Declaration of W. Robert Majure

DECLARATION OF W. ROBERT MAJURE

1. My name is W. Robert Majure. I am an economist employed by the Antitrust Division of the United States Department of Justice. During my tenure at the Antitrust Division, I have analyzed numerous mergers in the telecommunications industry and in other industries. I have also been involved in settling concerns in many of these matters by consent decree. Examples of the mergers I have reviewed are WorldCom/Sprint, Cingular/AT&T Wireless, and SBC/Ameritech. In addition, I have been involved in analyzing and helping to develop the Division's position on policy issues relating to the telecommunications industry, and have assisted in preparing comments on telecommunications-related issues that were filed with the FCC. I currently serve as Chief of the Division's Competition Policy Section, a group of economists who are routinely assigned to work with Division lawyers on investigations. A copy of my curriculum vitae is attached.
2. I supervised the economists who worked with staff attorneys on the Division's investigations of SBC's proposed merger with AT&T and Verizon's proposed merger with MCI. I was involved in all aspects of their work including the development of potential economic theories of harm and the analysis of whether these theories were supported by the documents and data produced by the merging parties and competitive local exchange carriers ("CLECs"), as well as by information gathered from industry and public sources. I also supervised their work in evaluating the proposed remedies, and in recommending that the Department accept the divestitures contained in the proposed Final Judgments as solutions to the harms that have been identified in the Complaints.
3. I have been asked to prepare this declaration in order to help explain how the proposed remedies address the harm identified by the Department and described in the Complaints. Below, I discuss local private line services and the companies that provide such service, and describe in particular the extent to which these companies have competing facilities. Next, I describe the anticompetitive harm from the mergers as discussed in the Complaints, which focused on specific buildings where the merging firms had the only facilities that can be used to provide local private line services. Finally, I address why the remedies are sufficient and appropriate, including how they provide an appropriate buyer with all of the assets it will need to step into the shoes of AT&T or MCI and replace the competition in these services lost through the mergers. I also describe how the remedies are crafted to limit disruption to customers' existing telecommunications services. Throughout this declaration, I reference examples of the types of materials that support the Department's decisions to bring the present cases and to adopt the remedies.
4. The focus of antitrust analysis of mergers is on predicting when a proposed merger will likely lead to a loss of competition. Two large companies in the same industry do not necessarily compete vigorously with one another. For example, they may only rarely sell comparable products to the same customers. As an economist trying to determine whether a proposed merger should be challenged, I look at where and to what extent the companies compete, who else competes, how competition occurs, and who else is likely to enter if conditions change. The goal is to understand how competition, not competitors, will be affected. So the focus is on the overall mechanics of competition and the particular ways in which each of the merging parties influences the other's competitive activities, such as pricing. In the present cases, the question is whether the loss of either AT&T or MCI is likely to affect competition for any particular product in a particular geographic area. More specifically, I assess whether there is evidence supporting a valid economic theory that indicates that the merger is likely to harm customers.⁽¹⁾
5. The questions posed in an antitrust analysis are not ones that companies study in the normal course of business. The hypothetical question of how firms in the industry would alter their behavior after a merger, a question which is essential to antitrust analysis, is far removed from the usual things with which businesses are concerned (e.g., how to produce the product most efficiently or where to sell a product for the most profit). So it is rare for internal business documents to provide complete answers. Rather, in reviewing a merger, I build a picture of the industry and competitive conditions by assembling relevant information from numerous sources and applying pertinent economic theories. This declaration references the types of materials collected and indicates the relevant facts suggested by such materials.

The Products and Providers

6. SBC and Verizon initially were only local phone companies. They were formed as a result of the 1984 Modification of Final Judgement ("MFJ") that was entered in the antitrust case the Department brought against the Bell System. The MFJ separated the old Bell System into a number of local telephone companies, which were owned by regional bell operating companies ("RBOCs"), such as SBC and Verizon, and one long distance company, AT&T. Each of the local companies created by the MFJ had a distinct service area wherein it was effectively the monopoly provider of local phone service. At the same time, AT&T was positioned to compete against new entrants, primarily MCI, for the provision of long distance service.
7. Since 1984, the industry has seen dramatic changes including the entry of additional competitors and the construction of additional network facilities to provide both local and long distance service. Of particular note, the Telecommunications Act of 1996 altered the regulatory regime to encourage competition in long distance and local markets. This precipitated a period of aggressive investment in telecommunications facilities as many firms ­ including AT&T and MCI ­ spent considerable sums expanding their network infrastructure to offer local services. As a result there are multiple local networks in most major U.S. cities operated by these competitive local exchange carriers ("CLECs"), including MCI and AT&T.
8. At the time its merger with SBC was proposed, AT&T owned local area networks insome cities where SBC was the RBOC. As the incumbent, SBC provides ubiquitous service ­ its network in these cities is extensive and capable of providing service essentially anywhere. By contrast, AT&T's network was focused on providing service where it is most profitable to do so ­ in areas where the density of high-revenue business customers justifies the construction costs. As is typical of CLEC networks, AT&T local area networks have two components, transport and laterals. Transport networks are high-capacity fiber⁽²⁾ deployed to provide the capability to move traffic around a city.⁽³⁾ Laterals are the connections from this fiber network into individual buildings. Most of the buildings along a fiber transport network will not have a lateral connection to the network (referred to as "off-net"), but many others will ("on-net").⁽⁴⁾ Wherever AT&T offered facilities-based service it can be assumed that SBC offered a competing service.
9. Likewise, at the time their merger was proposed, MCI owned local area networks (including both transport⁽⁵⁾ and laterals into buildings⁽⁶⁾) in some cities where Verizon was the incumbent local service provider. The MCI networks were built, similar to AT&T's, primarily in central business districts, but as to all these locations it can be assumed that MCI was competing with Verizon.
10. In addition, CLECs other than AT&T and MCI have also built local area networks in the same cities where either AT&T or MCI owned local facilities.⁽⁷⁾ In fact, multiple CLECs do business in each of the major cities within SBC's and Verizon's service territory.⁽⁸⁾ These CLEC's include cable companies (such as Cox Communications and Cablevision), traditional long distance carriers (Sprint), and a number of companies whose primary business is providing either long distance and/or local facilities (such as AboveNet, Level 3, XO Communications, and Telcove). CLECs' networks tend to cover the same high-density areas covered by AT&T and MCI, and they therefore overlap substantially with the transport networks of AT&T and MCI.⁽⁹⁾ In addition, CLECs (sometimes multiple CLECs) have laterals into the same buildings as AT&T or MCI.⁽¹⁰⁾ Some, though not all, CLECs continue to build both transport and laterals into individual buildings.⁽¹¹⁾ At the time the mergers were proposed, for example, [REDACTED TEXT] was adding new buildings at a faster rate than either MCI or AT&T.⁽¹²⁾
11. These local area networks are used to provide business customers with a common fundamental building block of telecommunications services ­ local private lines. These are dedicated circuits (meaning that the customer has the exclusive right to use the transmission path) connecting two points in the same metropolitan area. The customer for such a circuit might be a business seeking to connect computers in two office buildings, or another carrier seeking to sell services to business customers. Access into a building through owned or leased facilities is necessary to enable a carrier to serve customers in that building. Virtually any telecommunications service one intends to use needs a local connection to the buildings where the service originates or terminates, and, except where the expected volume of traffic on the connection is low, a business will use local private lines to make that connection. In each of the cities identified in the Complaints, local private lines are provided by the merging parties and by other CLECs.⁽¹³⁾

The Harm Alleged

12. The existence of overlapping assets that can be used to provide competing products and services (e.g., those described above) often suggest areas of concern in the Department's merger investigations. But the fact that two companies have overlapping assets does not, in and of itself, prove that customers would experience higher prices as a result of a merger of those companies. To determine whether a merger that would result in common ownership of overlapping assets would rise to the level of an antitrust violation, the Department attempts to evaluate all relevant facts about the market to determine how firms compete. Based on its understanding of the facts and the nature of competition, the Department then attempts to predict the effect of the merger on competition.
13. As set forth in the Complaints, the Department identified harm in the market for the sale of local private lines. This harm is predicted in situations where only AT&T and SBC or MCI and Verizon, respectively, were capable of supplying local private lines before the merger and no other CLEC was likely to connect the building to its network. After the merger, SBC or Verizon would be the only possible supplier of local private lines to those buildings, and they could raise prices without fear of competition. In practice, the fact that the merged firms would no longer face competition from a CLEC in these buildings is likely to result in higher prices or lower quality (e.g., less responsiveness to service outages or requests to provide new circuits) for local private lines, or for packages of telecommunications services that include local private lines into the affected buildings, all to the detriment of consumers.
14. Of course, harm to consumers would be unlikely if the merged firms knew that raising prices, for example, would make it profitable for a CLEC to construct a lateral connection from its network into these buildings and thereby offer customers a choice. When entry such as that is likely, firms often recognize that fact and avoid giving the would-be entrant cause to enter. As a result, the buildings identified by the Department as likely to experience harm from the mergers do not include all situations where only AT&T and SBC or only MCI and Verizon were present before the mergers. Rather, they include only those 2-1 buildings where entry is unlikely.⁽¹⁴⁾ The likelihood of entry can be assessed by examining the criteria CLECs use in deciding whether to make such investments, i.e., whether the potential revenue to be earned will be sufficiently greater than the cost of building a lateral.⁽¹⁵⁾ This provides an understanding of the relationship between cost (primarily the distance of the building from the provider's network) and the revenue the CLEC was likely to receive from customers in the building.⁽¹⁶⁾ Estimates of the revenue opportunity (based on the current traffic being generated in the building adjusted for special circumstances) and the distance to the closest CLEC fiber provide bases for identifying the subset of 2-1 buildings for which long-term harm was not likely to be offset by entry.⁽¹⁷⁾ These more than 700buildings are listed in the Appendices to the proposed Final Judgments.

The Remedies

15. In fashioning a remedy here, the Department's goal was to restore the competition that the Complaints allege would likely be lost due to the mergers. In general, the Department aims to craft a remedy that has a "close logical nexus" to the alleged violation.⁽¹⁸⁾In addition, the proposed remedy should "effectively redress the violation and, just as importantly, be no more intrusive on market structure and conduct than necessary to cure the competitive harm."⁽¹⁹⁾
16. In these cases, the proposed remedies are straightforward. They require the divestiture of connections into the buildings identified as problematic in the Complaints. In each building, the buyer of the divested assets would step into the shoes of AT&T or MCI. As new sales opportunities arise in the buildings, the buyer will be positioned to offer an alternative to SBC or Verizon. All customers ­ the tenants in the building as well as the carriers who 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.
17. If AT&T or MCI had some unique qualifications as a competitor in selling local private lines, there might be concerns that no other CLEC would be as competitive as AT&T or MCI. However, I found no evidence suggesting a unique competitive role for either of these firms in selling local private lines. In fact, any supplier that can provide a technically reliable point-to-point connection is a competitive option for purchasers interested in accessing those two locations.⁽²⁰⁾
18. Replacing the lost competition often presents some practical difficulties that requires a balance of the competitive benefits from a divestiture with any burdens the divestitures could create for customers. For example, some commentators have suggested that the merging firms should be required to sell customer contracts along with the live circuits that serve those customers. These proposals, however, would create concerns that forcing customers to change carriers when they had not anticipated or planned for such a change would be disruptive and costly for the customers.⁽²¹⁾ The proposed remedies are a balance between these potential problems and the benefits of requiring divestiture of these assets. Because most business contracts are relatively short in duration (typically one to three years), many of the existing contracts will be expiring within the next two years, and, consequently, a buyer of the divestiture assets can compete directly or indirectly (through [REDACTED TEXT] leasing access to another CLEC) for this business as contracts come up for renewal. Under the proposed remedies, therefore, it is likely that a buyer will be successful without these assets, and it is better not to burden customers with the disruption associated with transferring the customer contracts and most of the facilities providing services under those contracts.²²
19. The specific terms of the remedies embodied in the proposed Final Judgments are crafted to address practical concerns such as this. The proposed remedies achieve a desirable balance between minimizing unintended negative consequences and giving a buyer of the assets the ability to replicate the competitive significance of AT&T or MCI in providing local private line services to the buildings of concern. Consequently, I believe the proposed remedies appropriately limit collateral harm while effectively addressing the harm in the Complaints.
20. The proposed decrees call for divestiture of the greater of half of the unused fiber or eight of the strands. At least eight strands have to be divested. I believe this amount of fiber will be sufficient to serve likely customer demand in the affected buildings today, as it is as much fiber as CLECs generally use in a building.²³ Going forward, the amount of bandwidth customers demand will likely continue to grow but so will the amount of bandwidth that advanced fiber-optic electronics can produce from a single strand. Advances in optronics continue to make this equipment cheaper and more efficient every year.²⁴ Therefore, even if there is an increase in demand in the building, upcoming advances in the state of the art for this equipment will likely make each fiber strand even more productive.
21. The proposed decrees do not require divestiture of wiring or electronics inside the buildings. The cost of wiring and electronics are relatively small compared to the other costs of connecting a building to one's network in order to connect a new customer. A CLEC's decision whether to enter a building is instead generally driven by the distance to the CLEC's network.²⁵ Putting this another way, once the buyer has purchased the divestiture assets relevant to a particular building, the remaining costs of serving a particular customer in that building are not likely to prevent entry if the customer is expected to produce any substantial amount of revenue. Therefore, not requiring that the buyer obtain wiring or other equipment inside a building at the point of the divestiture is not likely to lessen the ability of the buyer to compete. On the other hand, forcing the divestiture of the inside wiring and electronics that are currently being used to provide services to customers in the building could raise all of the unintended negative consequences for those customers discussed above.
22. Rather than a divestiture of full ownership of fiber strands, the proposed decrees use 10-year IRUs (indefeasible rights of use) effectively to divest a portion of a facility. It is common practice in the industry for multiple carriers to share in the ownership of a fiber cable through IRU arrangements (each carrier has access to a strand or multiple strands in the fiber cable). As I understand it, the IRU conveys essentially all ownership rights to the current IRU holder.²⁶ These types of agreements have been developed through ordinary commercial dealings between carriers to achieve a high level of control by the IRU holder. In fact, carriers routinely use IRUs to supplement their owned facilities. For example, AT&T uses IRUs to support a large part of its local service networks, including some of the buildings identified in the Complaints.²⁷
23. The 10-year terms for the IRUs are appropriate given the dynamic nature of the industry and the difficulties inherent in predicting the future in the long term. Generally, the Department has tried to limit decrees to 10 years²⁸ precisely because industries change dramatically over time, and a decree which started out in the public interest may end up doing more harm than good over the decades. In this industry, 10 years is a relatively long time. Customer contracts are typically one to three years, so the IRUs allow the buyer to compete through multiple contracting cycles. At the end of 10 years, the buyer may choose to replace the IRUs with its own facilities. Another possibility is that another CLEC will build overlapping facilities in the interim. The proposed decrees do not, of course, prohibit longer IRU terms. For example, some of the prospective buyers of the divested assets elected to negotiate extensions.²⁹
24. The proposed decrees provide that the buyer must be provided with transport from the customer's premises back to the transport network of the buyer. Clearly, a carrier that is not AT&T or MCI might have some difficulties using a connection that starts at the building but ends at the AT&T or MCI transport network. The concept in the proposed Final Judgments is to divest something that connects the identified buildings to the buyer's local area network. The decrees do not require divestiture of the entire fiber transport network of the merged entity that runs close to these buildings and terminates at the buyer's network, because this infrastructure is being used simultaneously to provide service to many more buildings. The problem of customer disruption and lost efficiencies associated with such a divestiture would be significantly greater than for the customer divestitures described above. Again, the proposed decrees solve this problem by granting the buyer some of the rights of ownership without transferring ownership of the asset in its entirety. The buyer gets the right to use the merged firms' transport fiber networks to move the traffic from the customer's premises to the buyer's network as well as the right to add other buildings.³⁰
25. In short, the proposed remedies address the alleged violations. In each building where lost competition between AT&T and SBC or between MCI and Verizon was identified as likely to cause a substantial harm, the proposed remedies introduce a new competitor that will have the assets necessary to replicate the competition AT&T or MCI was providing for local private line services before the mergers. At the same time, the proposed decrees minimize the disruptions to other services and other customers using the same facilities. This balance is struck by focusing the divestitures as much as possible on the assets that are critical to the competitiveness of the buyer, and by relying on the types of joint ownership arrangements that are common in the industry.