability to compete for future business opportunities as they arise.(70) Moreover, it is not uncommon for acquirers to pay lower prices for assets divested under consent decrees. Buyers recognize that the merged entity has an obligation to sell the assets under the timetable set forth in a publicly available consent decree. Under those circumstances, economists generally expect the buyer to get the better end of the negotiations. For all of these reasons, the prices paid for the divested assets do not support the inference made by amici.(71) While the price would undoubtedly be higher if customers were also being divested, that does not mean that the purchasers of the divested assets will be weakened and thus unable to remedy the harm alleged in the Complaints. As soon as the acquirer has the ability to serve these buildings without having to make an extensive investment in infrastructure, it will be able to compete aggressively for each new business opportunity.

    1. Arguments Made by Amici About Harms Beyond the Scope of the Complaints Are Irrelevant to the Court's Public Interest Determination

Several amici harbor fundamental misconceptions about the authority of the Department of Justice and the task before this Court in a Tunney Act proceeding. The Department of Justice is not a regulatory agency -- it is a law enforcement agency empowered to bring suit in federal court to challenge specific violations of the antitrust statutes.(72) Unlike regulatory agencies such as the FCC and state utility commissions, the Department of Justice is not authorized to challenge or block a merger on "public interest" grounds that are not antitrust violations, nor can it seize upon a merger as an opportunity to improve the state of competition in an industry or a market beyond remedying the effects of an unlawful merger. Congress delegated regulatory responsibility for telecommunications mergers to the FCC, which weighs whether merger applicants have shown that, on balance, a transaction will serve the "public interest, convenience, and necessity."(73) Many states authorize state regulators, such as the New York Public Service Commission, to conduct similar "public interest" reviews.(74) Unsatisfied with the outcome of those proceedings,(75) as well as the manner in which the Department of Justice exercised its prosecutorial discretion, various amici have seized upon these Tunney Act proceedings as a forum to revisit the merits of the underlying transactions and to achieve goals they failed to achieve before the FCC and other regulatory agencies.(76)

This expansive view of the scope of the current proceedings is patently inconsistent with the plain language of the Tunney Act, which directs the Court to evaluate the impact of the proposed consent decree, rather than of the underlying transaction, on the public interest. Specifically, the statute directs the Court to consider whether "entry of such judgment is in the public interest."(77) The Court's role under the Tunney Act is limited to reviewing the remedy in relationship to the violations that the United States has alleged in its Complaints.(78) It should not base its public interest determination regarding the proposed Final Judgments on antitrust concerns that would not have been part of the government's case had these cases gone to trial on the existing Complaints.

Amici also misconstrue the burden on the United States in a Tunney Act proceeding. Amici, principally ACTel and COMPTEL, attack the "evidentiary" weight of the materials(79) that the United States provided to the Court and argue that the United States failed to "prove" various elements of the allegations in the Complaints. As movant, the United States has the burden of persuading the Court that the entry of the consent decrees is in the "public interest." Nothing in the Tunney Act or in Tunney Act jurisprudence, however, suggests that the United States is required to prove by a preponderance of the evidence (or any other standard) each of the elements of cases that it has settled. Such a requirement would substantially undercut the reasons for entering settlements – saving time and expense and avoiding the risk of losing at trial.

Amici continue to raise a host of issues that go beyond the scope of the United States' Complaints despite this Court's statement "for the record" that it would not consider such issues.(80) These issues include alleged harms in broader LPL markets as well as product markets (e.g., mass-market telephony) that are separate and distinct from LPL. The United States explained below why it did not allege a broader harm than described in its Complaints, and it would be impracticable and improper to require the United States to disprove the cases that it did not bring. Our system of separation of powers leaves the balancing of competing interests affecting the scope of the United States' case "in the first instance, to the discretion of the Attorney General."(81)

    1. The Mergers Are Not Likely to Harm Competition for LPL Beyond the Two-to-One Buildings Alleged in the Complaints

All of the amici have suggested that if the mergers cause harm to competition in two-to-one buildings, it necessarily follows that competition will also be harmed in buildings where the number of competitors is reduced from four to three or three to two.(82) Sprint goes so far as to assert that such harm is a "fact," of which the Court should take judicial notice.(83) ACTel likewise contends that "[t]he very same analysis [that] 'predicts' harm in 2-to-1 buildings also predicts harm in 4-to-3 and 3-to-2 buildings."(84) The loss of a competitor, however, does not always translate into a loss of competition. After careful investigation, the United States did not find sufficient evidence to support allegations by amici that AT&T and MCI are uniquely capable competitors for LPL. Moreover, the application of well-settled antitrust principles to LPL markets does not suggest that the mergers are likely to significantly harm competition beyond the two-to-one buildings where harm is alleged in the Complaints.

      1. Amici Have Distorted the Competitive Significance of AT&T and MCI

Several amici have alleged that the elimination of the acquired firm (AT&T in SBC's territory or MCI in Verizon's territory) will lessen competition because this company has unique characteristics that allow it to compete more aggressively against the RBOC than other CLECs. Amici point primarily to the size and extent of AT&T's or MCI's network, particularly the number of buildings on-net, as advantages that distinguish the acquired firm from other CLECs, and they rely also on the existence of network effects(85) in telecommunications networks.(86) The amici's arguments are misplaced because the acquired firms' networks are not the most extensive in most geographic areas of concerns or in any event significantly larger than the networks of other CLECs. In addition, whatever network effects exist, AT&T and MCI do not benefit from such effects substantially more than other CLECs.

The evidence provided by the United States with its Submission of August 7, 2006, demonstrates that amici have vastly overstated AT&T's and MCI's networks and competitive significance as providers of LPL. Indeed, AT&T and MCI had only about [REDACTED TEXT] total on-net buildings in SBC's and Verizon's territories respectively,(87) and their sales of LPL were relatively small, particularly in relation to the RBOCs. A report prepared by NASUCA's consultant, which was submitted in this proceeding by amicus Sprint, indicates that on a national basis, AT&T and MCI each had approximately the same number of on-net buildings as TWT does today.(88) Moreover, in most of the metropolitan areas identified in the United States' Complaints, the acquired CLEC is not the largest CLEC in terms of number of buildings on-net.(89) In [REDACTED TEXT] of the nineteen divestiture cities, a carrier other than the acquired firm had the most on-net buildings of any CLEC.(90) Even in cities where the merging party was the largest CLEC, one or more other CLECs is typically not far behind.(91)

The arguments made by amici as to network effects also have no merit. Although enterprise customers want to buy a telecommunications service that allows them to reach particular locations, no carrier, not even the RBOCs, has facilities that connect to every building nationwide or worldwide. Providers, therefore, rely on their ability to interconnect their networks with other carriers' networks in order to be able to meet the particular needs of their customers. For example, a carrier that bids to serve a sophisticated enterprise customer by constructing an advanced data network connecting all of the customer's offices usually has to obtain some building connections from other carriers.(92) The carrier would purchase LPLs to connect those buildings to its network so that the customer gets what appears to be one network that reaches all its offices. The fact that the seller of the LPL has connections to many other buildings is irrelevant to the carrier purchasing it. The only way a seller might have a unique benefit to the buyer is if it has many of the customer's buildings on its network. The benefit to the buyer is that it avoids the need to have to negotiate with more sellers, but this benefit may be quite small. All CLECs, including AT&T and MCI, own facilities that connect to only a small percentage of the buildings in an area, and therefore, it would be very unlikely for an enterprise customer to be interested in multiple buildings on the same CLEC's network. Thus, AT&T and MCI generally are unlikely to have an advantage over other CLECs in selling LPLs due to such network effects .

Perhaps in recognition of the fact that the acquired firm typically does not have a uniquely large facilities-based network in the metropolitan areas in question, several amici have argued that the impact of the mergers extends to buildings where neither AT&T nor MCI owns a connection. They contend that the loss of AT&T and MCI in SBC's and Verizon's regions, respectively, will have anticompetitive consequences due to the loss of those firms as re-sellers of LPL to other carriers.(93) According to these amici, AT&T and/or MCI were particularly important sellers of "Type II" circuits – circuits that include at least a portion that is purchased wholesale from the RBOC.(94) The United States considered this issue but did not find evidence to support such allegations.(95) To the contrary, AT&T's and MCI's sales of "Type II" LPL pre-merger were relatively small.(96) Indeed, as a document attached to NASUCA's reply shows, AT&T had decided before the merger that it could not profitably offer Type II LPL except where it is able to provision the service using primarily its own network facilities.(97) Neither AT&T nor MCI have uniquely extensive transport networks that they could combine with RBOC access circuits to become especially effective sellers of Type II circuits.(98) Nor is there any evidence to suggest that AT&T or MCI obtained discount terms from the RBOCs for LPL that were significantly, materially better than those available to other CLECs.(99) For the reasons above, the United States did not conclude that the elimination of AT&T or MCI as a reseller of LPL was likely to have a significant anticompetitive effect on LPL.(100)

Finally, contrary to ACTel's suggestions, the evidence does not show that the loss of AT&T and MCI would have a disproportionate effect because of their purported roles as "low price leaders."(101) ACTel's primary evidence in support of this allegation consists of [REDACTED TEXT] (102) [REDACTED TEXT] (103) As Dr. Majure notes, other evidence the United States analyzed suggests that, [REDACTED TEXT] [REDACTED TEXT] (104) [REDACTED TEXT] (105)

      1. Application of Well-Established Antitrust Theories Does Not Suggest that the Mergers will Cause Broader Harm in LPL Markets

As Dr. Majure explains, the two basic theories of competitive harm discussed in the Merger Guidelines are coordinated effectsand unilateral effects.(106) Whether a merger is likely to cause either unilateral or coordinated effects depends on the nature of competition in a particular market. Notwithstanding conclusory statements by amici to the contrary, neither theory provides a sound basis for alleging harm beyond the two-to-one buildings identified in the proposed Final Judgments.(107)

The United States investigated whether the mergers were likely to cause broader harm under a unilateral effects theory. As discussed above, the United States did not find that AT&T or MCI is uniquely situated to compete for LPL.(108) The evidence shows that LPL is close to a pure commodity product in which buyers perceive no substantial difference between the various LPL options and make purchasing decisions based primarily on price.(109) In commodity markets, antitrust theory does not suggest unilateral effects unless the remaining competitor or competitors in the market are capacity constrained such that they would be unable to supply customers seeking to escape a price increase by the merged firm.(110) That is not a concern here because the incremental cost of expanding capacity is relative minor once a carrier has established a connection into a building.(111) As long as at least one other CLEC has a connection to a building there is no reason to believe that that CLEC can not adequately replace AT&T or MCI.

The United States also investigated the potential for the merger to have coordinated effects in broader LPL markets. Successful collusion depends on the ability of the parties to reach and police an agreement regarding pricing or output.(112) Dr. Majure's reply declaration explains that LPL markets are not conducive to collusion for several reasons, including the fact that carriers that invested to bring a building "on-net" have the strong incentive to compete aggressively for every customer in the building in order to recover the large fixed costs associated with building facilities.(113) A reasoned application of antitrust principles, therefore, suggests that neither unilateral effects nor coordinated effects are likely where the merged firm faces competition from another CLEC after the mergers.

    1. The Mergers Are Not Likely to Harm Retail Competition for Telecommunications Services Provided over LPL Beyond the Two-to-One Buildings Alleged in the Complaints

A number of amici suggest that the mergers are likely to cause competitive harm to retail enterprise customers beyond the harm in the 2-to-1 buildings that was alleged in the Complaints. The United States carefully investigated whether the mergers in question would cause competitive harm to retail enterprise customers for a broad array of telecommunications services and concluded that it would not. Based on its review of the documents and data submitted by the parties as well as approximately 200 customer interviews,(114) the United States concluded that there are numerous other firms competing to provide telecommunications services to large businesses.

In fact, the United States' investigation revealed that the retail business offerings of the merging firms are largely complementary rather than overlapping.(115) Whereas AT&T and MCI tend to be strong competitors for long distance voice service and advanced data services sold to customers with nationwide or international reach, SBC and Verizon are focused on providing business services to retail customers with primarily "in-region" needs. Both SBC and Verizon have explained that it was partly their inability to win large enterprise customers that provided the impetus for these mergers.(116) The United States did not find harm in retail enterprise markets except for services provided over LPL to customers in certain two-to-one buildings.(117)

    1. The Mergers are Not Likely to Harm Retail Mass-Market Competition

Some amici reprise their demands that the Court base its public interest determination on unsubstantiated claims of harm in consumer markets that are not even arguably within the scope of the Complaints. NASUCA, NJRC, and NYAG argue that the Court should refuse to enter the proposed Final Judgments because they fail to remedy purported harm to mass-market customers who purchase products like "plain old telephone" or residential "long distance" services.(118) Their concerns, however, have little to do with the mergers, much less with the harm alleged in the Complaints. Mass-market products utilize a different technology (switched access) and are sold to a different set of customers (residential and small business). The primary grievance stated by amici appears to be with the FCC, and its recent decisions to relieve RBOCs of regulatory obligations to make switched-access facilities available to competitive CLECs under a cost-based tariff.(119) In fact, at least one amicus specifically asks the Court to overturn the FCC's Order with respect to the merging parties.(120)

The United States fully investigated the mergers' potential to harm competition for mass-market services and found insufficient evidence to allege a violation of Section 7 of the Clayton Act. The United States' decision was based on a number of factors discussed in Dr. Majure's declaration, including regulatory changes that diminished the ability of AT&T and MCI to compete with the RBOCs to serve mass-market customers.(121) Several regulatory agencies that reviewed the mergers found that AT&T and MCI had already begun to exit mass markets before the mergers.(122) Under those circumstances, the United States' conclusion that the mergers were unlikely to harm competition in mass markets was clearly reasonable. In any event, the law is well settled that such claims, which are far beyond any conceivable interpretation of the Complaints, cannot serve as the basis for the Court to find that the proposed Final Judgments are not in the public interest.(123)


Accordingly, the United States respectfully requests that the Court grant the United States' motion for entry of the proposed Final Judgments.

    Respectfully submitted,

Laury E. Bobbish
Assistant Chief

Claude F. Scott, Jr. (D.C. Bar No. 414906)
John M. Snyder (D.C. Bar No. 456921)
Jared A. Hughes

Trial Attorneys

Telecommunications and Media Section
Antitrust Division
U.S. Department of Justice
1401 H Street, N.W., Suite 8000
Washington, D.C. 20530
(202) 514-5621
Attorneys for the United States

Dated: September 19, 2006

Attachment 1

Attachment 2

Attachment 3

Attachment 1



AT&T Corp.

In connection with the proposed transaction, SBC intends to file a registration statement, including a proxy statement of AT&T Corp., and other materials with the Securities and Exchange Commission (the "SEC"). Investors are urged to read the registration statement and other materials when they are available because they contain important information. Investors will be able to obtain free copies of the registration statement and proxy statement, when they become available, as well as other filings containing information about SBC and AT&T Corp., without charge, at the SEC's Internet site (www.sec.gov). These documents may also be obtained for free from SBC's Investor Relations web site (www.sbc.com/investor_relations) or by directing a request to SBC Communications Inc., Stockholder Services, 175 E. Houston, San Antonio, Texas 78205. Free copies of AT&T Corp.'s filings may be accessed and downloaded for free at the AT&T Relations Web Site (www.att.com/ir/sec) or by directing a request to AT&T Corp., Investor Relations, One AT&T Way, Bedminster, New Jersey 07921.

SBC, AT&T Corp. and their respective directors and executive officers and other members of management and employees may be deemed to be participants in the solicitation of proxies from AT&T shareholders in respect of the proposed transaction. Information regarding SBC's directors and executive officers is available in SBC's proxy statement for its 2004 annual meeting of stockholders, dated March 11, 2004, and information regarding AT&T Corp.'s directors and executive officers is available in AT&T Corp.'s proxy statement for its 2004 annual meeting of shareholders, dated March 25, 2004. Additional information regarding the interests of such potential participants will be included in the registration and proxy statement and the other relevant documents filed with the SEC when they become available.

Certain matters discussed in this statement, including the appendices attached, are forward-looking statements that involve risks and uncertainties. Forward-looking statements include, without limitation, the information concerning possible or assumed future revenues and results of operations of SBC and AT&T, projected benefits of the proposed SBC/AT&T merger and possible or assumed developments in the telecommunications industry. Readers are cautioned that the following important factors, in addition to those discussed in this statement and elsewhere in the proxy statement/prospectus to be filed by SBC with the Securities and Exchange Commission, and in the documents incorporated by reference in such proxy statement/prospectus, could affect the future results of SBC and AT&T or the prospects for the merger: (1) the ability to obtain governmental approvals of the merger on the proposed terms and schedule; (2) the failure of AT&T shareholders to approve the merger; (3) the risks that the businesses of SBC and AT&T will not be integrated successfully; (4) the risks that the cost savings and any other synergies from the merger may not be fully realized or may take longer to realize than expected; (5) disruption from the merger making it more difficult to maintain relationships with customers, employees or suppliers; (6) competition and its effect on pricing, costs, spending, third-party relationships and revenues; (7) the risk that Cingular Wireless LLC could fail to achieve, in the amount and within the timeframe expected, the synergies and other benefits expected from its acquisition of AT&T Wireless; (8) final outcomes of various state and federal regulatory proceedings and changes in existing state, federal or foreign laws and regulations and/or enactment of additional regulatory laws and regulations; (9) risks inherent in international operations, including exposure to fluctuations in foreign currency exchange rates and political risk; (10) the impact of new technologies; (11) changes in general economic and market conditions; and (12) changes in the regulatory environment in which SBC and AT&T operate.

The cites to webpages in this document are for information only and are not intended to be active links or to incorporate herein any information on the websites, except the specific information for which the webpages have been cited.


AT&T Corp.

  1. My name is Anthony Fea. My business address is 200 Laurel Ave Middletown, New Jersey. I am a Director responsible for the Program and Project Management of AT&T's Local Network Services ("LNS") organization, the group within AT&T Corp. that provides local service to AT&T Business customers. I am currently responsible for LNS' national integrated Program and Project Management activities. Integrated Program and Project Management planning activities includes Program and Project Management activities for the Switch, Transport, Node, Digital Cross-Connect Systems and Outside Plant technologies used in AT&T's local networks, as well as interexchange carrier ("IXC") collocations and network optimization. As part of my job, I am also responsible for supporting the current and future years' capital budgets, along with current year capital management responsibilities. I am a graduate of Stevens Institute of Technology, with a B.S. in Electrical Engineering. Since obtaining my degree, I have worked at a number of telecommunications firms including Bell Atlantic (now Verizon), Telcordia Technologies (BellCore), and most recently TCG and AT&T.
  2. My name is Anthony J. Giovannucci. My business address is 207-209 F Street, South Boston, Massachusetts. I am a Director for AT&T's Engineering organization, specifically overseeing AT&T's Media Engineering organization which is responsible for national planning and deploying AT&T's transmission media (fiber and microwave), for both Local and Long Distance applications. In my current position I am responsible for a number of key areas of Outside Plant activity, including the development of an Outside Plant ("OSP") Plan of Record for capital deployment, negotiation and completion of agreements controlling rights-of-way, building rights-of-entry, franchises and joint facility builds as well as the evaluation of distressed assets for their potential acquisition and incorporation into AT&T's network footprint. Prior to my employment by AT&T, I performed OSP Engineering on a contract basis at various regional Bell companies (New England Telephone and BellSouth) between 1987 and 1993. From 1993 to 1998, I worked at TCG which was acquired by AT&T in 1998. Along with Mr. Fea, I am the principal sponsor of the testimony describing the engineering, operation and location of AT&T's local networks.
  3. My name is Bob Handal. My business address is One AT&T Way Bedminster, NJ 07921. I am a Director responsible for Alternate Supply within the Local Services and Access Management ("LSAM") organization that is responsible for access procurement. Specifically, the Alternate Supply team manages relationships with suppliers other than incumbent local exchange carriers ("LECs") to procure access services. As part of my job, I am responsible for developing relationships with suppliers that can offer alternatives to the special access services offered by incumbent LECs. I am responsible for the execution of supplier agreements, the associated budgets, and unit cost reduction targets. I have worked at AT&T in a variety of positions since I graduated from the University of Vermont in 1989. I have been in my current assignment since January of 2003. I am the principal sponsor of the portions of the testimony pertaining to AT&T's purchase of dedicated access from competitive carriers.
  4. My name is Michael E. Lesher. My business address is One AT&T Way, Room 5C212F, Bedminster, NJ 07921. I am employed by AT&T Corp. ("AT&T") as the Director of Access Product Management within AT&T's Business Services organization. My current duties include the development and lifecycle management of AT&T's point-to-point and ring access services, including responsibility for product costing and pricing, feature development, service implementation and process improvement of both local and private line services. Prior to this, I have held a number of positions at AT&T with responsibility for AT&T's local network and services. I hold a B.S. degree in Accounting from Virginia Polytechnic Institute and State University, and an M.B.A. in Finance and Computer Science from Southern Methodist University. I am the principal sponsor of the testimony pertaining to AT&T's supply of local private line services.
  5. My name is C. Michael Pfau. My business address is One AT&T Way, Room 3A158, Bedminster, New Jersey 07921. I have a Bachelors of Science degree in Mechanical Engineering and a Master of Business Administration. I have a Professional Engineering license from the state of Pennsylvania. I am employed by AT&T Corp. ("AT&T"), and I serve as Director - Public Policy Analysis, Network Engineering & Technologies. My responsibilities include developing public policy as it relates to interconnection with incumbent ILECs and the use of unbundled network elements that they are obligated to provide under the Telecommunications Act of 1996 ("the Act") and the Commission's rules implementing the Act. In that capacity I am required to understand the operational needs of the various business units so that their interests are reflected in the policy positions taken by AT&T. I also help those units understand how provisions of the Act and the Commission's rules affect their business plans. I support the other affiants in this testimony regarding the presentation of the data that AT&T retains regarding the scope of its local network and the availability of alternative access arrangements from other competitive carriers.
  6. The purpose of our current declaration is to provide additional factual background regarding AT&T's deployment of loop and transport facilities, and the extent to which it both purchases alternatives to incumbent LEC special access services from other competitive carriers and provides such alternatives to other carriers. Specifically, we will describe (1) AT&T's local network architecture, particularly the limited scope of AT&T's local network facilities in SBC's service territory; (2) why, as a matter of basic network engineering, AT&T's dedicated building access facilities are not "unique"; (3) the limited extent to which AT&T provides wholesale local private line services that can be viewed as an alternative to incumbent LEC special access service; and (4) the extent to which AT&T's purchases of dedicated access alternatives from competitive carriers are widely dispersed among numerous carriers. Each point is discussed in turn below.


  1. Although opponents of the AT&T-SBC merger characterize AT&T's local network as extensive, it is in fact quite limited. AT&T's local network employs the "spoke/hub" basic architecture used by most competitive local carriers. This means that when AT&T enters a local market, it typically does so first by deploying a metropolitan fiber facility (metro fiber), generally in the "downtown" area of the market, that connects strategic network locations such as local switches, nodes and AT&T's local points of presence ("POPs"). As is the case for other competitive local carriers, AT&T does not have direct access to individual customer locations in a large majority of instances, so, in the majority of cases, AT&T must lease loop (and often transport facilities) from the incumbent LEC. These facilities are accessed only at the incumbent LEC wire center. To connect its network to that of the incumbent and pick up the traffic from the loop and transport facilities that it is leasing, AT&T will collocate in an incumbent local serving office ("LSO") and extend a fiber lateral from its metro fiber to that facility. Such "facilities-based" collocations connect directly to the AT&T network and serve as a point where the demand generated by AT&T customers at that particular wire center is placed on AT&T's network.
  2. AT&T can also use its facilities-based collocation as a point where it accesses traffic served by other incumbent LSOs that are not directly connected to AT&T's local fiber network. AT&T leases ILEC transport to connect the LSOs in which it has established a facilities-based collocation to the LSOs where it does not have a facilities-based collocation. In connection with this activity, AT&T will sometimes deploy "non- facilities-based" collocations. Non-facilities-based collocations do not involve the deployment of local metro fiber but generally are instances in which AT&T has deployed multiplexing equipment that allows more efficient utilization of incumbent LEC special access services that are used to bring traffic to AT&T's fiber-based collocations. As such, non-facilities-based collocations are not part of AT&T's local fiber network.1
  3. In a minority of instances, AT&T is able to economically justify extending its local network to individual buildings that generally share three characteristics in common: (1) there is an AT&T customer willing to place substantial business directly onto the AT&T network; (2) the building is located in close proximity to its metro fiber; and (3) if spare conduit does not already exist, it is practically feasible to engage in new construction to connect the building to AT&T's metro fiber. When these conditions are met, and the business case demonstrates that the investment is prudent to undertake, AT&T extends a fiber lateral from a "splice point" (a pre-deployed physical point of connection to the metro fiber) from its metro fiber to the building. Typically, such splice points are established about every [REDACTED] along the metro fiber route.
  4. In the past, particularly prior to AT&T's acquisition of TCG in 1998, TCG deployed fiber extensions to buildings "on spec" in the hopes that it would ultimately win business to fill up that capacity. AT&T, however, discontinued that practice several years ago. Because of capital constraints, AT&T deploys fiber laterals only when it has a firm customer commitment to purchase service that independently justifies the construction. This is true with respect to both retail and wholesale service.
  5. AT&T has deployed local metro fiber networks in only 61 markets nationwide. The network includes metro fiber and associated dedicated fiber lateral connections to about [REDACTED] buildings where there is an active commercial presence in the building - a tiny faction of the buildings where AT&T serves retail and wholesale customers through the use of dedicated local loop facilities.
  6. In SBC's region, AT&T has deployed metro fiber facilities in only 19 metropolitan areas. 2 Like other competitive carriers in SBC's region, AT&T's metro fiber serves only the most urban portions of those markets where demand is most highly concentrated. As a result, AT&T has facilities-based collocations in only about [REDACTED] of SBC's central offices.3
  7. The substantial majority of AT&T's facilities-based collocations are in wire centers that are located in the areas of each local market where demand is most highly concentrated. Specifically, AT&T has [REDACTED] facilities-based collocations associated with its metro fiber in SBC territory. Most [REDACTED] of those collocations are in an SBC office that satisfies the "triggers" the Commission established for de-listing both DS1 and DS3 transport, and an additional [REDACTED] are in offices that satisfy the "triggers" the Commission established for de-listing DS3 transport.4 Thus, nearly [REDACTED] of AT&T's fiber-based collocations are in locations where the Commission has held that there are multiple competitors present or substantial potential revenues that would permit collocation by multiple competitors, or both.5
  8. AT&T has extended its network to [REDACTED] buildings in SBC's region that also have active commercial customers of AT&T.6 This is a tiny fraction of the hundreds of thousands of commercial buildings in SBC's service region that we understand are served with dedicated facilities.7
  9. AT&T is only one of many competitive carriers that operate in SBC's states. In 2004, AT&T purchased special access alternatives from [REDACTED] different suppliers in SBC states that in virtually all instances provide AT&T dedicated building access using their own network facilities. These carriers include: [REDACTED].
  10. Because of the breadth and scope of competitors in SBC's region, AT&T reaches only a small fraction of the total buildings served by other competitive carriers. In addition to keeping detailed data regarding the building locations served by AT&T's local network, AT&T has also developed a database regarding the buildings served by competitive carriers. The underlying data were provided to AT&T by competitive carriers seeking to provide AT&T with special access services to the buildings that they serve. These data are typically updated monthly or quarterly by the competitive carriers.
  11. Because AT&T uses the data for its own commercial purposes, it has a strong interest in ensuring that they are accurate as possible. AT&T generally seeks to eliminate a building from its "on net" list if AT&T learns that the building is not, in fact, currently served by competitive fiber. AT&T also eliminates from the database competitive carriers that do not satisfy AT&T's quality standards. As such, AT&T's data do not include the entirety of available competitive special access supply because the data do not reflect carriers that do not actively market special access services to AT&T, nor do they include carriers from which AT&T does not purchase special access services.8 For example, AT&T's data do not include buildings served by Sprint.
  12. As noted, AT&T has [REDACTED] commercial buildings "on net" in the SBC service areas. Competitive carriers serve many times that number. According to AT&T's competitive building inventory, [REDACTED] different competitive carriers have lit fiber connections to [REDACTED] buildings in the SBC service territories - a tiny fraction of commercial buildings in SBC's service territories.9 In addition, AT&T's competitive inventory shows that competitive carriers have "unlit" fiber connections to [REDACTED] buildings in SBC service territories. Thus, CLECs in aggregate have [REDACTED] direct fiber building connections.
  13. In a substantial number of instances, these competitive carriers serve the same buildings as AT&T. In SBC's service territories, [REDACTED] of AT&T's "on net" buildings are also served by "lit" CLEC fiber and [REDACTED] are also served by "unlit" CLEC fiber.
  14. Relying on information supplied by GeoTel, Cbeyond claims that the "loss" of AT&T as an independent competitor would result in a substantial reduction in the number of buildings directly served by competitive fiber facilities. Cebyond, however, limited its analysis to two markets: Cbeyond claims that AT&T serves 53% of unique buildings in Cleveland (Cbeyond at 26 & Wilkie Dec. H 18) and 64% of unique buildings in the Milwaukee, Wisconsin MSA. (Cbeyond at 26 & Wilkie Dec. H 20). AT&T's detailed data regarding the location of its network and the buildings served by competitive carriers - data AT&T relies upon for its own commercial purposes - demonstrate that these claims are overblown. AT&T's local network in these metro areas reaches only a small fraction of the buildings served either by SBC or other competitive carriers.
  15. Cleveland. As is the case nationally, AT&T's local network in Cleveland is limited. AT&T has only [REDACTED] commercial buildings on net in Cleveland, and [REDACTED] of those locations also served by competitive carriers. On the other hand, competitive carriers serve [REDACTED] additional unique buildings that are not directly served by AT&T's network as well as [REDACTED] buildings that are served by AT&T. AT&T's building inventory also shows that competitive carriers have deployed unlit fiber to another [REDACTED] buildings in Cleveland.
  16. As is the case generally, most of the buildings AT&T serves in Cleveland are "high demand" locations that generate at least one DS3 equivalent of retail demand and are candidates for competitive deployment by other carriers if AT&T's current customer(s) wished to switch providers.10 Of the [REDACTED] AT&T buildings not served by either lit or unlit competitive facilities, all but [REDACTED] have 1 DS3 equivalent or more of demand.
  17. AT&T's metro fiber in Cleveland is concentrated in dense urban areas. AT&T has [REDACTED] fiber-based collocations in the Cleveland MSA. [REDACTED] are in Tier 1 wire centers and [REDACTED] are in Tier 2 wire centers. AT&T's [REDACTED] collocations represent only [REDACTED] percent of SBC's switches in the Cleveland area.
  18. Milwaukee. The statistics for the Milwaukee, Wisconsin MSA are similar. Competitive carriers in Milwaukee serve [REDACTED] unique buildings with lit fiber and have deployed unlit fiber to another [REDACTED] buildings; AT&T has only [REDACTED] commercial buildings "on net." Of these [REDACTED] buildings, competitive carriers have deployed lit fiber to [REDACTED] of them. Moreover, of those [REDACTED] AT&T buildings not served by active competitive fiber, [REDACTED] have more than one DS3 of demand.
  19. AT&T's local metro fiber in Milwaukee is largely built out to the same wire centers as other competitive carriers in that market. AT&T has only [REDACTED] fiber-based collocations in the Milwaukee MSA. In contrast, there are 36 SBC switch locations in the Milwaukee area. [REDACTED] of AT&T's fiber-based collocations are in Tier 1 MS As that satisfy both the Commission's "triggers" for DS1 and DS3 transport.


  1. We understand that a particular concern raised by commenters in this proceeding is the act that AT&T has deployed last-mile fiber laterals to individual commercial buildings in SBC's service areas and that, as a result, competition must be analyzed on a route-by- route basis. See Broadwing at 22-23; Cbeyond at 25-30; CompTel at 16; Global Crossing at 11-13 & Farrell Dec. 1fl[ 23-28. In particular, we understand that they claim that even to the extent there is generally competition throughout an MSA, the loss of AT&T with respect to particular buildings is competitively significant. The evidence, however, shows that the fact that AT&T is the only carrier currently serving a building does not mean that other carriers could not economically deploy to that building too.
  2. As described above, AT&T's network is only connected to a small fraction of the total buildings served by competitive carriers, and in many cases, competitive carriers serve the same buildings as AT&T. Even with respect to the small number of AT&T's fiber laterals where there currently is no other competitor in the building, these buildings are potentially addressable by competitors that have demonstrated their ability to deploy fiber to many times more buildings than AT&T. AT&T today in most instances builds fiber laterals only where the customer has demand sufficient to support at least OC3-level service. The Commission has found, however, that competitive carriers are not "impaired" with respect to OCn-level loop facilities because the revenue opportunities associated are sufficient to overcome the economic barriers to deploying local loops. Triennial Review Order H 316 ("Services offered over OCn loops produce revenue levels which can justify the high cost of loop construction, providing the opportunity for competitive LECs to offset the fixed and sunk costs associated with loop construction."). Indeed, in the TR Remand Order C[fl{ 177-85), the Commission made a national finding of non-impairment that limits requesting carriers to leasing only a single DS3 loop facility and further limited DS1 and DS3 loops in many "high demand" locations where AT&T has deployed its own fiber laterals. See supra note 10.
  3. This is confirmed by the data on the extent to which AT&T's self-supplies or leases access to OCn facilities from competitive carriers as opposed to incumbent LECs in SBC's region. At the OC-3 level, AT&T self-provided about [REDACTED] of the circuits it uses in support of its service offerings and leases about [REDACTED] of those circuits from competitive carriers. At the OC-12 level, AT&T self-supplies about [REDACTED] of the circuits it uses in support of its service offerings and leases about [REDACTED] of those circuits from competitive carriers. Finally, AT&T self-supplies [REDACTED] of its OC-48 level circuits. These data thus show that self-supply is generally feasible at the OCn-level and that there is substantial supply of competitive OCn-level special access services.
  4. The fact that AT&T was able to deploy a fiber lateral to serve a customer in a particular location generally means that one (or more) customers in the location has OCn-level (or near OCn-level) of demand sufficient to support competitive deployment of facilities. Indeed, the very fact that AT&T was able to construct facilities to a particular building to serve a particular customer is evidence that the customer is willing to purchase services from a facilities-based competitor and that another carrier could also economically construct facilities to that same customer provided that it has a reasonably proximate metro fiber. Thus, when AT&T's contract with that customer expires and the customer's business is again "up for grabs," other carriers have a comparable ability to deploy their own facilities and win the customer that AT&T had when it initially won the customer's business.
  5. The evidence suggests that the majority of buildings served only by AT&T could also be economically served by other competitive carriers. There are [REDACTED] commercial buildings in SBC's territories that are served only by AT&T. Over [REDACTED] of those buildings have at least 1 DS3 equivalent of demand.
  6. Nor does AT&T have any special ability to build to additional buildings. Foremost, as shown above, the balance of the competitive carrier industry addresses many times the number of buildings in SBC's territory that AT&T reaches. While there may be some instances in which AT&T has the "closest" network, AT&T's fiber facilities are typically located in the dense urban areas that are also typically served by numerous other competitive carriers. Indeed, AT&T has examined the opportunities that exist in buildings within a mile of its network where it is currently leasing special access service to serve retail customers. As such, this analysis identified buildings where AT&T might be said to have an advantage because of the proximity of its network. Compared to the thousands of buildings that AT&T currently reaches using leased dedicated access facilities, only [REDACTED] could potentially satisfy AT&T's business case for construction designed to achieve access cost savings - i.e., where the savings from access cost reduction would by itself support deployment. Further, only [REDACTED] of those buildings have one DS3 or less of demand and are candidates for a potential build because of their close proximity to AT&T's metro fiber.11 But even with respect to these few "near net" buildings that AT&T estimates that it could potentially serve with its own fiber despite their relatively "low" demand, other competitive carriers may be as close, or closer, to these buildings and thus be in an equal or better position to build their own facilities.
  7. In the minority of cases where AT&T has deployed loops and currently serves retail service below the levels deemed to establish "impairment" by the Commission, even those situations do not necessarily indicate unique circumstances in which other parties would be unable to serve similarly situated (or even the same) customers. Foremost, the service provided to a customer and a building at any particular time is simply a snapshot that represents current conditions. Customers routinely add and disconnect demand as needs change, businesses relocate and/or contracts expire and are put out for bid. The fact that it was economically justified to place a customer location on AT&T's network in the past is not altered by subsequent changes in the customer's needs and/or shifts in its serving carrier.
  8. Thus, locations that appear as "low volume" today are a natural outcome of competitive forces at work. Because demand typically does not "disappear" from a location, it remains available to support future deployment by another competitive carrier. Indeed, in many locations where AT&T currently serves a "low volume" customer, it is because it has lost some of that customer's business to another competitive carrier after AT&T's initial customer contract expired. Such instances are evidence of the feasibility of multiple competitive deployment to a building.
  9. And even in the small number of instances in which AT&T (or its predecessors) deployed facilities when its customers in the building had "low" demand, it usually did so under conditions that would typically permit other carriers to do the same. For example, multi-location customers will sometimes not award a contract unless a carrier agrees to place all of their locations "on net." In such cases, the total revenues from the contract were sufficient to allow AT&T to economically deploy facilities to some low demand locations. Other carriers in similar circumstances would be able to extend their network to such low volume locations. In other instances, a low demand retail customers may be in a building where AT&T has also established a network location, such as a point-of-interconnection with another carrier or a network node.12 There are also buildings where AT&T is able to "hub" multiple buildings on a "campus" to a central point of aggregation - a build that other carriers could feasibly undertake in similar circumstances.13 Finally, other carriers, like AT&T are able to serve a "low demand" building via a fixed wireless loop in the situations that permit such deployment.14


  1. Merger opponents also greatly overstate the role of AT&T as a supplier of alternatives to incumbent LEC special access services. AT&T's local network is different from that of many other competitive carriers in one important respect. It was primarily designed and deployed to service AT&T's own retail customers, not to support wholesale "special access" alternatives to other carriers.
  2. As a result of AT&T's retail focus, AT&T sells less than [REDACTED] a year in wholesale local private line services in SBC's region. To put this figure in perspective, AT&T expects to generate over [REDACTED] in revenue from its local and long distance private line services.
  3. Indeed, AT&T cannot be considered a key supplier of private line services to the competitive carriers who have opposed this merger. Specifically, 25 competitive carriers assert have alleged that the combination of SBC's and AT&T's local network facilities raises competitive concerns.15 But [REDACTED] of these carriers do not purchase any local private line service at all from AT&T in SBC's region. Overall, AT&T supplies only about [REDACTED] local private line circuits to these 25 carriers that generate about [REDACTED] a month in revenues - which averages to only [REDACTED] per each of these competitive carrier.16 And [REDACTED] of these revenues are for OCn-level service where for which the Commission has held that there are relatively low barriers to competitive supply.
  4. This conclusion is not called into question by merger opponents' economic testimony that AT&T is a "key" bidder on private line services and in some circumstances offers the lowest price of rival competitive suppliers. See Cbeyond, Wilkie Dec. ]fl[ 22-27'. Quite obviously, if AT&T had the substantial competitive cost advantage suggested by Professor Wilkie, AT&T would have more than a miniscule share of the dedicated access "market" in SBC's territory. But more fundamentally, AT&T's ability to offer "low" private line rates depends heavily on the relative location of the buildings to be served in relation to AT&T's network. For locations that are already on AT&T's network (or in very close proximity to access points to AT&T's metro fiber such that AT&T can deploy a fiber lateral at a relatively low cost), AT&T may have the ability to supply the private line service at a "low" rate. And while AT&T's network is occasionally the closest to the location in question, the data discussed above show that this occurs very rarely, and that many other competitive carriers usually have a comparable (or superior) ability to serve those locations.
  5. Finally, AT&T is not a "reseller" of special access services, as some merger opponents have claimed. AT&T does not purchase special access services from SBC (or any ILEC for that matter) and then resell them to other CLECs. Thus, AT&T is not using resale as a means to engage in arbitrage and put pressure on SBC's special access prices.
  6. The reason why AT&T does not engage in such pure resale is simple: such a practice is unlikely to generate any profits. Even where AT&T obtains from other incumbent LECs "volume-based" discounts that are greater than those earned by most other special access purchasers - and we understand that SBC does not offer such discounts - the spread between the discounts AT&T obtains and other carriers obtain is small. The transaction costs of engaging in the resale business would wipe out any margins AT&T might hope to earn.
  7. AT&T does use SBC special access services as an input to many of its local and long distance service offerings, including, in some instances, AT&T's local private line services that are purchased at wholesale by other carriers. AT&T refers to services for which it leases a portion of the local network from another carrier as "Type II" service. With only a literal handful of exceptions, however, AT&T provides Type II local private line services only where AT&T has self-supplied the transport portion of the service and one of the tails of the service. Thus, most of the private line circuits AT&T sells are "Type I" services provided over AT&T's own facilities and only a minority of are provided over special access leased from SBC.
  8. Further, the vast majority of AT&T's Type II sales are to existing customers. AT&T sells very little Type II private line service to new customers because of the inherent disadvantage in selling the service in competition with carriers able to supply the service entirely over their own facilities.
  9. In all events, AT&T's sales of Type II local private line service are not significant. AT&T currently provides less than [REDACTED] a year in wholesale local private line services in SBC's territory. Of that amount, only [REDACTED] is associated with Type II services for which AT&T leases a portion of the circuit from SBC. And, as explained, the majority of these revenues are associated with AT&T's own local facilities because most of the circuit is provided over AT&T's network.


  1. We next address the concern raised by some merger opponents that the loss of AT&T as a purchaser of access services from competitive carriers threatens the viability of these carriers. See CompTel at 19. The facts belie this claim. According to the Commission, the overall special access market is over $14 billion a year. See Statistics on Common Carriers, Table 2.8 (Oct. 12, 2004) (reporting that the RBOCs by themselves had over $14 billion in special access revenues in 2003). Not only are these services purchased by other major IXCs such as MCI, Sprint, Qwest, Global Crossing and Level 3, but also by wireless carriers, system integrators and any retail provider of bandwidth intensive telecommunications or data applications. And these other purchasers represent the majority of special access purchases nationwide. In fact, AT&T's nationwide special access expenditures on special access (from both incumbent and competitive carriers) amount to about [REDACTED] a year.
  2. Nationwide, AT&T spends only [REDACTED] on alternative access services provided by competitive carriers, and within SBC's region, AT&T spends only about [REDACTED] annually with competitive carriers. In stark contrast, AT&T purchases over [REDACTED] a year in special access from SBC.
  3. AT&T's purchases are also spread among a wide variety of carriers. In 2004, AT&T purchased special access services from over [REDACTED] different competitive carriers nationwide. Over [REDACTED] of these carriers do not provide special access service at all in the SBC region and thus are unaffected by the merger. And with regard to the remainder that sell special access alternatives in SBC's region, [REDACTED].
  4. The following table lists AT&T's 10 largest competitive special access suppliers in SBC's region for the calendar year 2004, and shows the relative percentage of AT&T's purchases from those carriers in SBC's region versus AT&T's special access purchases nationwide.17 [REDACTED]


I declare under penalty of perjury that the foregoing is true and correct.

May 9, 2005

________/s/_ Anthony Fea ____________
Anthony Fea


I declare under penalty of perjury that the foregoing is true and correct.

May 9, 2005

_______/s/Anthony Giovannucci ________
Anthony Giovannucci


I declare under penalty of perjury that the foregoing is true and correct.

May 9, 2005

Bob Handal


I declare under penalty of perjury that the foregoing is true and correct.

May 9, 2005

Michael Lesher


I declare under penalty of perjury that the foregoing is true and correct.

May 9, 2005

C. Michael Pfau

Attachment 2





Before the
Washington, DC 20554

In the Matter of

Verizon Communications Inc. and   
MCI, Inc.
Applications for Approval of
Transfer of Control

WC Docket No. 05-75


  1. My name is Jonathan P. Powell. I am Director, Wholesale Pricing - Data for MCI. My responsibilities include the competitive positioning and pricing of MCI's wholesale Metro Private Line service. My business address is 6929 North Lakewood, Tulsa, Oklahoma.
  2. My name is Peter H. Reynolds. I am Director, National Carrier Management and Initiatives for MCI. My responsibilities include managing MCI's relationships with CLECs and other access vendors. My business address is 22001 Loudoun County Parkway, Ashburn, Virginia.
  3. My name is Edwin A. Fleming. I am Senior Manager of Strategic Business Planning for MCI. My responsibilities include evaluating and managing building additions to MCI's local network. My business address is 2655 Warrenville Road, Downers Grove, Illinois.
  4. The purpose of this declaration is to (1) explain that any volume discount that MCI may obtain for Verizon special access services plays little or no role in MCI's Metro Private Line pricing; and (2) discuss the large number of competitive alternatives to MCI's wholesale Metro Private Line service.

I. MCI's Limited Use of Verizon Special Access Services

  1. As was discussed in the Declaration of Jonathan P. Powell and Stephen M. Owens (Powell/Owens Declaration), MCI has constructed local fiber networks in several cities in Verizon's territory. Those local fiber networks extend to approximately [BEGIN PROPRIETARY END PROPRIETARY] "on-net" buildings in Verizon's territory,1 a figure that includes [BEGIN PROPRIETARY END PROPRIETARY] fiber-based collocations in Verizon central offices. Most of these on-net buildings - approximately [BEGIN PROPRIETARY END PROPRIETARY] - are in the Verizon-East region.
  2. In order to reach off-net customer locations, MCI obtains high-capacity circuits from other CLECs or, more commonly, from Verizon's special access tariffs. MCI purchases most of those special access circuits pursuant to one of Verizon's term plans. The rates that MCI pays Verizon for those special access circuits are the same rates that MCI pays in those areas in which MCI does not have local facilities. More generally, Verizon's rates do not vary by MSA or wire center; they vary only according to tariff filing entity, state, or, in the case of some services, one of three pricing zones in a state.
  3. MCI uses its local fiber networks both (1) to provide MCI retail customers with access to MCI's long-haul voice, data, and Internet services; and (2) to provide retail and wholesale "Metro Private Line" services. Depending on the application, MCI's Metro Private Line service is equivalent to either the incumbent LECs' special access service or local private line service. Metro Private Line circuits are dedicated intraLATA high-capacity circuits that connect carrier hotels, incumbent LEC central offices, IXC POPs, wireless POPs, ISP POPs, office buildings, and other end user buildings. MCI's wholesale Metro Private Line customers include IXCs, CLECs, wireless carriers, and ISPs.
  4. MCI classifies Metro Private Line circuits into four different categories, depending on the mix of MCI facilities and third-party facilities that MCI uses to provision the Metro Private Line circuit. A Type I circuit is provisioned entirely "on-net," i.e., it connects two on-net buildings using only MCI fiber. The other three types of Metro Private Line circuits - Type II, Type III, and Type IV - are provisioned, to varying degrees, using special access circuits obtained from another local carrier - usually, but not exclusively, the incumbent LEC.
  5. A Type II circuit connects an on-net building to an off-net building. Most of the circuit is provisioned using MCI's local fiber, but a small piece is provisioned using the facilities of another local carrier - typically, an incumbent LEC special access "channel termination" that extends MCI's network to the off-net building. A Type III circuit uses two incumbent LEC channel terminations, to reach an off-net building at each end of the circuit, and MCI fiber in the "middle." A Type IV circuit uses no MCI facilities; it is simple resale of an incumbent LEC special access circuit.
  6. Although Metro Private Line Type II, Type III, and Type IV circuits use incumbent LEC special access services, and although MCI is a large purchaser of incumbent LEC special access services, any volume discounts that MCI may receive on its special access purchases are not a significant factor in the pricing of Metro Private Line services.
  7. Notably, more than [BEGIN PROPRIETARY END PROPRIETARY] percent of MCI's wholesale Metro Private Line revenue is derived from circuits that are entirely on-net and do not use incumbent LEC special access at all, i.e., Type I circuits. Consequently, any special access volume discount that MCI may receive plays no role in MCI's pricing of a substantial majority of its Metro Private Line services.
  8. Most of the remainder of MCI's wholesale Metro Private Line revenue is derived from Type II circuits, which generally use only a single channel termination. Less than 2 percent of MCI's wholesale Metro Private Line revenue is derived from Type III circuits. And MCI does not currently offer Type IV circuits, i.e., MCI does not currently engage in the simple resale of incumbent LEC special access services.2
  9. Little or none of the differential between the price that a wholesale customer would pay for an MCI Type II circuit and the incumbent LEC's price for an equivalent circuit is attributable to any special access volume discount that MCI may receive. Generally, the only special access component of a Type II circuit is a single channel termination. In most cases, the incumbent LECs' channel termination prices are largely independent of volume. For example, MCI obtains channel terminations under [BEGIN PROPRIETARY END PROPRIETARY]
  10. Because the incumbent LECs' special access rates are largely independent of volume, the price that MCI pays for the special access service used in a Type II circuit -typically, only a single channel termination -- is much the same as the price that a Metro Private Line customer would pay if it purchased the channel termination, for the same term, directly from the incumbent LEC. And even in those instances in which MCI may obtain some modest additional volume discount, that additional discount is largely offset by MCI's internal costs, including the cost of submitting the order for the channel termination to the incumbent LEC. Any differential between the MCI Metro Private Line price for a Type II circuit and the incumbent LEC's price for an equivalent circuit is thus almost exclusively attributable to the on-net part of the circuit, not to any volume discount that MCI may receive for the special access part of the circuit.

II. Competition for MCI's Wholesale Metro Private Line Service

  1. MCI's wholesale Metro Private Line business represents only a small fraction of MCI's total revenue. In the Verizon-East region, for example, MCI's wholesale Metro Private Line revenue is only approximately [BEGIN PROPRIETARY END PROPRIETARY] per year.
  2. In each of the areas in which MCI provides wholesale Metro Private Line services in the Verizon territory, it faces competition from several other CLECs. As was discussed in the Powell/Owens Declaration, other CLECs have pursued much the same market entry strategy as MCI, constructing their fiber networks on high-density routes in the downtown core or in suburban areas with high business concentration. Depending on the city, service providers competing with MCI's Metro Private Line service in the Verizon region include CLECs such as AT&T, Time Warner, XO, and TelCove; new fiber wholesalers such as AboveNet, FiberNet, and OnFiber, and utilities such as Progress Telecom, ConEd Communications, and PPL Telecom.
  3. In some cases, MCI is also a customer of these CLECs and fiber providers. MCI has entered into agreements to purchase dedicated circuits from several CLECs in the Verizon region, including [BEGIN PROPRIETARY END PROPRIETARY]. MCI has also obtained dark fiber from utilities and other fiber wholesalers, including [BEGIN PROPRIETARY END PROPRIETARY].
  4. MCI maintains a database of buildings that have been "lit" by MCI or one of the CLECs with which MCI has an agreement to purchase dedicated access services. In Albany, NY, MCFs database shows [BEGIN PROPRIETARY END PROPRIETARY] lit buildings. MCI is the sole CLEC in no more than [BEGIN PROPRIETARY END PROPRIETARY] of those buildings. Similarly, in Baltimore, MD, MCI is the sole CLEC in no more than [BEGIN PROPRIETARY END PROPRIETARY] of the [BEGIN PROPRIETARY END PROPRIETARY] lit buildings in MCI's database; in Pittsburgh, PA, MCI is the sole CLEC in no more than [BEGIN PROPRIETARY END PROPRIETARY] of the [BEGIN PROPRIETARY END PROPRIETARY] lit buildings in MCI's database; in Philadelphia, PA, MCI is the sole CLEC in no more than [BEGIN PROPRIETARY END PROPRIETARY] of the [BEGIN PROPRIETARY END PROPRIETARY] lit buildings in MCI's database; in New York, NY, MCI is the sole CLEC in no more than [BEGIN PROPRIETARY END PROPRIETARY] of the [BEGIN PROPRIETARY END PROPRIETARY] lit buildings in MCI's database; and in Washington, DC, MCI is the sole CLEC in no more than [BEGIN PROPRIETARY END PROPRIETARY] of the [BEGIN PROPRIETARY END PROPRIETARY] lit buildings in MCI's database.
  5. The CLECs and fiber wholesalers that have networks in the Verizon region are well-positioned to compete for MCI's Metro Private Line revenue. First, CLECs are already present in a substantial fraction of MCI's on-net buildings. For example, the lit building lists provided to MCI by the CLECs with which MCI has agreements to purchase dedicated access services show that those CLECs alone have a presence in at least [BEGIN PROPRIETARY END PROPRIETARY] of MCI's approximately [BEGIN PROPRIETARY END PROPRIETARY] on-net buildings in the Verizon-East region.
  6. It should be stressed that this figure understates the extent to which CLECs are present in MCI lit buildings in Verizon-East territory. MCI only has information about the buildings that have been lit by the CLECs with which MCI has an agreement to purchase dedicated access services. MCI does not know which MCI on-net buildings have also been lit by the other CLECs that have networks in Verizon-East territory. Other CLECs that are known to have lit buildings in the Verizon-East region, and thus may be present in MCI on-net buildings, include [BEGINPROPRIETARY END PROPRIETARY]. Cavalier, for example, advertises high-bandwidth "metro transport" services that rely on Cavalier's "dense footprint in 215 Verizon central offices."3
  7. Furthermore, MCI's wholesale Metro Private Line demand is concentrated in the subset of buildings that are most likely to be served by multiple CLECs or fiber providers. Specifically, MCI's Metro Private Line wholesale business has been focused on the provision of high-capacity circuits between "carrier" buildings such as IXC POPs, wireless POPs, ISP POPs, carrier hotels, and incumbent LEC central offices. For example, the Metro Private Line circuits that MCI sells to wholesale customers at the 60 Hudson Street and 111 8th Avenue carrier hotels in New York are typically OC-n level circuits. Because those carrier hotels and other carrier buildings are very high traffic locations, they are also the locations in which MCI faces the most competition for its wholesale business. For example, MCI faces competition at the 60 Hudson Street carrier hotel from at least AT&T, Time Warner, Level 3, and XO.
  8. Finally, most MCI on-net buildings - including those that to date have been lit only by MCI - are readily addressable by multiple CLECs or fiber providers. As is shown in Attachment 1, which relies on data previously presented in Exhibit 12B of the Lew/Lataille Declaration, 80 percent of MCI's on-net buildings are concentrated in only 111 of the [BEGIN PROPRIETARY END PROPRIETARY] Verizon wire centers that have MCI on-net buildings. Attachment 1 also shows that all but 10 of those 111 Verizon wire centers have three or more competitive fiber providers, and that those 111 wire centers have an average of 10 competitive fiber provider networks.
  9. Any of the multiple CLECs and fiber wholesalers that have constructed networks in the Verizon wire centers in which MCI on-net buildings are concentrated could readily extend their networks to an MCI on-net building.
  10. Verizon Central Offices As is discussed above, some of MCI's on-net buildings are Verizon central offices. In determining which Verizon central offices to bring on-net, MCI targeted those central offices that had the highest levels of demand and, consequently, provided sufficient revenue to warrant the cost of facilities construction. Because the MCI fiber-based collocations are in such high-demand central offices, and because MCI was able to "prove in" the fiber-based collocations, it is apparent that other CLECs could also extend their networks to those central offices (if they have not done so already).
  11. Of the [BEGIN PROPRIETARY END PROPRIETARY] Verizon central offices in which MCI's local network has a fiber-based collocation, [BEGIN PROPRIETARY END PROPRIETARY] or 74 percent, have been designated by Verizon as either Tier 1 or Tier 2 central offices under the transport impairment tests that the FCC adopted in the Triennial Review Remand Order (see Attachment 2). And, in many applications, Metro Private Line circuits that terminate in Verizon wire centers are equivalent to entrance facilities, for which the FCC has made a finding of non-impairment.
  12. End User Buildings CLECs and other fiber providers could also readily extend their networks to office buildings, corporate campuses, and other MCI on-net end user buildings. The fact that MCI has lit a building shows that there are no building access issues and that the building is a communications-intensive location that generates sufficient revenue to justify the cost of facilities construction.
  13. MCI generally does not even consider a building for a "building add" unless there is customer demand of a DS3 or more, and adds a building only if the available revenue is sufficient to recover the cost of construction within the payback period specified by MCI's corporate guidelines. A sample consisting of the most recent 20 approved building adds in Verizon-East territory for which "day one" circuit counts were specified in the proposal showed that all but two had initial circuit demand of a DS3 equivalent or more, and demand in the two buildings whose initial circuit demand was below the DS3 level was projected to ultimately increase above that level.4
  14. Furthermore, a review of current circuit data for the on-net buildings with MCI localfiber in Verizon territory showed that a significant majority have current demand at the OCn or near-OCn level. Specifically, [BEGIN PROPRIETARY END PROPRIETARY] percent of those buildings have current demand of two or more DS3 equivalents.5 And [BEGIN PROPRIETARY END PROPRIETARY] percent have current demand of one or more DS3 equivalent.6
  15. In every city, MCI's on-net buildings exhibit high levels of circuit demand. In Albany, NY, for example, [BEGIN PROPRIETARY END PROPRIETARY] percent of MCI's on-net buildings have current circuit demand of 2 DS3 equivalents or more; in Baltimore, MD, [BEGIN PROPRIETARY END PROPRIETARY] percent of MCI's on-net buildings have current circuit demand of 2 DS3 equivalents or more; in New York, NY, [BEGIN PROPRIETARY END PROPRIETARY] percent of MCI's on-net buildings have current circuit demand of 2 DS3 equivalents or more; in Philadelphia, PA, [BEGIN PROPRIETARY END PROPRIETARY] percent of MCI's on-net buildings have current circuit demand of 2 DS3 equivalents or more; in Pittsburgh, PA, [BEGIN PROPRIETARY END PROPRIETARY] percent of MCI's on-net buildings have current circuit demand of 2 DS3 equivalents or more; and in Washington, DC, [BEGIN PROPRIETARY END PROPRIETARY] percent of MCI's on-net buildings have current circuit demand of 2 DS3 equivalents or more. Overall, in those six cities, [BEGIN PROPRIETARY END PROPRIETARY] percent of MCI's on-net buildings have current circuit demand of 2 DS3 equivalents or more.
  16. Because MCI's on-net buildings are high-demand locations, and because MCI has no material cost advantage in constructing outside plant, obtaining building access, or preparing the "POP space" in the building, the fact that MCI was able to "prove in" the extension of its network to a building shows that other CLECs could do the same.
  17. At least 80 percent of MCI's on-net buildings are either in wire centers that meet the "triggers" that the FCC established for de-listing unbundled DS3 loops, or have sufficient demand to justify the use of OC-n circuits, which are not available as unbundled network elements. Specifically, 51 percent of MCI's on-net buildings are in wire centers that meet the DS3 impairment test adopted in the Triennial Review Remand Order, and an additional 29 percent of MCI's on-net buildings have current circuit demand at the OC-n or near OC-n level, i.e., two or more DS3 equivalents.

Attachment 3



1. Reply Declaration of W. Robert Majure ("Majure Reply Decl."), ¶¶ 3-10; see also U.S. Dep't of Justice & Federal Trade Comm'n, Horizontal Merger Guidelines (rev. ed. 1997) ("Merger Guidelines").

2. See Merger Guidelines § 1.11. The approach set forth in the Merger Guidelines is consistent with legal precedent, which defines product markets by evaluating the extent of cross-elasticity of demand between a product and possible substitutes; in other words, the readiness and ability of consumers of one product to turn to the other product. See Brown Shoe Co. v. United States, 370 U.S. 294, 325 (1962); Rothery Storage & Van Co. v. Atlas Van Lines, Inc., 792 F.2d 210, 218 (D.C. Cir. 1986) (definition of the "relevant market" rests on a determination of available substitutes).

3. Local private lines are dedicated point-to-point circuits offered over copper and/or fiber-optic transmission facilities that originate and terminate within a single metropolitan area. Complaints ¶ 3.

4. Majure Reply Decl. ¶ 6.

5. See, e.g., Order and Notice of Proposed Rulemaking, In re: Special Access Rates for Price Cap Local Exchange Carriers; AT&T Corp. Petition for Rulemaking to Reform Regulation of Incumbent Local Exchange Carrier Rates for Interstate Special Access Services, 20 F.C.C.R. 1994 (Jan. 31, 2005). To avoid burdening the Court with paper, the United States has not provided copies of publicly available records, including materials available on the Federal Communications Commission website. The United States would be pleased to provide copies of these materials at the Court's request.

6. The NYAG even makes the unsupportable argument that a geographic market can never be smaller than a metropolitan area. Memorandum of Eliot Spitzer, Attorney General of the State of New York, in Response to the August 7, 2006 Submission of the United States at 3 (Sept. 5, 2006) ("NYAG Resp."). However, Brown Shoe, the case on which NYAG relies, merely requires that the geographic market "correspond to the commercial realities of the industry and be economically significant." 370 U.S. at 336-37 (internal quotation omitted). It places no lower limit on the size of a geographic market. As described infra, a single-building geographic market reasonably reflects the "commercial reality" of the LPL product market.

7. See, e.g., Complaint, United States v. Echostar Communications Corp., Civ. No. 1:02CV02138 (D.D.C. filed Oct 31, 2002), ¶¶ 30-31 (alleging that each residence constitutes a separate geographic market but aggregating, for ease of analysis, residences that face the same choice of providers). In the present cases aggregation is not useful because different providers own the connections to each building. The State of New York, represented by amicus NYAG, joined the Echostar complaint.

8. In reviewing these two mergers, the FCC asserted that the relevant geographic market "is a particular customer's location, since it would be prohibitively expensive for an enterprise customer to move its office location in order to avoid" a price increase. Memorandum Opinion and Order, In re: SBC Communications Inc. and AT&T Corp. Applications for Approval of Transfer of Control, 20 F.C.C.R. 18,290 (rel. Nov. 17, 2005), ¶ 28 ("FCC SBC/AT&T Order"); Memorandum Opinion and Order, In re: Verizon Communications Inc. and MCI, Inc. Applications for Approval of Transfer of Control, 20 F.C.C.R. 18,433 (rel. Nov. 17, 2005), ¶ 28 ("FCC Verizon/MCI Order") (collectively, "FCC Merger Orders").

Similarly, in its 2003 Triennial Review Order, the Commission directed the states to evaluate the extent of competition separately for each "particular customer location," to assist in determining impairment pursuant to under Section 251 of the Telecommunications Act of 1996. Report and Order, In re: Review of the Section 251 Unbundling Obligations of Incumbent Local Exchange Carriers, 18 F.C.C.R. 16,978 (Aug. 21, 2003), ¶ 328, vacated and remanded in part, aff'd in part, United States Telecom Ass'n v. FCC, 359 F.3d 554 (D.C. Cir. 2004), cert denied, 543 U.S. 925 (2004). In its 2005 Triennial Review Remand Order, subsequent to litigation, the FCC recognized that "a properly designed building-specific test could assess variations in impairment far more subtly than could a wire center or MSA-based approach . . . ." Order on Remand, In re: Review of the Section 251 Unbundling Obligations of Incumbent Local Exchange Carriers, 20 F.C.C.R. 2533 (Feb. 4, 2005) ("FCC TRRO"),¶¶ 155, 157-61, petition for review denied by Covad Communications Co. v. FCC, 450 F.3d 528 (D.C. Cir. 2006).

9. Merger Guidelines § 1.21.

10. FTC v. Arch Coal, Inc., 329 F. Supp. 2d 109, 123 (D.D.C. 2004) (holding that the "relevant geographic market in which to examine the effects of a merger is 'the region in which the seller operates, and to which the purchaser can practicably turn for supplies'" (quoting FTC v. Cardinal Health, Inc., 12 F. Supp. 2d 34, 49 (D.D.C. 1998) (citing Tampa Elec. Co. v. Nashville Coal Co., 365 U.S. 320 (1961))).

11. Brown Shoe, 370 U.S. at 336.

12. Majure Reply Decl. ¶ 7.

13. Declaration of Lee L. Selwyn (Sept. 5, 2006) ("Selwyn NASUCA Decl."), attached to Reply of Amicus the National Association of State Utility Consumer Advocates Regarding the Public Interest (Sept. 5, 2006) ("NASUCA Reply").

14. Declaration of Lee L. Selwyn on Behalf of AT&T Corp., In re: Petition of Qwest Corp. for Forbearance Pursuant to 47 U.S.C. § 160(c) in the Omaha Metropolitan Statistical Area, WC Docket No. 04-223 (Aug. 24, 2004), ¶ 32.

15. Id. ¶ 23 (emphasis in original).

16. Id. ¶ 32.

17. Id.

18. COMPTEL's Response to the Department of Justice's Supplemental Submission at 19-21 (Sept. 5, 2006) ("COMPTEL Resp.").

19. See Covad, 450 F.3d at 544 ("[T]he CLECs argue that impairment should be assessed on a building-by-building (as opposed to a wire center-by-wire center) basis."). Petitioners in Covad included COMPTEL, XO Communications, Xspedius Communications, LLC, Covad Communications Co., Time Warner Telecom Inc., Alpheus Communications, LP, ATX Communications, Inc., CTC Communications Corp., McLeodUSA Telecommunications Services, Inc., Mpower Communications Corp., Pac-West Telecomm, Inc., RCN Telecom Services, Inc., TDS Metrocom, LLC, Eschelon Telecom, Inc., NuVox, Inc., and US LEC Corp. See Opening Brief of CLEC Petitioners and Intervenor in Support, Covad v. FCC, No. 05-1095 (D.C. Cir. filed July 26, 2005).

20. See, e.g., Comments of Time Warner Telecom, In re: Special Access Rates for Price Cap Local Exchange Carriers, WC Docket No. 05-25 (June 13, 2005), at 1, 7, 25 ("[A]ny regulatory framework for special access channel terminations must reflect the fact that carriers decide to construct loop facilities to commercial customers based on the characteristics of particular buildings."); Letter from Broadview, Covad, CBeyond, Eschelon, KMC, NuVox, XO, and Xspedius, to Marlene H. Dortch, FCC, Re: WC Docket No. 04-313, CC Docket No. 01-338 (Dec. 8, 2004), at 1-5 (supporting building-specific analysis); Letter from Thomas Jones, Counsel for Time Warner Telecom, to Marlene H. Dortch, FCC, Re: CC Docket No. 01-338, WC Docket No. 04-313 (Dec. 1, 2004), at 3-6 (advocating building-by-building assessments for both actual deployment and potential entry).

21. Declaration of Keith L. Kassien (Sept. 1, 2006) ("Kassien Sprint Decl."), ¶¶ 13-14, attached to Response of Sprint Nextel Corporation to the United States' Submission in Response to the Court's Minute Order of July 25, 2006 (Sept. 5, 2006) ("Sprint Resp.").

22. ACTel suggests that the appropriate geographic market may consist of each individual floor in each building. ACTel's Response to the United States' Submission Pursuant to the Court's Minute Order of July 25, 2006, at 7-8, 11-14 (Sept. 5 , 2006) ("ACTel Resp.").

23. Several amici have suggested that a broader geographic market is dictated by the way that retail enterprise customers purchase the services that are provided over LPL. COMPTEL argues that the customer statements show that retail customers look for suppliers of these services that can serve multiple locations, and based on past practices of the Department, the market could be the entire United States. COMPTEL Resp. at 28-30. The NYAG's economist notes that retail customers purchase network services that use LPL to connect multiple locations. Declaration of Nicholas Economides (Sept. 5, 2006) ("Economides NYAG Decl."), ¶¶ 25-27, attached to NYAG Resp. Both conclude that a geographic market for retail services, therefore, cannot be a single building and Dr. Economides suggests that it must be at least as large as a metropolitan area. As discussed in Dr. Majure's Reply Declaration, the way end-user customers purchase services provided over LPL does not preclude a finding that each individual building is a separate market for LPL. Majure Reply Decl. ¶¶ 8-9 & n.7.

24. Declaration of W. Robert Majure (Aug. 7, 2006) ("Majure Decl."), ¶ 14, attached to United States' Submission in Response to the Court's Minute Order of July 25, 2006 (Aug. 7, 2006) ("United States' Submission").

25. Majure Reply Decl. ¶ 10.

26. Although the merged firm has changed its name to AT&T, Inc., the United States refers to it herein as "SBC" to avoid confusion.

27. Majure Decl. ¶ 13.

28. See, e.g., NYAG Resp. at 11-13; Reply of the New Jersey Rate Counsel to the United States' Submission in Response to the Court's Minute Order of July 25, 2006 at 5-11 (Sept. 5, 2006) ("NJRC Reply").

29. Majure Reply Decl. ¶ 11.

30. Several amici offer HHI calculations for product markets that are clearly not alleged in the United States Complaints. Much of the NJRC's response, for example, offers an HHI analysis for five mass-market and retail products. NJRC does not even attempt to argue that these markets are within the scope of the United States' Complaints. NJRC Reply at 5-11. NYAG's economist likewise reported that the New York Public Service Commission staff calculated HHIs for telecommunications services to various retail market segments. Economides NYAG Decl. ¶ 55. Sprint and COMPTEL offer HHIs that attempt to reflect concentration levels for LPL markets, but as explained in Dr. Majure's reply declaration, these numbers are unreliable. Majure Reply Decl. ¶ 13 & n.15.

31. See COMPTEL Resp. at 30-31; Sprint Resp. at 25-26; NJRC Reply at 7-11.

32. U.S. Dep't of Justice & Federal Trade Comm'n, Commentary on the Horizontal Merger Guidelines at 15-16 (2006) ("Merger Guidelines Commentary"); see also United States v. Baker Hughes Inc., 908 F.2d 981, 984 (D.C. Cir. 1990) (stating that evidence of market concentration simply provides a convenient starting point for a broader inquiry into future competitiveness); Arch Coal, 329 F. Supp. 2d at 130 (recognizing that "this circuit has cautioned against relying too heavily on a statistical case of market concentration alone").

33. Merger Guidelines Commentary at 15.

34. Merger Guidelines § 3.0; see also Majure Decl. ¶ 15.

35. Baker Hughes, 908 F.2d at 988.

36. This investigative effort included information requests contained in CIDs issued to CLECs. Some of the responses have previously been provided to the Court. Majure Decl., Attachs. Tab 9, CLEC Interrogatory Responses.

37. Majure Decl. ¶ 14. ACTel complains that the fiber maps that the Department used do not tell whether the fiber is "available for wholesale at all." ACTel Resp. at 7, 20. But all of the carriers whose fiber maps the United States used offer LPL on a wholesale basis. See Majure Decl., Attachs. Tab 7, Note on Overlapping Fiber Maps. The FCC Order approving the Verizon/MCI merger identifies more than 25 CLECs that sell LPL at wholesale. FCC Verizon/MCI Order ¶ 30.

38. One amicus declarant, however, complains that the United States did not state what its entry criteria were. See Selwyn NASUCA Decl. ¶ 50. This is incorrect. See Majure Decl. ¶ 14 n.7 (stating the distance and demand thresholds the United States utilized in determining the likelihood of entry).

39. Sprint Resp. at 10 (stating that Sprint rarely enters for DS-1 or DS-3, but will occasionally for OC-3). Consistent with this, the United States' entry criteria never predicted a likelihood of entry for buildings with a single DS-3 or less of demand for dedicated services, but did predict entry for instances where demand exceeded an OC-3 and competitive fiber was sufficiently close. (For the intermediate case, 2 DS-3s, which is between a single DS-3 and an OC-3 (roughly equivalent to 3 DS-3s), the United States' screen predicts a likelihood of entry only where competitive fiber is extremely close – within one-tenth of a mile.)

40. ACTel Resp. at 14-16 (stating that "the Government's formula simply ignored all the other factors in Paragraph 27 - specifically, (3) the availability of capital, (4) physical barriers such as railbeds, and, most importantly, (5) the difficulty in securing consents from building owners and municipalities"); Selwyn NASUCA Decl. ¶ 46 (same); Sprint Resp. at 7-8 (stating that competitive access vendors face large capital expenditures, often do not have the economies of scale to justify construction, and have difficulty obtaining access to buildings and rights of way); Economides NYAG Decl. ¶¶ 68-70 (stating that DOJ only considered two of the five factors stated in the Complaints).

41. Building-specific barriers such as building access costs or physical barriers (such as rivers) may, in rare instances, make a difference in the likelihood of entry, but those instances are by far the exception rather than the rule. Building access costs, i.e., the costs of "securing the necessary consent from building owners," Complaints ¶ 27(e), typically constitute only a small fraction of the costs of constructing fiber laterals to a building. The fact that AT&T or MCI has already built into all the 2-to-1 buildings subject to the United States' entry analysis tends to suggest that these buildings are not subject to some unusually troublesome entry obstacle such as a river or particularly obstructionist landlord.

42. Majure Reply Decl. ¶¶ 16, 18; Majure Decl. ¶ 14.

43. See, e.g., COMPTEL Resp. at 22-24; ACTel Resp. at 16-17; Sprint Resp. at 32.

44. See, e.g., Baker Hughes, 908 F.2d at 988 ("If barriers to entry are insignificant, the threat of entry can stimulate competition in a concentrated market, regardless of whether entry ever occurs." (emphasis in original)).

45. Economides NYAG Decl. ¶¶ 11-12.

46. Selwyn NASUCA Decl. ¶ 51.

47. Majure Reply Decl. ¶ 17; U.S. Dep't of Justice, Antitrust Division Policy Guide to Merger Remedies at 3 (2004) ("Merger Remedies Guide").

48. Merger Remedies Guide at 4.

49. Majure Decl. ¶ 16.

50. COMPTEL has directed the Court's attention to the settlement recently announced by the Department relating to the acquisition of Midwest Wireless Holdings LLC by ALLTEL Corporation as an example of "the Guidelines 'Done Right.'" Motion to Supplement COMPTEL's Response to the DOJ's Supplemental Submission (Sept. 14, 2006), at 2. Based on the outcome in that case, COMPTEL argues that the settlements here should require broader divestitures. However, the "one size fits all" approach for crafting remedies suggested by COMPTEL is inappropriate. The Department determines what remedies are appropriate on a case-by-case basis, taking into account the facts of the particular merger, the nature of competition in the market, and the specific harm identified. The Department has obtained a wide range of remedies including, where appropriate, divestitures that include on-going businesses and customers when necessary to address the harm identified. In these cases, for the reasons discussed in Dr. Majure'sdeclaration, divestitures of customers were not warranted or necessary. Majure Decl. ¶ 18.

51. ACTel Resp. at 6.

52. Proposed Final Judgments § IV(H).

53. Competitive Impact Statements at 9 (emphasis added) ("CISes").

54. See FCC Verizon/MCI Order ¶ 30 (listing wholesale providers, including [REDACTED TEXT] ).


56. Merger Remedies Guide at 32 ("[T]he Division will perform a 'fitness' test to ensure that the purchaser has sufficient acumen, experience, and financial capability to compete effectively in the market over the long term.").

57. Although a firm's long-term financial viability is relevant to whether it is an acceptable divestiture buyer, the fact that a firm is "losing money" does not necessarily suggest that it is not financially viable nor that it could not be an effective competitor. Indeed, AT&T – [REDACTED TEXT] – lost money in its final fiscal year as an independent firm, far more than [REDACTED TEXT]. See AT&T Corp. SEC Form 10-K Filing at 25 (Mar. 10, 2005) (reporting an operating loss of approximately $10 billion in 2004).


59. Economides NYAG Decl. at 33 (sub-head E); see also Selwyn NASUCA Decl. ¶ 37 (divestiture buyer will be "no match for the two mega-RBOCs").

60. Majure Reply Decl. ¶ 19.

61. Economides NYAG Decl. ¶ 66.

62. A divestiture of "lit" fiber – fiber currently in use by a customer – could potentially raise unnecessary complications, including the possibility of outages or other harm to the end-users in question. See, e.g., Majure Decl. ¶¶ 18, 21.

63. Economides NYAG Decl. ¶ 67.

64. See Proposed Final Judgments §§ IV(A), (H)(2).

65. Of course, the divestiture buyer also has to agree to the IRU terms, including those involving maintenance, and it is unlikely that the sophisticated carriers buying these assets would agree to terms that did not provide adequate guarantees of maintenance. Indeed, as previously noted, carriers in the industry routinely use IRUs to add fiber to their networks and negotiate terms that, in their business judgment, adequately protect them. Majure Decl. ¶ 22.

66. See, e.g., Majure Decl., Attachs. Tab 16, Divestiture Purchase Agreements


The SBC agreements have been submitted to the Court because the terms have been conditionally approved by the United States. The Verizon agreements have not been because they were more recently submitted to the United States and the terms have not yet been approved.

67. See, e.g., id. [REDACTED TEXT]

68. Selwyn NASUCA Decl. ¶¶ 53-54.

69. Majure Reply Decl. ¶ 21.

70. Id. ¶ 20.

71. See id. ¶¶ 20-21.

72. Clayton Act § 15, 15 U.S.C. §§ 4, 25.

73. 47 U.S.C. § 310(d); see also 47 U.S.C. § 214(a).

74. For example, under New York law, the New York Public Service Commission must determine whether a merger between telephone companies "is in the public interest." N.Y. Pub. Serv. Law § 100 (Consol. 2006). In fact, NYAG participated in the New York Public Service Commission proceedings unsuccessfully raising many of the arguments that it brings here in objecting to the Verizon/MCI merger.

75. These agencies reached conclusions about the mergers that are consistent with those reached by the United States. FCC Merger Orders ¶ 2; Order Approving Merger, In re: Joint Petition of SBC Communications Inc., AT&T Corp., N.Y. Public Service Commission CASE 05-C-0242, at 11 (Sept. 21, 2005) ("NYPSC SBC/AT&T Order"); Order Asserting Jurisdiction and Approving Merger Subject to Conditions, In re: Joint Petition of Verizon Communications Inc. and MCI, Inc. for a Declaratory Ruling Disclaiming Jurisdiction Over or in the Alternative for Approval of Agreement and Plan of Merger, N.Y. Public Service Commission CASE 05-C-0237, at 60-64 (Nov. 22, 2005) ("NYPSC Verizon/MCI Order"); Order of Approval, In re: Joint Petition of Verizon Communications Inc. and MCI, Inc. for Approval of Merger, N.J. Board of Public Utilities Docket No. TMO5030189, at 49-50 (Apr.12, 2006) ("NJBPU Verizon/MCI Order"); Order, In re: Joint Petition of SBC Communications Inc. and AT&T Corp., Together with its Certificated Subsidiaries for Approval of Merger, N.J. Board of Public Utilities Docket No. TMO5030168, at 23 (Oct. 4, 2005) ("NJBPU SBC/AT&T Order").

76. Some amici go so far as to ask the Court to use its authority to deal with issues not before it and that are inappropriate for a merger remedy. See, e.g., NJRC Reply at 22-23 (asking Court to "direct the parties to consider . . . having Verizon and SBC compete out of region" for mass market customers); Sprint Resp. at 33 (suggesting that "[a] holding by this Court that the proposed consent decrees are inadequate may cause the federal government to take the steps necessary to lower the merging parties['][special access] prices to just and reasonable levels").

77. 15 U.S.C. § 16(e) (emphasis added).

78. See United States v. Microsoft Corp., 56 F.3d 1448, 1459 (D.C. Cir. 1995) (statute does not authorize the Court to "construct [its] own hypothetical case and then evaluate the decree against that case"); United States v. BNS, 858 F.2d 456, 462-63 (9th Cir. 1998) (holding that a district court may not base its public interest determination on antitrust concerns in markets other than those alleged in the government's complaint); United States v. Archer-Daniels-Midland Co., 272 F. Supp. 2d 1, 6 (D.D.C. 2003) (noting that a court is not to "review allegations and issues that were not contained in the government's complaint").

79. On August 7, 2006 the United States provided the Court with materials relating to the harm identified and the remedy crafted to address this harm. See United States' Submission Attachs. The materials include data and documents produced by the merging parties, competing telecommunications carriers (including members of ACTel and COMPTEL), and other third parties. These materials demonstrate the reasoning behind the claims brought and the remedy and clearly support a finding that the entry of the proposed decrees is in the public interest.

80. See Motion Hr'g Tr. at 14:5-14:13 (July 12, 2006):

MR. SCOTT: The case law, which we believe is still good, indicates that going beyond the complaint itself is not appropriate for the context of these proceedings as it will impinge upon our role as the prosecutors and it will result in the Court looking at issues that we have not brought before you.

COURT: I agree with you. Let me make clear for the record. It's clearly my job to scrutinize the only complaint that's pending before the Court.

81. United States v. Bechtel Corp., 648 F.2d 660, 666 (9th Cir. 1981) ("The balancing of competing social and political interests affected by a proposed antitrust consent decree must be left, in the first instance, to the discretion of the Attorney General. The court's role in protecting the public interest is one of insuring that the government has not breached its duty to the public in consenting to the decree. The court is required to determine not whether a particular decree is the one that will best serve society, but whether the settlement is 'within the reaches of the public interest.'" (citations omitted)); cf. Heckler v. Chaney, 470 U.S. 821, 831 (1985) ("[A]n agency's decision not to prosecute or enforce, whether through civil or criminal process, is a decision generally committed to an agency's absolute discretion. This recognition of the existence of discretion is attributable in no small part to the general unsuitability for judicial review of agency decisions to refuse enforcement." (citations omitted)); Wayte v. United States, 470 U.S. 598, 607 (1983) (stating that prosecutor's discretion about what charges to file reflects recognition that decision to prosecute is ill-suited to judicial review); United States v. Nixon, 418 U.S. 683, 693 (1974) (stating that the executive branch "has exclusive authority and absolute discretion to decide whether to prosecute a case").

82. To justify arguments related to harm not alleged in the Complaints, amici are reduced to arguing that they know the scope of the United States' Complaints better than the United States does. The arguments made by amici, however, rely on snippets of the Complaints taken out of context. Read in their entirety, the Complaints allege harm only in the buildings identified in the proposed Final Judgments. See, e.g., Complaints ¶¶ 3, 25. Any suggestion that this is a post hoc argument developed to narrow the scope of these proceedings is demonstrably false; a contemporaneous press release issued by the United States describes the Complaints as follows:

According to the complaint[s] against Verizon [and SBC], [the merging firms] are the only two firms that own or control a direct wireline connection to hundreds of buildings in the metropolitan areas of Washington–Baltimore; Boston; New York; Philadelphia; Tampa; Richmond, Virginia; Providence, Rhode Island; and Portland, Maine [and Chicago; Dallas–Fort Worth; Detroit; Hartford–New Haven, Connecticut; Indianapolis; Kansas City; Los Angeles; Milwaukee; San Diego; San Francisco–San Jose; and St. Louis]. Therefore, in the absence of new entry, the merger would eliminate competition for facilities-based local private line service to those buildings.

Press Release, October 27, 2005, available at http://www.usdoj.gov/atr/public/press_releases/2005/212407.htm.

83. Sprint Resp. at 20.

84. ACTel Resp. at 29.

85. Network effects refers to the idea that the more locations or points a supplier has on a network, the more valuable that network is to customers. See, e.g., Selwyn NASUCA Decl. ¶ 18.

86. Id. ¶¶ 18-19, 28, 36; Economides NYAG Decl. ¶¶ 45-46, 52-53, 78, 80, 86.

87. Majure Decl., Attachs. Tab 3, AT&T Building List; Tab 5, MCI Building List.

88. Compare Sprint Resp. at 13 n.14 (citing estimate by Dr. Selwyn that pre-merger AT&T and MCI had approximately 6,400 and 6,000 on-net buildings respectively) with Time Warner Telecom, Inc. SEC Form 10-Q Filing at 26 (Aug. 9, 2006) (reporting that as of June 30, 2006 TWT had 6,433 on-net buildings).

89. AT&T was the largest CLEC in only [REDACTED TEXT] of the eleven SBC divestiture cities. MCI was the largest CLEC in only [REDACTED TEXT] Verizon divestiture cities. See Majure Decl., Attachs. Tab 3, AT&T Buildings List; Tab 5, MCI Building List; Tab 6, CLEC Network Maps and Building Lists. Dr. Selwyn even seems to acknowledge this when he notes that "pre-merger AT&T and MCI each possessed what were among the largest CLEC facilities-based fiber optic networks in terms of the quantity of nodes (buildings) being served." Selwyn NASUCA Decl. ¶ 19 (emphasis added).

90. Majure Decl., Attachs. Tab 3, AT&T Buildings List; Tab 5, MCI Buildings List; Tab 6, CLEC Network Maps and Buildings Lists.

91. Notwithstanding Sprint's argument that AT&T and MCI are "uniquely positioned," by its own admission there are dozens of other competitors for LPL. Kassien Sprint Decl. ¶ 14. In Verizon's region, for example, nearly [REDACTED TEXT] of the LPL that Sprint purchased from CLECs were from a competitor other than MCI. Id. ¶ 16. In SBC's region, almost [REDACTED TEXT] of the LPL that Sprint acquired from CLECs were purchased from someone other than AT&T. Id.

92. Enterprise customers sometimes buy telecommunications services from companies that own no telecommunications facilities. These companies, known as aggregators, purchase facilities from carriers and combine them with other services to provide and manage for example a data network. A number of the customer statements identify aggregators as competitors. See Majure Decl., Attachs. Tab 1, Retail Customer Statements.

93. See, e.g., Economides NYAG Decl. ¶¶ 50-51 (claiming that AT&T and MCI "resold surplus ILEC private lines at prices well below those that smaller CLECs could obtain from the ILECs directly"); Selwyn NASUCA Decl. ¶¶ 12-16 (suggesting that DOJ suffers from serious misconception, i.e., that the only source of competition for LPL comes from competing facilities-based providers). Dr. Selwyn concedes that this concern is beyond the scope of the Complaints. Selwyn NASUCA Decl. ¶ 16 (stating that "the Department's Complaints ignore this important source of Local Private Line competition in its entirety").

94. "Type II circuits" typically refer to circuits of which a part is resold from the RBOCs. Often, it refers to a circuit where the competitive carrier (like AT&T or MCI) provides the transport portion, and the RBOC provides the "access" portion – the portion between the customer and the nearest RBOC central office. These are distinguished from "Type I circuits," which are circuits provided entirely over the seller's own facilities.

95. United States' Response to Public Comments at 45-46; Majure Reply Decl. ¶ 31; see also FCC Merger Orders ¶¶ 41-47.

96. AT&T's and MCI's "Type II" sales amounted to a small minority of their overall LPL sales, which in turn were relatively small to begin with. Declaration of Anthony Fea, Anthony Giovannucci, Bob Handal, Michael Lesher and C. Michael Pfau (May 9, 2005), ¶ 43 (attached to Joint Opposition of SBC Communications Inc. and AT&T Corp. to Petitions to Deny and Reply to Comments, In re: SBC Communications Inc. and AT&T Corp. Applications for Transfer of Control, WC Docket No. 05-65 (May 10, 2005) ("AT&T currently provides less than [REDACTED TEXT] a year in wholesale local private line services in SBC's territory. Of that amount, only [REDACTED TEXT] is associated with Type II services for which AT&T leases a portion of the circuit from SBC.")), United States' Reply Submission Attach.1; Declaration of Jonathan P. Powell, Peter H. Reynolds, and Edwin A. Fleming (May 20, 2005), ¶ 11 (attached to Joint Opposition of Verizon Communications Inc. and MCI, Inc. to Petitions to Deny and Reply to Comments, In re: Verizon Communications Inc. and MCI, Inc. Applications for Transfer of Control, WC Docket No. 05-75 (May 24, 2005) ("Notably, more than [REDACTED TEXT] percent of MCI's wholesale Metro Private Line revenue is derived from circuits that are entirely on-net and do not use incumbent LEC special access at all, i.e., Type I circuits.")), United States' Reply Submission Attach. 2.


97. Selwyn NASUCA Decl., Attach. 2., Declaration of Alan G. Benway, Robert G. Holleron, Jeffrey King, Michael C. Lesher, Michael C. Mullan and Maureen Swift on Behalf of AT&T Corp., In re: Unbundled Access to Network Elements, WC Docket No. 04-313, Review of the Section 251 Unbundling Obligations of Incumbent Local Exchange Carriers, CC Docket No. 01-338 (Oct. 1, 2004), ¶ 101.

98. Majure Decl. ¶ 10.

99. Majure Reply Decl. ¶ 31.

100. A number of amici point to the fact that AT&T purchases large quantities of special access from the RBOCs and serves the vast majority of its customer locations via such special access as evidence that the loss of AT&T as a reseller will adversely impact competition. See, e.g., Selwyn NASUCA Decl. ¶ 13 (claiming that of 186,000 buildings where AT&T was providing retail DS-1 or higher services, only 5.7% was provided over its own, or CLEC, LPL); see also Economides NYAG Decl. ¶ 51. But this misses the point. These sales are not sales of "Type II" LPL to other carriers. These sales are principally sales of inputs to telecommunications services ultimately sold to enterprise customers. The United States did not find evidence of an anticompetitive problem related to the sale of "Type II" LPL, nor did it find evidence sufficient to suggest that either merger would significantly reduce competition to large enterprise customers that rely on LPL as an input – except in the rare situation where a substantial portion of the customer's needs was in an AT&T (or MCI) on-net building and there was no other CLEC with a connection to that building or likely to build in.

101. See ACTel Resp. at 26 (AT&T and MCI were in many cases low price leaders); [REDACTED TEXT].


103. See Reply of the United States to ACTel's Opposition to the United States' Motion for Entry of Final Judgments at 15-18 (June 1, 2006). The significant problems with those materials are discussed in more detail in Dr. Majure's Reply Declaration and accompanying attachment. See Majure Reply Decl. ¶ 30.

104. Majure Reply Decl. ¶ 28.

105. In an SEC filing, MCI reported that its 2004 wholesale revenue fell by $0.9 billion, in part due to the "elimination of certain incentive discounts" that it had offered in 2003. MCI, Inc., SEC Form 10-K Filing at 52 (Aug. 30, 2005). Similarly, before the merger, MCI predicted the continued decline in its wholesale revenues as it continued to "strive to achieve higher margins in this business." MCI, Inc., SEC Form 10-Q Filing at 29 (May 9, 2005).

106. Merger Guidelines § 2.0; see also Majure Reply Decl. ¶ 25.

107. ACTel has attempted to imply that prices have increased since the mergers, but the evidence it conjures up is all smoke and mirrors. The pricing tariffs it previously submitted show a price increase for no more than a handful of services in a handful of states in SBC's territory. There is no indication that the increases were merger-related or that they are symptomatic of "dramatic" across-the-board price increases as ACTel claims. See Majure Reply Decl. ¶ 29 n.44; see also AT&T, Inc.'s Memorandum in Response to ACTel's Reply Memorandum in Opposition to the Department of Justice's Motion for Entry of Final Judgments at 1-4 (June 27, 2006) (pointing to similar pricing tariff increases for the period before the consummation of the merger and thereby suggesting that these sorts of increases are not at all atypical, and have nothing to do with the mergers). ACTel's citations to documents and press statements are equally misleading and taken out of context. ACTel Resp. at [REDACTED TEXT] , 31-32.

108. See supra Part 2.b.i.

109. Majure Reply Decl. ¶ 24.

110. Merger Guidelines § 2.22 (stating that unilateral effects would be unlikely in commodity market unless competitors "of the merged firm likely would not respond to the price increase and output reduction by the merged firm with increases in their own outputs sufficient in the aggregate to make the unilateral action of the merged firm unprofitable").

111. Sprint Resp. at 7 ("Most of the cost of deploying access facilities lies in the supporting structures, placement, rights of way, and access to buildings, and not in the fibert strand or copper wires themselves."); see also Economides NYAG Decl. ¶ 45 n.34 (noting that because the incremental cost of installing additional capacity is small CLECs typically install ample capacity to supply customers that might be won over later ).

112. Merger Guidelines § 2.11.

113. Speculation by amici that the two merged firms will not compete in each other's region in the future is similarly unreasonable. After acquiring AT&T or MCI, each merged firm will own substantial facilities in the other's region. Amici have not suggested why the merged firms would not have every incentive to utilize those facilities to compete aggressively, particularly for enterprise customers whose office locations may extend beyond a single RBOC's region. See Majure Reply Decl. ¶ 32.

114. Id. ¶ 34. COMPTEL criticizes the United States' purported reliance on the customer statements submitted to the Court earlier in this proceeding. COMPTEL Resp. at 27. As Dr. Majure has explained, although these statements by themselves may be of limited evidentiary value they are consistent with the interviews that the United States conducted, and relied on in part, in making its decision not to pursue allegations of competitive harm to large enterprise customers. Majure Reply Decl. ¶ 34.

115. The statements previously submitted by the United States are consistent on this point. Of the 129 retail customer statements previously submitted by the United States, at least 33 specifically characterize the merging parties as primarily offering complementary, rather than competitive retail products. These customers included all sizes of customer, from the largest Fortune 50 customers ( [REDACTED TEXT] ) to small customers such as [REDACTED TEXT] , which spends less than [REDACTED TEXT] on telecommunications services annually. See Majure Decl., Attachs. Tab 1, Retail Customer Statements.

116. SBC Communications, Inc., Description of the Transaction, Public Interest Showing, and Related Demonstrations at 67, 96-101, In re SBC Communications Inc. and AT&T Corp. Applications for Transfer of Control, WC Docket No. 05-65 (Feb. 22, 2005); Verizon Communications, Inc., Public Interest Statement at 25-26, In re Verizon Communications Inc. And MCI, Inc. Applications for Transfer of Control, WC Docket No. 05-75 (Mar. 11, 2005).

117. NYAG points to a customer survey to suggest that customers are concerned about the merger. However, the survey, which was prepared at the request of opponents to the transaction, is unpersuasive for reasons explained by Dr. Majure. See Majure Reply Decl. ¶ 34 & n.57.

118. See NJRC Reply at 2 (remedies fail to protect mass-market customers who want plain old telephone service); NASUCA Reply at 7 (asserting that the Complaints take an unreasonably narrow view of the merger's impact, which will have broad impacts on residential and other customers); see also Economides NYAG Decl. ¶ 17 (claiming that DOJ has not identified all relevant antitrust markets, including wireless, cable packet voice and long distance).

119. See NJRC Reply at 3-5 (discussing impact of FCC TRRO on local switched service).

120. See, e.g., Motion of the New Jersey Division of the Ratepayer Advocate to Intervene and Memorandum of Points at 7 (July 18, 2006) (asking the Court to force Verizon and AT&T to sell switched residential access facilities to competitors at "TELRIC" rates).

121. Majure Reply Decl. ¶ 33.

122. See, e.g., NJBPU SBC/AT&T Order at 9 (finding no harm to competition from the AT&T/SBC merger based on the "inescapable [and] firmly established" fact that "AT&T has ceased attempting to compete with other CLEGs in New Jersey for mass market customers"); NJBPU Verizon/MCI Order at 36 (finding "MCI has essentially ceased to compete vigorously [in mass markets]"); NYPSC Verizon/MCI Order at 29 ("We find that MCI, already deemphasizing its presence in the mass market, would no longer be in a position to exert significant influence over this market in the absence of this merger.").

123. See United States v. Pearson plc, 55 F. Supp. 2d 43, 45 (D.D.C. 1999) (citing BNS, 858 F.2d at 462-63).


1. In addition, AT&T has deployed a limited number "rifle shot" collocations, which are also not associated with AT&T's metro fiber deployment but instead are used in connection with AT&T's long distance services and its "digital link" service. When providing its digital link service, AT&T uses long distance 4ESS switches to provide a limited functionality local voice service to business customers that also use AT&T long distance service.

2. The metropolitan areas are Austin, Chicago, Cleveland, Columbus, Dallas, Detroit, Dayton, Hartford, Houston, Indianapolis, Kansas City, Los Angeles, Milwaukee, Reno, St. Louis, Sacramento, San Antonio, San Diego, and San Francisco. AT&T has also deployed local facilities in Cincinnati, but that area is outside of SBC's incumbent territories in Ohio.

3. In addition, AT&T has [REDACTED] so-called "rifle shot" collocations, which as noted, are not associated with AT&T's metro fiber deployment but instead are used in connection with AT&T's long distance services and its "digital link" service.

4. This calculation, and related calculations below, were conducted by comparing the locations of AT&T's facilities-based collocations with the list of "Tier 1" and "Tier 2" wire centers provided to the Commission by SBC.

5. For the remaining minority of locations - which represent about [REDACTED] of SBC's switch locations - there still may be multiple collocators.

6. Such locations are hereinafter referred to as "commercial buildings." AT&T's network also connects to non-commercial buildings, such as incumbent LSOs, for purposes of interconnecting with other networks. Such "network locations" typically do not house retail or wholesale customers. Thus, they are not locations for which competitive carriers would seek to purchase local private line service from AT&T.

7. In this regard, we are not aware of any building served by AT&T that is not also served by SBC.

8. AT&T's data includes more than "lit" buildings. AT&T identifies a building with flags noting whether the building is currently lit by a competitive carrier, could be lit by deploying terminal equipment, or is a building that a competitive carrier would be willing to put on net under appropriate conditions. Unless otherwise specified, we refer only to currently lit buildings and, as such, are thus taking a conservative view of a list that already understates competitive deployment. AT&T also does not know whether any of the buildings listed are pure non commercial locations where AT&T (or any other retail provider) is unlikely to have customers.AT&T believes most of the building locations are locations with at least some commercial customers, for it would be pointless for competitive carriers to provide AT&T with "on net" buildings where AT&T (or any other carrier) would never expect to have a customer.

9. AT&T estimates that these [REDACTED] connections serve approximately [REDACTED] unique buildings.

10. Although a single DS3 of demand is not always sufficient to support bypass deployment, this threshold is a useful proxy for considering whether competitive deployment is possible in light of the Commission's nationwide non-impairment findings in the TR Remand Order fl[ 117), which limited competitive carriers to one DS3 loop per location and the Commission's further decision (id. ][ 146) to eliminate access to DS3 (and DSl) loops altogether in high demand areas where many AT&T buildings are located. The Commission also limited requesting carriers to 10 DSls per location. Id. ^f 128.


12. Such network locations independently justify the deployment of fiber by AT&T. Other competitive carriers likewise have the same ability to deploy fiber to serve "low" capacity customers that are located in buildings that house their network locations.

13. In those circumstances, some of the individual buildings might have less than 2DS3 equivalents of demand, but because of their proximity to other served buildings and the unique opportunity to use another commercial location as a network hub to aggregate traffic onto a fiber connection service becomes economically feasible. Again, because of the aggregate revenue opportunity presented in such circumstances, other competitive carriers would have the same economic ability to self deploy such facilities.

14. Because fixed wireless "loops" do not require the same investment as wireline fiber loops, AT&T is able economically to provide lower capacity services using fixed wireless connections than with fiber loops. However, AT&T's experience is that about [REDACTED] of commercial buildings can be served with fixed wireless loops because of "line of sight" limitations on the technology, inability to get access to building rooftops, and gaps in AT&T's spectrum ownership.

15. These carriers are ACN, ATX,, Broadwing, Bullseye, Cavalier, Cbeyond, Cimco, Conversent, Cox, CTS, Eschelon, Gillette, Global Crossing, Granite, Lightship, Lightyear, NuVox, Pac-West, RCN, Sawis, TDS, Tele-Pacific, US LEC, Xspedius, and XO.

16. These figures are based on AT&T's 2005 local private line sales.

17. AT&T's other competitive special access suppliers in the SBC region provided AT&T with less than [REDACTED] in special access service in SBC's region. Providers not among the top ten accounted for less than [REDACTED] of AT&T's in-region purchases from competitive carriers.


1. This figure includes both fiber-served buildings and a limited number of buildings served over copper facilities. It also includes some buildings that have MCI fiber but are not currently active i.e., have no transmission electronics.

2. The only Type IV circuits currently provided by MCI consist of a handful of "grandfathered" Type IV circuits.

3. http://www.cavtel.com/wholesale/index.shtml

4. The "day one" circuit count does not necessarily indicate the ultimate traffic level in a building. The two buildings with day one circuit counts below the DS3 level were approved because the revenue guaranteed by the customer justified the cost of construction. It is likely that, in order to meet their revenue commitments, the customers in these buildings would have to subsequently increase their circuit demand above the day one level.

5. This analysis includes both Verizon-East and Verizon-West buildings. MCI on-net buildings that are Verizon wire centers were excluded from this analysis.

6. Current demand below the DS3 level typically signifies that a customer whose traffic justified the facilities construction has relocated or switched to a different carrier. Included in those buildings with current demand below the DS3 level are buildings that have MCI fiber but are not currently active, i.e., MCI has no customers in the building and has removed the transmission electronics.

Updated June 30, 2015