******************************************************** NOTICE ******************************************************** This document was converted from WordPerfect or Word to ASCII Text format. Content from the original version of the document such as headers, footers, footnotes, endnotes, graphics, and page numbers will not show up in this text version. All text attributes such as bold, italic, underlining, etc. from the original document will not show up in this text version. Features of the original document layout such as columns, tables, line and letter spacing, pagination, and margins will not be preserved in the text version. If you need the complete document, download the WordPerfect version or Adobe Acrobat version, if available. ***************************************************************** Before the Federal Communications Commission Washington, D.C. 20554 In the Matter of ) Implementation of the ) CC Docket No. 96-98 Local Competition Provisions ) of the Telecommunications Act of 1996 ) THIRD REPORT AND ORDER AND FOURTH FURTHER NOTICE OF PROPOSED RULEMAKING Adopted: September 15, 1999 Released: November 5, 1999 Comment Date: January 12, 2000 Reply Date: February 11, 2000 By the Commission: Commissioner Ness issuing a statement; Commissioner Furchtgott- Roth concurring in part, dissenting in part and issuing a statement; Commissioner Powell dissenting in part and issuing a statement. TABLE OF CONTENTS Paragraph I. INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . . . .1 II. EXECUTIVE SUMMARY . . . . . 16 III. BACKGROUND. . . . . . .17 IV. FRAMEWORK FOR DECIDING WHAT NETWORK ELEMENTS MUST BE UNBUNDLED PURSUANT TO SECTION 251. . . . . . . . . . . . . . . 22 A. Overview. . . . . . . . . . . . . . . . . . . . . . . . . 22 B. The "Necessary" and "Impair" Standards of Section 251(d)(2) . . . . . . .30 1. Application of the "necessary" and "impair" standards to proprietary and non- proprietary elements under Section 251(d)(2) . . . . . . . .30 a. Background . . . . . . 30 b. Discussion . . . . . . 32 2. Definition of "Proprietary in Nature". . . . . . . . . .33 a. Background . . . . . . 33 b. Discussion . . . . . . 35 3. The "Necessary" Standard of Section 251(d)(2)(A) . . . . . . 42 a. Background . . . . . . 42 b. Discussion . . . . . . 45 4. The "Impair" Standard of Section 251(d)(2)(b). . . . . .49 a. Background . . . . . . 49 b. Discussion . . . . . . 52 C. Adoption of a National List of Unbundled Network Elements . . . . .118 1. Background . . . . . . . . . . . . . . . . . . . . . . 118 2. Discussion . . . . . . . . . . . . . . . . . . . . . . 121 a. Legal Authority. . . . . . 122 b. Goals of the Act . . . . . 125 D. Modification of the National List . . . . . . . . . . . .145 1. Background . . . . . . . . . . . . . . . . . . . . . . 145 2. Discussion . . . . . . . . . . . . . . . . . . . . . . 149 a. Modification of the National List by the Commission. . . . . .149 b. Modification of the National List by the States. . . . . 154 V. APPLICATION OF THE STANDARD TO INDIVIDUAL NETWORK ELEMENTS. . . . . .163 A. Loops . . . . . . . . . . . . . . . . . . . . . . . . . .163 1. Background . . . . . . . . . . . . . . . . . . . . . . 163 2. Discussion . . . . . . . . . . . . . . . . . . . . . . 166 a. Definition . . . . . .167 b. Proprietary Concerns Associated with the Loop. . . . . . 181 c. Unbundling Analysis for the Loop in General. . . . . . . 182 B. The Subloop . . . . . . . . . . . . . . . . . . . . . . .203 1. Background . . . . . . . . . . . . . . . . . . . . . . 203 2. Discussion . . . . . . . . . . . . . . . . . . . . . . 206 a. Definition of the Subloop. . . . . . 207 b. Proprietary Concerns Associated with Subloops. . . . . . 209 c. Unbundling Analysis for Subloops . . . . .210 d. Technical Feasibility. . . . . .221 C. Network Interface Devices (NIDs). . . . . . . . . . . . .231 1. Background . . . . . . . . . . . . . . . . . . . . . . 231 2. Discussion . . . . . . . . . . . . . . . . . . . . . . 233 a. Definition of the NID. . . . . .234 b. Proprietary Concerns Associated with the NID . . . . . . 237 c. Unbundling Analysis . . . . . .238 D. Local Switching . . . . . . . . . . . . . . . . . . . . .242 1. Local Circuit Switching. . . . . . . . . . . . . . . . 242 a. Background . . . . . .242 b. Discussion . . . . . .245 2. Packet Switching . . . . . . . . . . . . . . . . . . . 301 a. Background . . . . . .301 b. Discussion . . . . . .303 E. Interoffice Transmission Facilities . . . . . . . . . . .319 1. Background . . . . . . . . . . . . . . . . . . . . . . 319 2. Discussion . . . . . . . . . . . . . . . . . . . . . . 322 a. Dedicated Transport. . . . . . .323 b. Shared Transport . . . . . 370 F. Signaling Networks and Call-Related Databases . . . . . .381 1. Signaling Networks . . . . . . . . . . . . . . . . . . 381 a. Background . . . . . .381 b. Discussion . . . . . .384 2. Call-Related Databases . . . . . . . . . . . . . . . . 401 a. Background . . . . . .401 b. Discussion . . . . . .403 G. Operations Support Systems. . . . . . . . . . . . . . . .422 1. Background . . . . . . . . . . . . . . . . . . . . . . 422 2. Discussion . . . . . . . . . . . . . . . . . . . . . . 425 a. Definition of OSS. . . . . 426 b. Proprietary Concerns Associated with OSS . . . . . .433 c. Unbundling Analysis for OSS. . . . . 434 H. Operator Services and Directory Assistance. . . . . . . .439 1. Background . . . . . . . . . . . . . . . . . . . . . . 439 2. Discussion . . . . . . . . . . . . . . . . . . . . . . 442 a. Definition of Operator Services and Directory Assistance . . . . . 444 b. Proprietary Concerns Associated with OS/DA . . . . .446 c. Unbundling Analysis. . . . . . .447 VI. MISCELLANEOUS ISSUES. . . . . . 466 A. Section 271-Related Issues. . . . . . . . . . . . . . . .466 1. Background . . . . . . . . . . . . . . . . . . . . . . 466 2. Discussion . . . . . . . . . . . . . . . . . . . . . . 469 B. Combinations of Unbundled Loops and Transport Network Elements. . . . . 475 1. Enhanced Extended Link . . . . . . . . . . . . . . . . 476 a. Background . . . . . .476 b. Discussion . . . . . .479 2. Use of unbundled network elements to provide exchange access services. . . . . .484 a. Background . . . . . .484 b. Discussion . . . . . .485 C. Nondiscrimination Obligations of Incumbent LECs . . . . .491 VII. FOURTH FURTHER NOTICE OF PROPOSED RULEMAKING. . . . . . .493 A. Background. . . . . . . . . . . . . . . . . . . . . . . .493 B. Discussion. . . . . . . . . . . . . . . . . . . . . . . .495 VIII. PROCEDURAL ISSUES . . . . .498 A. Final Regulatory Flexibility Analysis . . . . . . . . . .498 1. Need for, and Objectives of, the Third Report and Order. . . . . 499 2. Summary of Significant Issues Raised by the Public Comments in Response to the IRFA. . . . . . 500 3. Description and Estimate of the Number of Small Entities to which Rules will Apply . . . . . . 501 4. Description of Projected Reporting, Recordkeeping, and Other Compliance Requirements . . . . . . . . . . . . . . . . . . . . . . . 506 5. Steps Taken to Minimize the Economic Impact of this Order on Small Entities, and Alternatives Considered. . . . . . . . . . . . . . . . . . 507 6. Report to Congress . . . . . . . . . . . . . . . . . . 510 B. Initial Regulatory Flexibility Analysis (IRFA). . . . . .511 1. Need for, and Objectives of, the Proposed Rules. . . . 512 2. Legal Basis. . . . . . . . . . . . . . . . . . . . . . 514 3. Description and Estimate of the Number of Small Entities to Which the Proposed Rules Will Apply . . . . . . . . . . . . . . . . . . . . . . . . 515 4. Description of Projected Reporting, Recordkeeping, and Other Compliance Requirements . . . . . . . . . . . . . . . . . . . . . . . 518 5. Steps Taken to Minimize Significant Economic Impact on Small Entities, and Significant Alternatives Considered. . . . . . . . . . . . . . . . . . 519 6. Federal Rules that May Duplicate, Overlap, or Conflict with the Proposed Rules . . . . . .520 IX. ORDERING CLAUSES. . . . . .521 APPENDIX A -- LIST OF COMMENTERS Appendix B -- top 50 metropolitan statistical areas APPENDIX C -- final rules APPENDIX D -- current unbundled network elements I. INTRODUCTION 1. In this proceeding, we respond to the Supreme Court's January 1999 decision that directs us to reevaluate the unbundling obligations of section 251 of the Telecommunications Act of 1996 (1996 Act). The Supreme Court's decision removed many of the uncertainties surrounding the requirements of section 251 by upholding the majority of the Commission's rules implementing that section of the Act, including the Commission's jurisdiction to implement sections 251 and 252 of the Act, the Commission's definitions of network elements, and the Commission's rule requiring incumbent local exchange carriers (LECs) to offer combinations of unbundled network elements that are already combined. The Court has directed us, however, to revise the standards under which the unbundling obligations of section 251(c)(3) are determined. Specifically, the Court has required us to give some substance to the "necessary" and "impair" standards in section 251(d)(2), and to develop a limiting standard that is "rationally related to the goals of the Act." In addition, as we develop the "necessary" and "impair" standards, the Court has required us to consider the availability of alternative network elements outside the incumbent's network. 2. In passing the 1996 Act, Congress overhauled many aspects of federal regulation of telecommunications services by establishing a pro-competitive and deregulatory framework designed to benefit "all Americans by opening all telecommunications markets to competition." Two of the fundamental goals of the 1996 Act are to open the local exchange and exchange access markets to competition and to promote innovation and investment by all participants in the telecommunications marketplace. Congress sought to foster this competition by fundamentally changing the conditions and incentives for market entry and by attempting to open any remaining local service bottlenecks. As a result, the provisions of the 1996 Act set the stage for a new competitive paradigm in which carriers in previously segmented markets are able to compete in a dynamic and integrated telecommunications market that promises lower prices and more innovative services to consumers. 3. Central to the new statutory scheme is section 251 of the Act, which seeks generally to reduce inherent economic and operational advantages possessed by incumbent local exchange carriers. Toward this end, section 251 imposes specific market- opening mechanisms, such as mandatory interconnection, unbundling, and resale requirements on incumbent LECs, in order to break the incumbents' control over local facilities. Congress directed the Commission to implement the provisions of section 251, and to specifically determine which network elements should be unbundled pursuant to section 251(c)(3). 4. Pursuant to our statutory mandate and the directives of the Supreme Court, we reevaluate the unbundling obligations of incumbent LECs, pursuant to sections 251(c)(3) and 251(d)(2). The new standards and framework we adopt in this Order for determining which network elements incumbent LECs must make available on an unbundled basis will remove the uncertainties surrounding the incumbents' unbundling obligations since passage of the Act. More importantly, however, they will define the competitive landscape of telecommunications markets for the foreseeable future. 5. The standards and unbundling obligations that we adopt in this Order are designed to create incentives for both incumbent and competitive LECs to innovate and invest in technologies and services that will benefit consumers through increased choices of telecommunications services and lower prices. We recognize that there will be a continuing need for all three of the arrangements Congress set forth in section 251 to remain available to competitors so that they can serve different types of customers in different geographic areas. We continue to believe that the ability of requesting carriers to use unbundled network elements, including various combinations of unbundled network elements, is integral to achieving Congress' objective of promoting rapid competition to all consumers in the local telecommunications market. Moreover, in some areas, we believe that the greatest benefits may be achieved through facilities-based competition, and that the ability of requesting carriers to use unbundled network elements, including various combinations of unbundled network elements, is a necessary precondition to the subsequent deployment of self-provisioned network facilities. 6. Although Congress did not express explicitly a preference for one particular competitive arrangement, it recognized implicitly that the purchase of unbundled network elements would, at least in some situations, serve as a transitional arrangement until fledgling competitors could develop a customer base and complete the construction of their own networks. In particular, Congress stated: "[I]t is unlikely that competitors will have a fully redundant network in place when they initially offer local service because the investment necessary is so significant. Some facilities and capabilities . . . will likely need to be obtained from the incumbent [LEC] as network elements pursuant to new section 251." Implicit in this recognition, and in section 271's requirement that the Bell Operating Companies (BOCs) provide access and interconnection to their network facilities in accordance with the requirements in the competitive checklist, is Congress's expectation that new competitors would use unbundled elements from the incumbent LEC until it was practical and economically feasible to construct their own networks. 7. We fully expect that over time competitors will prefer to deploy their own facilities in markets where it is economically feasible to do so, because it is only through owning and operating their own facilities that competitors have control over the competitive and operational characteristics of their service, and have the incentive to invest and innovate in new technologies that will distinguish their services from those of the incumbent. Unbundling rules that encourage competitors to deploy their own facilities in the long run will provide incentives for both incumbents and competitors to invest and innovate, and will allow the Commission and the states to reduce regulation once effective facilities-based competition develops. Accordingly, the unbundling rules we adopt in this proceeding seek to promote the development of facilities-based competition. 8. We believe that the "necessary" and "impair" standards we adopt below address the Supreme Court's mandate and implement the statutory language and goals of the Act. The standards we adopt take into consideration alternatives outside the incumbent LEC's network, and whether those alternatives are actually available to the requesting carrier as a practical, economic, and operational matter. We consider not only the direct costs, but also other costs and impediments associated with using alternative elements that may constitute barriers to entry. We believe the Commission must assess these factors to determine the availability of alternatives, and whether access to the incumbent's network element thereby satisfies the "necessary" and "impair" standards of section 251(d)(2). 9. The unbundling standards we adopt in this Order also seek to encourage the rapid introduction of competition in all markets, including residential and small business markets. They seeks to create incentives for both incumbents and requesting carriers to invest and innovate in new technologies by establishing a mechanism by which regulatory obligations to provide access to network elements will be reduced as alternatives to the incumbent LECs' network elements become available in the future. In addition, the standards provide reasonable certainty regarding the availability of unbundled elements, thereby allowing requesting carriers to attract investment capital and move forward with implementing national and regional business plans that will allow them to serve the greatest number of consumers 10. To date, we have seen the development of facilities-based competition among providers of particular services in certain sectors of the market. For example, as discussed in more detail below, competitors have deployed their own fiber rings and approximately 700 circuit switches to provide local exchange and exchange access services primarily to medium and large business customers in high-density metropolitan areas. In addition, the record in this proceeding suggests that a growing number of carriers are deploying packet switches to provide data services in a number of markets, particularly for end users with substantial telecommunications needs. 11. Other local markets, however, particularly the residential and small business markets, and geographic markets outside of major metropolitan areas, have seen minimal competition. This may be due to the uncertainty surrounding the ability of competitive LECs to use reasonably priced unbundled network elements to serve these areas as a result of litigation concerning the Commission's unbundling rules. Because unbundled network elements have not been made fully available to requesting carriers as the Commission expected in 1996, we do not yet know the extent to which competition will develop once all of the unbundling rules are actually implemented by incumbent LECs. 12. Only recently have incumbent LECs provided access to combinations of unbundled loops, switches, and transport elements, often referred to as "the platform." Since these combinations of unbundled network elements have become available in certain areas, competitive LECs have started offering service in the residential mass market in those areas. For example, in January of this year, Bell Atlantic, as part of an agreement with the New York Public Service Commission, began offering the unbundled network element platform out of particular end offices in New York City. As a result, MCI WorldCom had acquired upwards of 60,000 new local residential customers in New York as of June 1999. AT&T also plans to serve local residential customers over the platform in Texas. 13. For effective competition to develop as envisioned by Congress, competitors must have access to incumbent LEC facilities in a manner that allows them to provide the services that they seek to offer, as contemplated in section 251(d)(2) of the Act. Despite the development of competition in some markets, incumbents still control the vast majority of the facilities that comprise the local telecommunications network, giving them advantages of economies of scale and scope not enjoyed by competitive LECs. Because competitors do not yet enjoy the same economies of scale, scope and ubiquity as the incumbent, they may be impaired if they do not have access, at least initially, to certain network elements supplied by the incumbent LEC. For example, without access to unbundled network elements, a competitive LEC may choose not to enter a particular market because the cost and delays associated with deploying its own facilities would be too high given the revenues obtainable from that market and the relative attractiveness of other potential new markets. Similarly, a competitive LEC may decline to enter a market because certain of their facilities are subject to economies of scale and scope such that the competitor would need a larger market share than it is likely to have initially. In such cases, competitors may choose to enter a certain market if they can obtain access to particular unbundled network elements on sufficiently favorable terms that such scale economies are overcome, and other potential markets no longer appear more attractive. 14. The standards and rules we adopt in this Order seek to build on industry experience and technological changes that have occurred in the telecommunications marketplace since the 1996 Act was enacted three years ago. Today, both incumbent LECs and requesting carriers are at the early stages of deploying innovative technologies to meet the ever-increasing demand for high-speed, high- capacity advanced services. To encourage competition among carriers to develop and deploy new advanced services, the marketplace for these services must be conducive to investment, innovation, and meeting the needs of consumers. Accordingly, our unbundling rules are designed to facilitate the rapid and efficient deployment of all telecommunications services, including advanced services. Specifically, unbundling rules that are based on a preference for development of facilities-based competition in the long run will provide incentives for both incumbents and competitors to invest and innovate, and should allow the Commission to reduce regulation once true facilities-based competition develops. 15. The unbundling standards we adopt in this order also are designed to be administratively practical and respond to changes in the marketplace as alternatives to the incumbent LECs' network elements become available. We are committed to reviewing the unbundling obligations in three years, and as the marketplace changes with the development of new technologies and increased facilities-based competition, we will modify the list of unbundled elements, as warranted. XVI. EXECUTIVE SUMMARY Section 251(d)(2)'s "Necessary" and "Impair" Standards. Section 251(d)(2)(A)'s "necessary" standard is a stricter standard that applies to proprietary network elements. Section 251(d)(2)(B)'s "impair" standard applies to non-proprietary network elements. Applying a stricter standard to proprietary network elements is consistent with Congress' intention to spur innovation and investment by both incumbent and competitive LECs. In applying these standards, we look first to what is occurring in the marketplace today. · Necessary. A proprietary network element is "necessary" within the meaning of section 251(d)(2)(A) if, taking into consideration the availability of alternative elements outside the incumbent's network, including self-provisioning by a requesting carrier or acquiring an alternative from a third party supplier, lack of access to that element would, as a practical, economic, and operational matter, preclude a requesting carrier from providing the services it seeks to offer. There are limited circumstances under which we may unbundle proprietary information or functionalities even if those elements are not strictly "necessary," as long as the "impair" standard is met. These circumstances are: (1) where an incumbent LEC, for the primary purpose of causing a particular network to be evaluated under the stricter "necessary" standard in order to avoid its unbundling obligation, implements only a minor modification to the network element to make the element proprietary; (2) where an incumbent LEC cannot demonstrate that the information or functionality that it claims is proprietary differentiates its services from its competitors' services, or is otherwise competitively significant; or (3) where lack of access to the proprietary element would jeopardize the goal of the 1996 Act to bring rapid competition to the greatest number of consumers. · Impair. The incumbent LECs' failure to provide access to a non-proprietary network element "impairs" a requesting carrier within the meaning of section 251(d)(2)(B) if, taking into consideration the availability of alternative elements outside the incumbent's network, including self-provisioning by a requesting carrier or acquiring an alternative from a third-party supplier, lack of access to that element materially diminishes a requesting carrier's ability to provide the services it seeks to offer. In order to evaluate whether there are alternatives actually available to the requesting carrier as a practical, economic, and operational matter, we look at the totality of the circumstances associated with using an alternative. In particular, our "impair" analysis considers the cost, timeliness, quality, ubiquity, and operational issues associated with use of the alternative. Goals of the Act. We also interpret the obligations imposed in section 251(d)(2) within the larger statutory framework of the 1996 Act. Congress apparently contemplated that we would consider additional factors by directing the Commission, in section 251(d)(2), to "consider at a minimum" the "necessary" and "impair" standards. The Supreme Court decision requires us to apply a limiting standard "rationally related to the goals of the Act." Accordingly, in addition to the factors set forth above, we may consider the following factors: · Rapid Introduction of Competition in All Markets. We may consider whether the availability of an unbundled network element is likely to encourage requesting carriers to enter the local market in order to serve the greatest number of consumers as rapidly as possible. We also note that Congress required Bell Operating Companies to demonstrate that they are providing loops, switching, transport, signaling and databases, and operator services/directory assistance in order to obtain in-region, interLATA approval. While the section 271 checklist does not determine definitively which elements all incumbent LECs are required to unbundle pursuant to section 251, it sheds some light on what Congress believed was required to open local markets to competition. Accordingly, we believe that we may consider whether requiring all incumbent LECs to unbundle these same elements would promote the rapid introduction of competition on a nationwide basis. · Promotion of Facilities-Based Competition, Investment, and Innovation. We may consider the extent to which the unbundling obligations we adopt will encourage the development of facilities-based competition by competitive LECs, and innovation and investment by both incumbent LECs and competitive LECs, especially for the provision of advanced services. · Reduced Regulation. We may consider the extent to which we can encourage investment and innovation by reducing regulatory obligations to provide access to network elements, as alternatives to the incumbent LECs' network elements become available in the future. · Certainty in the Market. We may consider how the unbundling obligations we adopt can provide the uniformity and predictability that new entrants and fledgling competitors need to develop national and regional business plans. We also consider whether the rules we adopt provide financial markets with reasonable certainty so that carriers can attract the capital they need to execute their business plans to serve the greatest number of consumers. · Administrative Practicality. We may consider whether the unbundling obligations we adopt are administratively practical to apply. Modification of the National List. · The Order recognizes that rapid changes in technology, competition, and the economic conditions of the telecommunications market will require a reevaluation of the national unbundling rules periodically. In order to encourage a reasonable period of certainty in the market, the Commission expects to reexamine the national list of unbundled network elements in three years. · Section 251(d)(3) permits state commissions to require incumbent LECs to unbundle additional elements as long as the obligations are consistent with the requirements of section 251 and the national policy framework instituted in this Order. · Removal of elements from the national list on a state-by-state basis would not be consistent with section 251 and the goals of the Act. Network Elements that Must be Unbundled. Applying the above factors, the Order concludes that the following network elements must be unbundled: · Loops. Incumbent local exchange carriers (LECs) must offer unbundled access to loops, including high-capacity lines, xDSL-capable loops, dark fiber, and inside wire owned by the incumbent LEC. The unbundling of the high frequency portion of the loop is being considered in another proceeding. · Subloops. Incumbent LECs must offer unbundled access to subloops, or portions of the loop, at any accessible point. Such points include, for example, a pole or pedestal, the network interface device, the minimum point of entry to the customer premises, and the feeder distribution interface located in, for example, a utility room, a remote terminal, or a controlled environment vault. The Order establishes a rebuttable presumption that incumbent LECs must offer unbundled access to subloops at any accessible terminal in their outside loop plant. · To the extent there is not currently a single point of interconnection that can be feasibly accessed by a requesting carrier, we encourage parties to cooperate in any reconfiguration of the network necessary to create one. If parties are unable to negotiate a reconfigured single point of interconnection at multi-unit premises, we require the incumbent to construct a single point of interconnection that will be fully accessible and suitable for use by multiple carriers. · Network Interface Device (NID). Incumbent LECs must offer unbundled access to NIDs. The NID includes any potential means of interconnection with customer premises inside wiring at the point where the carrier's local loop facilities end, such as at a cross connect device used to connect the loop to customer-controlled inside wiring. This includes all features, functions, and capabilities of the facilities used to connect the loop to premises wiring, regardless of the specific mechanical design. · Circuit Switching. Incumbent LECs must offer unbundled access to local circuit switching, except for local circuit switching used to serve end users with four or more lines in access density zone 1 in the top 50 Metropolitan Statistical Areas (MSAs), provided that the incumbent LEC provides non-discriminatory, cost- based access to the enhanced extended link throughout zone 1. (An enhanced extended link (EEL) consists of a combination of an unbundled loop, multiplexing/concentrating equipment, and dedicated transport. The EEL allows new entrants to serve customers without having to collocate in every central office in the incumbent's territory.) Local circuit switching includes the basic function of connecting lines and trunks on the line-side and port-side of the switch. The definition of the local switching element encompasses all of the features, functionalities, and capabilities of the switch. · Packet Switching. Incumbent LECs must offer unbundled access to packet switching only in limited circumstances in which the incumbent has placed digital loop carrier systems in the feeder section of the loop or has its Digital Subscriber Line Access Multiplexer (DSLAM) in a remote terminal. The incumbent will be relieved of this obligation, however, if it permits a requesting carrier to collocate its DSLAM in the incumbent's remote terminal on the same terms and conditions that apply to its own DSLAM. Packet switching is defined as the function of routing individual data message units based on address or other routing information contained in the data units, including the necessary electronics (e.g., DSLAMs). · Interoffice Transmission Facilities. Incumbent LECs must offer unbundled access to dedicated interoffice transmission facilities, or transport, including dark fiber. Dedicated interoffice transmission facilities are defined as incumbent LEC transmission facilities dedicated to a particular customer or carrier that provide telecommunications between wire centers owned by the incumbent LECs or requesting telecommunications carriers, or between switches owned by incumbent LECs or requesting telecommunications carriers. State commissions are free to establish reasonable limits governing access to dark fiber if incumbent LECs can show that they need to maintain fiber reserves. · Incumbent LECs must also offer unbundled access to shared transport where unbundled local circuit switching is provided. Shared transport is defined as transmission facilities shared by more than one carrier, including the incumbent LEC, between end office switches, between end office switches and tandem switches, and between tandem switches in the incumbent LEC's network. · Signaling and Call-Related Databases. Incumbent LECs must offer unbundled access to signaling links and signaling transfer points (STPs) in conjunction with unbundled switching, and on a stand-alone basis. The signaling network element includes, but is not limited to, signaling links and STPs. Incumbent LECs must also offer unbundled access to call-related databases, including, but not limited to, the Line Information database (LIDB), Toll Free Calling database, Number Portability database, Calling Name (CNAM) database, Operator Services/Directory Assistance databases, Advanced Intelligent Network (AIN) databases, and the AIN platform and architecture. We do not require incumbent LECs to unbundle access to certain AIN software that qualify for proprietary treatment. · Operations Support Systems (OSS). Incumbent LECs must offer unbundled access to their operations support systems. OSS consists of pre-ordering, ordering, provisioning, maintenance and repair, and billing functions supported by an incumbent LEC's databases and information. The OSS element includes access to all loop qualification information contained in any of the incumbent LEC's databases or other records, including information on whether a particular loop is capable of providing advanced services. Network Elements that Need Not be Unbundled. The following network elements need not be unbundled: · Operator Services and Directory Assistance (OS/DA). Incumbent LECs are not required to unbundle their OS/DA services pursuant to section 251(c)(3), except in the limited circumstance where an incumbent LEC does not provide customized routing to a requesting carrier to allow it to route traffic to alternative OS/DA providers. Operator services are any automatic or live assistance to a consumer to arrange for billing or completion of a telephone call. Directory assistance is a service that allows subscribers to retrieve telephone numbers of other subscribers. Incumbent LECs, however, remain obligated under the non-discrimination requirements of section 251(b)(3) to comply with the reasonable request of a carrier that purchases the incumbents' OS/DA services to rebrand or unbrand those services, and to provide directory assistance listing updates in daily electronic batch files. · Shared Transport where Circuit Switching is not Unbundled. Incumbent LECs are not required to unbundle shared transport where they are not required to offer unbundled local circuit switching, as described above. · Packet Switching. Incumbent LECs are not required to unbundle packet switching, except in a limited circumstance. Competitive LECs are actively deploying packet switches to serve high-volume customers, and are not impaired in their ability to offer service to such customers without access to the incumbent LEC's facilities. Competitive LECs are impaired, however, in their ability to provide services to small-volume users without access to unbundled packet switching. Nonetheless, we consider the other goals of the Act in making our unbundling determination, and conclude that given the nascent nature of the advanced services market and the Act's goal to provide incentives to all carriers to invest and innovate, incumbent LECs are generally not required to unbundle packet switching. Section 271-Related Issues. · If a network element on the section 271 competitive checklist is not required to be unbundled pursuant to section 251(c)(3) (i.e., local circuit switching and shared transport in certain circumstances), Bell Operating Companies are not required to offer unbundled access to any such checklist items in compliance with the Commission's pricing rules. While the applicable price, terms, and conditions for that element are subject to section 201(b) and 202(a) of the Act, it would be counterproductive to mandate that incumbents offer elements at forward looking prices when the market price, rather than a regulated rate, should prevail in circumstances when an element is not unbundled. Combinations of Network Elements. · Given the pendency of litigation in the Court of Appeals for the Eighth Circuit, the Order declines to define the enhanced extended link as a separate network element, nor does it address whether an incumbent LEC must combine network elements that are not already combined in the network. Further Notice of Proposed Rulemaking: Use of Unbundled Network Elements to Provide Exchange Access Service. · The Further Notice seeks comment on whether there is any basis in the statute or our rules under which incumbent LECs could decline to provide entrance facilities (the link between an interexchange carrier's point of presence and an incumbent's switch or serving wire center) at unbundled network element prices. · The Further Notice also invites parties to refresh the record on whether requesting carriers may use unbundled dedicated or shared transport facilities in conjunction with unbundled switching to originate or terminate interstate toll traffic to customers to whom the requesting carrier does not provide local exchange service. XXVI. BACKGROUND 27. On August 8, 1996, the Commission adopted the Local Competition First Report and Order, implementing the local competition provisions of the 1996 Act. In that order, the Commission established rules governing the obligations of incumbent LECs to open their local networks to competition pursuant to the requirements of section 251 of the 1996 Act. Among other things, the order adopted rules implementing the network unbundling requirements of sections 251(c)(3) and 251(d)(2) of the 1996 Act. Section 251(c)(3) imposes a duty on all incumbent LECs to provide to competitors access to network elements on an unbundled basis. Section 251(d)(2) provides that, in determining which network elements should be unbundled under section 251(c)(3), the Commission shall consider, "at a minimum, whether -- (A) access to such network elements as are proprietary in nature is necessary; and (B) the failure to provide access to such network element would impair the ability of the telecommunications carrier seeking access to provide the services that it seeks to offer." 28. In the Local Competition First Report and Order, the Commission applied its interpretation of the "necessary" and "impair" standards of section 251(d)(2) to the unbundling requirements of section 251(c)(3). Specifically, the Commission defined "necessary" to mean "an element is a prerequisite for competition," and it defined "impair" to mean "to make or cause to become worse; diminish in value." The Commission also determined that a requesting carrier's ability to offer service is "impaired" or "diminished in value" if "the quality of the service the entrant can offer, absent access to the requested element, declines" or if "the cost of providing the service rises." 29. After addressing the "necessary" and "impair" standards, the Commission adopted rule 51.319, which sets forth the network elements that incumbent LECs were required to make available to requesting carriers on an unbundled basis. Section 51.319 of the Commission's rules required incumbent LECs to offer unbundled access to the following network elements: (1) local loops; (2) network interface devices; (3) local switching; (4) interoffice transmission facilities; (5) signaling networks and call-related databases; (6) operations support systems; and (7) operator services and directory assistance. Section 51.317 of the Commission's rules allowed states to impose additional unbundling requirements pursuant to the Commission's interpretation of section 251(d)(2). 30. Following adoption of the Local Competition First Report and Order, incumbent LECs and state commissions filed various challenges to the Commission's rules; these appeals were consolidated in the Eighth Circuit. The Eighth Circuit, among other holdings, rejected the incumbent LECs' argument that, in determining which elements were subject to the unbundling requirements, the Commission had not properly applied the "necessary" and "impair" standards of section 251(d)(2). Accordingly, the Eighth Circuit upheld section 51.319. The Supreme Court granted several parties' requests to review the Eighth Circuit's decision. 31. In its January 25, 1999 opinion, the Supreme Court reversed the Eighth Circuit's decision on this issue, stated that section 51.319 should be vacated, and remanded the matter for further proceedings. While the Court affirmed that the Commission has jurisdiction to implement the local competition provisions of the 1996 Act, including the unbundling requirements in section 251, it concluded that the Commission had not adequately considered the "necessary" and "impair" standards of section 251(d)(2). The Court found, among other things, that the Commission, in deciding which elements must be unbundled, did not adequately take into consideration the "availability of elements outside the incumbent's network." The Court also faulted the Commission's "assumption that any increase in cost (or decrease in quality) imposed by a denial of a network element renders access to that element 'necessary,' and causes the failure to provide that element to 'impair' the entrant's ability to furnish its desired services. . ." In addition, the Court criticized the Commission's interpretation of section 251(d)(2) because it "allows entrants, rather than the Commission, to determine" whether the requirements of that section are satisfied. On April 16, 1999, we released a Second Further Notice in this docket seeking comment on the appropriate unbundling standard, and which network elements should be unbundled. XXXII. FRAMEWORK FOR DECIDING WHAT NETWORK ELEMENTS MUST BE UNBUNDLED PURSUANT TO SECTION 251 A. Overview 33. In this section, we discuss our framework for determining whether a particular network element should be unbundled. We interpret the terms "necessary," "impair," and "proprietary" in section 251(d)(2) in a manner that gives substance to those terms. We then discuss how we will evaluate alternative elements that are available through self-provisioning or from third-party suppliers. In considering whether to unbundle a particular network element, we look first to what is occurring in the marketplace today. For some network elements, we are beginning to see competitors using alternatives in discrete situations. In order to determine whether these alternative sources of network elements are actually available as a practical, economic, and operational matter, so that incumbents should be free of any unbundling obligations for that element, we look at several factors, including cost, ubiquity, quality, timeliness, and operational impediments. 34. We acknowledge that given the complexity associated with a competitive LEC's decision to enter a certain market, it is extremely difficult to identify one particular factor that is dispositive of whether or not a competitor will seek to offer a particular service in any given market. For example, even where a competitive LEC's costs to provide a service may be comparable to the incumbent's costs to provide a similar service, the competitive LEC, because it cannot enter all markets simultaneously, may choose not to enter a particular market at a particular time because there are other markets that are relatively more profitable to serve. The competitive LEC might also be dissuaded from entering a market because of subsidy distortions or other regulatory factors. Conversely, notwithstanding the fact that a competitive LEC's infrastructure costs in a particular market may be materially different from the incumbent LECs' costs, the competitive LEC may still choose to enter that market if it can provide more services over its network infrastructure, or offer marketing, service, or technical innovations for which customers will pay a premium. 35. Although we may not be able to identify with precision a competitor's incentives, or lack of incentives to enter a particular market, we nonetheless find that evidence demonstrating the lack of competition in certain areas of the country and among certain classes of customers is a strong indicator that there may exist economic and other types of barriers that may, at a minimum, impair a competitor's ability to compete vis-…-vis the incumbent. Accordingly, based on evidence provided in the record, we use our administrative judgment to identify several factors, including cost, ubiquity, quality, timeliness, and operational impediments, that we find particularly helpful in explaining whether a competitor's ability to provide the service it seeks to offer is impaired without access to a particular unbundled network element. Based on the actual state of competition, we look at these factors and their relationship to alternative sources of network elements to determine whether the alternatives are actually available as a practical, economic, and operational matter. 36. In particular, we examine both the direct and other costs a carrier incurs to substitute the alternative network element for the incumbent LEC's network element. We also consider whether self-provisioning or purchasing a network element from a third- party supplier would prevent a requesting carrier from entering the market within a reasonable time, or from expanding its operations to meet promptly the demand of its customers. In addition to costs and delays, we consider whether using alternative sources of network elements introduces quality differences or operational or technical impediments that may impair a requesting carrier's ability to provide the services that it seeks to offer. Specifically, we assess whether use of an alternative source of the network element would cause a requesting carrier's customers to experience degraded service. 37. We also consider the extent to which a requesting carrier can compete for customers on a wide-spread basis using alternative sources of the elements outside the incumbent's network. In some cases, to compete effectively with the incumbent LEC for the same customers, competitive LECs must be able to attain similar economies of scale that can only be achieved by serving a broad base of customers within a geographic area. Although theoretically, all or part, of an incumbent LEC's network can be replicated at some cost, as a practical matter, replication of elements in a ubiquitous manner may impair a requesting carrier's ability to compete vis-…-vis the incumbent. If the competitive LEC must deploy multiple facilities in order to be able to bring competition to a broad base of customers within a geographic area, the costs and delays associated with deploying facilities will likely be magnified, and could "materially diminish" that competitor's ability to provide the services that it seeks to offer. 38. We find that the language of section 251(d)(2) and the Supreme Court decision suggest that we should consider, in addition to the "necessary" and "impair" standards, the overall goals of the 1996 Act. Section 251(d)(2) states that the Commission shall "consider, at a minimum," the "necessary" and "impair" standards, thus leaving the Commission free to consider other relevant factors. In addition, the Supreme Court decision requires us to apply a limiting standard "rationally related to the goals of the Act." Moreover, as a policy matter, we believe that we may consider additional factors to ensure that the unbundling requirements promote the goals of the Act. 39. Accordingly, we may consider, in addition to the "necessary" and "impair" standards, whether the unbundling obligations we adopt are likely to: (1) encourage competitive LECs to rapidly enter the local market and serve the greatest number of consumers; (2) advance the development of facilities-based competition by competitive LECs, and encourage investment and innovation in new technologies and new services by both incumbent and competitive LECs; (3) reduce regulation of unbundled network elements as alternatives to the incumbent LECs' network elements become available in the future, (4) provide certainty in the marketplace that will allow new entrants and fledgling competitors to develop national and regional business plans and bring the benefits of competition to the greatest number of consumers; and (5) be administratively practical to apply. We conclude that these important policy goals can only be furthered by adoption of a national list of unbundled elements that takes into consideration, where appropriate, discrete geographic and product market variations that create exceptions to the incumbent LECs' general duty to unbundle the elements on the list. 40. We do not assign any particular weight to the factors we identify above. Rather, we consider the relationship among the various factors to determine whether an incumbent LEC's network element should be unbundled. Indeed, there may be circumstances in which there is significant evidence that competitors are impaired without unbundled access to a particular element, but requiring incumbent LECs to unbundle that element would be inconsistent with the goals of the Act. A. The "Necessary" and "Impair" Standards of Section 251(d)(2) 1. Application of the "necessary" and "impair" standards to proprietary and non-proprietary elements under Section 251(d)(2) a. Background 41. In the Local Competition First Report and Order, the Commission concluded that section 251(d)(2) establishes separate standards that apply to proprietary and non- proprietary network elements. Specifically, the Commission determined that the "necessary" standard of section 251(d)(2)(A) applies to proprietary elements, and that the "impair" standard in section 251(d)(2)(B) applies to non-proprietary elements. In the Notice, we sought comment on this interpretation of section 251(d)(2). In particular, we asked parties to comment on the difference between the "necessary" and "impair" standards. Noting that the Act employs two different terms, we asked if the Commission must apply different criteria to determine whether a network element meets these standards. 42. Only a couple of commenters dispute the Commission's previous decision to apply the "necessary" standard to proprietary elements and the "impair" standard to non- proprietary elements. In particular, Sprint argues that the "necessary" and "impair" standards apply only to proprietary elements and thus, all non-proprietary elements must be unbundled. a. Discussion 43. We find no reason to change our framework for analyzing network elements under section 251(d)(2). In subpart (A) of section 251(d)(2), "necessary' modifies elements that are "proprietary in nature" while in subpart (B), "impair" modifies all other network elements. We agree with the majority of commenters that the "necessary" standard of section 251(d)(2)(A) is a higher standard that applies to proprietary network elements or to proprietary functions within an element. We believe that our conclusion that section 251(d)(2) establishes a higher standard for proprietary network elements than for non-proprietary elements is consistent with both the language of the statute and the goals of the 1996 Act to encourage incumbent LECs and competitive LECs to innovate and invest in new technologies. Specifically, incumbent LECs will have an on-going incentive to innovate if we ensure that their investment in the proprietary portions of their network is protected. While competitive LECs will have access to the incumbent LEC's proprietary network elements where necessary, they will not have unlimited access to those elements. We believe that this balanced approach provides competitive LECs with an incentive to innovate and invest in new technologies that will differentiate their services from the incumbents' services. We note that applying the "necessary" standard to proprietary elements, and the "impair" standard to non-proprietary elements is also consistent with the Commission's previous interpretation of this section that was implicitly adopted by the Supreme Court. 1. Definition of "Proprietary in Nature" a. Background 44. Section 251(d)(2)(A) states that "[i]n determining what network elements should be made available for purposes of subsection (c)(3), the Commission shall consider, at a minimum, whether . . . . access to such network elements as are proprietary in nature is necessary." In the Local Competition First Report and Order, the Commission referred to proprietary network elements as including, for example, "those elements with proprietary protocols or elements containing proprietary information." The Commission found in the Local Competition First Report and Order that for most network elements subject to the unbundling obligations of section 51.319, parties had not identified any proprietary concerns. For those network elements where parties did identify proprietary concerns, the Commission found that access to those network elements was "necessary." 45. In the Notice, we sought comment on whether we should consider network elements as non-proprietary if the interfaces, functions, features and capabilities of the elements sought by the requesting carrier are defined by recognized standard-setting bodies (e.g. ITU, ANSI, or IEEE), are defined by Bellcore requirements, or otherwise are widely available from vendors. We further requested comment on whether the term "proprietary" should be limited to information, software, or technology that can be protected by patents, copyright or trade secret laws. There is general agreement among the parties that the Commission should define proprietary, under section 251(d)(2)(A), consistent with intellectual property categories. Several competitive LECs maintain that we must define the term "proprietary" narrowly so as not to create incentives for incumbent LECs to attempt to deny access to unbundled network elements on proprietary grounds. a. Discussion 46. In this Order, we adopt a limited definition of the phrase "proprietary in nature" that tracks the intellectual property categories of patent, copyright, and trade secrets. The majority of parties addressing this issue support using intellectual property law as a basis for defining "proprietary in nature." We agree, and find that the intellectual property laws governing patent, copyright and trade secrets find a common purpose in Congress' intention to protect proprietary interests under section 251(d)(2). The intellectual property laws are designed to protect the incentives of authors and inventors to innovate. Similarly, Congress recognized that an incumbent LEC's incentive to innovate could be adversely affected by requiring incumbent LECs to unbundle proprietary portions of network elements to requesting carrier-competitors. Congress therefore required the Commission to consider whether unbundling in such instances is "necessary." 47. We find that if an incumbent LEC can demonstrate that it has invested resources (time, material, or personnel) to develop proprietary information or network elements that are protected by patent, copyright, or trade secret law, the product of such an investment is "proprietary in nature" within the meaning of section 251(d)(2)(A). This definition is consistent with the 1996 Act's policy of preserving the incumbent LECs' innovation incentives. It is also consistent with the Commission's conclusion, in the Local Competition First Report and Order, that in some instances it will be "necessary" for new entrants to obtain access to proprietary elements. Finally, our decision to define interests that are "proprietary in nature" along established intellectual property categories is consistent with the Department of Justice and Federal Trade Commission Guidelines for the Licensing of Intellectual Property. 48. Our definition excludes elements that are based on widely accepted industry documents or on standards commonly used by a standards-setting body (e.g. ITU, ANSI, IEEE) or by vendors. There are few innovation incentives associated with elements that are based on well-recognized standards that are widely available in the market, and we therefore are not required to scrutinize such elements under the higher "necessary" standard. 49. Section 251(d)(2) directs the Commission to "consider, at a minimum" whether access to proprietary elements is necessary. As discussed below, this discretionary language permits us to consider other factors, in addition to the "necessary" standard, in making our unbundling determination. We find that there are several circumstances which, if they exist with regard to information or functionalities that the incumbent LEC claims are proprietary, will permit us to order unbundling of the proprietary information or functionalities even if unbundled access to the element is not strictly "necessary," as long as the "impair" standard is met. The first circumstance is where an incumbent LEC, for the primary purpose of causing a particular network to be evaluated under the stricter "necessary" standard in order to avoid its unbundling obligation, implements only a minor modification to the network element to make the element "proprietary in nature." Denying a requesting carrier access to the element in this circumstance would not encourage innovation and investment by the incumbent LEC, which is one of the goals of the 1996 Act, and would reduce consumer benefits by failing to facilitate rapid deployment of competitive services. The second circumstance is where an incumbent LEC cannot demonstrate that the information or functionality that it claims is proprietary differentiates its services from its competitors' services, or is otherwise competitively significant. Information or functionalities that do not distinguish an incumbent LEC's service from that of its competitors are unlikely to be the focus of an incumbent LEC's efforts to innovate, and therefore do not require the high level of protection normally afforded to proprietary elements under the "necessary" standard. The third circumstance is where we find that lack of access to the proprietary element would jeopardize the goal of the 1996 Act to bring rapid competition to the greatest number of customers. In such a circumstance, we may find that the incumbent LEC's asserted proprietary interest is outweighed by the benefits of facilitating more rapid deployment of competition for the greatest number of consumers. Given the significance of the incumbent LECs' proprietary interests, and our commitment to do nothing to discourage innovation and investment by all carriers, we do not envision, outside of these limited circumstances, unbundling a proprietary network element unless the "necessary" standard is satisfied. Moreover, we cannot imagine a situation where we would order unbundling of a proprietary element unless the "impair" standard has been met. 50. We agree with those commenters that argue that "proprietary in nature" applies only to the proprietary interests of the incumbent LEC and not to proprietary interests of third parties. Limiting the definition of "proprietary" to interests held by the incumbent LEC is consistent with section 251(d)(2)(A)'s goal of preserving the incumbent LECs' incentives to innovate. Moreover, sections 251(c) and 251(d)(2), by their terms, apply only to the proprietary interests of those parties subject to the Act's unbundling obligations -- incumbent LECs. Thus, section 251(d)(2) only indirectly affects, if at all, the innovation incentives of third parties. 51. Finally, we reject CompTel's argument that we should limit the application of proprietary network elements to those circumstances in which the incumbent LEC "discloses" proprietary information. CompTel argues that if unbundling merely provides a requesting carrier with the "use" of a proprietary methodology, but does not "disclose" or access the proprietary information itself, the element is not proprietary. We find that the "use" or "disclosure" rationale does not promote the goal of the Act to encourage investment and innovation, and thus is at odds with our decision to define "proprietary" along intellectual property categories. 52. Pursuant to patent law, patent holders trade monopoly rights in their inventions in exchange for a requirement that they disclose the technical details underlying the patent. Patent holders thus recover their investments by obtaining a monopoly on the "use" of their protected intellectual property. We agree with Ameritech that limiting the definition of "proprietary" to requests that would reveal proprietary information would turn intellectual property law and incentives to innovate on their head; "instead of granting exclusivity in exchange for disclosure, it would withhold exclusivity unless needed to avoid disclosure." Similarly, under copyright laws, an illegal copy or "use" of a protected work can damage an author's incentive to produce the work. We note, however, that the disclosure of sensitive customer information contained in unbundled network element must be consistent with the requirements of section 222. 1. The "Necessary" Standard of Section 251(d)(2)(A) a. Background 53. In the Local Competition First Report and Order, the Commission defined a "necessary" network element as one that is a "prerequisite" to competition. The Commission stated that "in some instances it will be 'necessary' for requesting carriers to obtain access to proprietary elements (e.g., elements with proprietary protocols or elements containing proprietary information) because without such elements, the ability of requesting carriers to compete would be significantly impaired or thwarted." It also acknowledged that prohibiting incumbents from refusing access to proprietary elements could reduce their incentives to offer innovative services. The Commission did not identify any proprietary elements subject to unbundling in the Local Competition First Report and Order, except that it acknowledged the claims of several parties that access to the incumbent LECs' advanced intelligent network (AIN) may raise proprietary concerns. It nevertheless concluded that access to AIN is "necessary" within the meaning of section 251(d)(2)(A). 54. In the Notice, the Commission sought comment on the definition of "necessary" for the purpose of determining proprietary network elements that must be unbundled pursuant to the requirements of section 251(d)(2)(A), as well as on what factors or criteria the Commission should apply in determining whether access to a network element is "necessary." 55. Several competitive LECs assert that in determining whether unbundling of a proprietary network element is "necessary," the Commission must evaluate whether the requesting carrier can obtain comparable functionality from an alternative network element. They maintain that if the requesting carrier would experience a material loss in functionality without the network element that that the incumbent LEC claims is proprietary, then the incumbent LEC's network element is "necessary" within the meaning of section 251(d)(2)(A). The incumbent LECs assert generally that both the "necessary" and "impair" standards require an analysis of whether lack of access to their networks elements, taking into consideration alternatives outside the incumbent's network, would deny an efficient competitor a meaningful opportunity to compete. These commenters argue that the "necessary" standard requires the Commission to accept a higher degree of proof that alternatives to the element are not available. a. Discussion 56. We conclude that a proprietary network element is "necessary" within the meaning of section 251(d)(2)(A) if, taking into consideration the availability of alternative elements outside the incumbent's network, including self-provisioning by a requesting carrier or acquiring an alternative from a third-party supplier, lack of access to that element would, as a practical, economic, and operational matter, preclude a requesting carrier from providing the services it seeks to offer. We agree with NTIA that the proper focus of the "necessary" standard is whether access to the incumbent LEC's proprietary element is absolutely required for the competitor's provision of its intended service. We find, therefore, that an incumbent LEC must provide access to a proprietary element, if withholding access to the element would prevent a competitor from providing the service it seeks to offer. In other words, we conclude that an incumbent LEC's proprietary network element would only be available to a competitor if the competitor is unable to offer service, without access to the element, because no practical, economic, and operational alternative is available, either by self-provisioning or from other sources. 57. The standard we assign to the term "necessary," as used in section 251(d)(2)(A), is consistent with the Supreme Court's decision because it considers alternatives available outside the incumbent's network and gives substance to the meaning of "necessary." Moreover, insofar as the standard focuses on the competitor's ability to furnish a desired service, and not merely on whether profits are increased by using the incumbent's network, the standard is also consistent with the Court's instruction that we must "apply some limiting standard, rationally related to the goals of the Act." 58. This "necessary" standard differs from the "impair" standard we adopt below because a "necessary" element would, if withheld, prevent a carrier from offering service, while an element subject to the "impair" standard would, if withheld, merely limit a carrier's ability to provide the services it seeks to offer. We therefore disagree with the standards proposed by ALTS and other competitive LECs that access to a proprietary element is "necessary" if the entrant would experience a material loss in functionality without access to the element. A standard based on a test of "material loss" in functionality requires only that the competitive LEC's ability to compete be materially affected in some way, as opposed to precluded, and ignores the higher degree of protection normally afforded intellectual property rights. The incumbent LECs argue that the "necessary" standard is a higher standard that is intended to preserve their incentive to invest in proprietary protocols, and that access to a proprietary element is "necessary" only if lack of access to that element would deny an efficient competitor a meaningful opportunity to compete. We agree with the incumbent LECs' concerns regarding the preservation of their investment incentives. We believe that our standard, by requiring that a requesting carrier be precluded as a practical, economic, and operational matter from providing service without access to the proprietary information, sufficiently protects the incumbents' proprietary property from nonessential access by competitors. 59. We reject, however, the incumbent LECs' proposal to base the "necessary" standard on the requirements of an efficient competitor. As we explain below in our discussion of the "impair" standard, we do not affirmatively base our unbundling standard on an efficient competitor because we conclude that the marketplace is better able than regulators to distinguish efficient competitors from inefficient competitors. We also note that GTE and SBC state that few, if any, network elements are entirely proprietary in nature. Other commenters point out that most network equipment and services are non-proprietary because of the need for interoperability of networks. We therefore expect that the "necessary" standard will be invoked only when there is a serious question of whether access to the element will infringe upon the incumbent's intellectual property. 1. The "Impair" Standard of Section 251(d)(2)(b) a. Background 60. In the Local Competition First Report and Order, the Commission adopted a dictionary definition of the term "impair" that means "to make or cause to become worse; diminish in value." The Commission stated that "generally . . . an entrant's ability to offer a telecommunications service is 'diminished in value' if the quality of the service the entrant can offer, absent access to the requested element, declines and/or the cost of providing the service rises." In particular, the Commission interpreted the "impair" standard as requiring an evaluation of whether the failure of an incumbent to provide access to a network element would decrease the quality, or increase the financial or administrative cost of the service a requesting carrier seeks to offer. 61. In the Notice, we sought comment on the meaning of the term "impair," and asked whether we should adopt a standard under which we examine whether the new entrant's ability to offer a telecommunications service in a competitive manner is materially diminished in value. We also sought comment on the factors or criteria we should adopt to determine whether failure to provide access to the incumbent LEC's network elements would impair an entrant's ability to provide service within the meaning of section 251(d)(2). 62. The incumbent LECs argue generally that a requesting carrier is impaired if, after taking into account the availability of elements from alternative sources outside the incumbent's network, lack of access to the requested element would deny a competitor a meaningful opportunity to compete. This standard is similar to the standard the incumbent LECs propose for the "necessary" standard under section 251(d)(2)(A). GTE argues that failure to provide access to a network element would impair a requesting carrier's ability to provide service only where the element is essential to competition, and there is convincing evidence that the carrier cannot compete effectively using an alternative network element. Several incumbent also maintain that we must consider all available alternatives, including those available from other suppliers and through self-provisioning by the requesting carrier. The Texas PUC proposes that a competitor is impaired if, looking at the marketplace as a whole, lack of access to the incumbent's network element causes it to incur an increase in cost such that the competitor does not have a meaningful opportunity to compete. The competitive LECs and the Illinois Commerce Commission propose a standard by which a carrier would be impaired if, after taking into account the availability of elements from alternative sources outside the incumbent's network, lack of access to the requested element would materially diminish the requesting carrier's ability to provide service. The difference between the standard proposed by the competitive LECs and the standard proposed by the incumbent LECs is essentially the difference between whether lack of access to an unbundled network element "denies" or "materially diminishes" the ability of a competitor to provide the services it seeks to offer. Many competitive LECs also assert that the incumbent LECs' failure to provide access to an element would impair a requesting carrier's ability to provide service where there is no competitive wholesale market for the requested element. a. Discussion (i) The "Impair" analysis 63. We conclude that the failure to provide access to a network element would "impair" the ability of a requesting carrier to provide the services it seeks to offer if, taking into consideration the availability of alternative elements outside the incumbent's network, including self-provisioning by a requesting carrier or acquiring an alternative from a third-party supplier, lack of access to that element materially diminishes a requesting carrier's ability to provide the services it seeks to offer. We find that a materiality component, although it cannot be quantified precisely, requires that there be substantive differences between the alternative outside the incumbent LEC's network and the incumbent LEC's network element that, collectively, "impair" a competitive LEC's ability to provide service within the meaning of section 251(d)(2). We therefore agree with the Illinois Commerce Commission that where a competing LEC's "ability to offer a telecommunications service in a competitive manner is materially diminished in value without access to that element," the competitor's ability to provide its desired services would be impaired. 64. We believe that a standard that includes a "materiality" component gives substance to the "impair" standard of section 251(d)(2)(B), and responds to the Supreme Court's concern that we "apply some limiting standard, rationally related to the goals of the Act." A standard that includes a materiality component preserves requesting carriers' ability to provide service using unbundled elements, as contemplated by the Act, and encourages them to invest and innovate. As envisioned by Congress, requesting carriers may need each of the three separate means of providing service (resale of the incumbent LEC's service, use of unbundled incumbent LEC network elements, deployment of self-provisioned facilities), or various combinations of these means, in order to serve different customer classes in different areas. The purchase of unbundled network elements from the incumbent should serve as a transitional strategy that will provide requesting carriers with the ability to gain a sufficient volume of business to justify economical deployment of their own facilities. 65. Although we recognize that the existence of some significant level of competitive LEC facilities deployment is probative of whether competitive LECs are impaired from providing service within the meaning of section 251(d)(2), we decline to adopt the incumbent LECs' position that the presence of a single competitor providing service, without using the incumbent's unbundled network elements, is dispositive evidence that a competitor's ability to provide service generally would not be impaired without access to such elements. According to Bell Atlantic, if an efficient competitor can and does provide service without access to the incumbent's network element, it is irrelevant whether a less efficient competitor might claim that it would be impaired without access to the element. We find that the "efficiency" argument raised by Bell Atlantic and other incumbent LECs is more relevant to the length of time a competitor has been in business than to the efficiencies created by the competitor's inherent capabilities or cost structure. More importantly, however, we agree with MCI WorldCom that the Act is not calibrated to the performance of the company whose business plan allows it to rely the least on the incumbent LEC's network elements. The provisions of the 1996 Act do not contemplate that either the incumbent LEC or the regulator will determine whether a particular carrier is "efficient." Rather, the Act is designed to create a regulatory framework that requires incumbent LECs to make network elements subject to the unbundling obligations of section 251 available to all requesting carriers, subject to the requirements of section 251(d)(2), and allows the marketplace to determine ultimately which competitors thrive or survive. 66. Moreover, the ability of one or more competitors to serve certain customers in a particular market is not dispositive of whether competitive LECs without unbundled access to the incumbent LEC's facilities are able to compete for other customers in the same market or for customers in other markets. In some markets, particularly those markets serving high-volume business customers, it may be practical and economical for competitive LECs to compete using self-provisioned facilities. In other markets, however, typically those markets consisting of residential consumers and small businesses, the delay and costs associated with self-provisioning a network element will preclude those same competitors, or others, from assuming the risk of entry, unless they can purchase unbundled elements from the incumbent. We agree with the commenters that point out that we cannot evaluate the needs of every potential carrier seeking access to each network element on a case-by-case basis. We conclude, however, that we should not adopt rules that would deny access to network elements to all competitors based on the presence of a single competitor that has been able to enter without the use of a particular unbundled network element from the incumbent LEC. 67. We believe that Congress rejected implicitly the argument that the presence of a single competitor, alone, should be dispositive of whether a competitive LEC would be "impaired" within the meaning of section 251(d)(2). For example, although Congress fully expected cable companies to enter the local exchange market using their own facilities, including self-provisioned loops, Congress still contemplated that incumbent LECs would be required to offer unbundled loops to requesting carriers. A standard that would be satisfied by the existence of a single competitive LEC using a non-incumbent LEC element to serve a specific market, without reference to whether other competitive LECs are "impaired" under section 251(d)(2), would be inconsistent with the Act's goal of creating robust competition in telecommunications. In particular, such a standard would not create competition among multiple providers of local service that would drive down prices to competitive levels. Indeed, such a standard would more likely create stagnant duopolies comprised of the incumbent LEC and the first new entrant in a particular market. An absence of multiple providers serving various markets would significantly limit the benefits of competition that would otherwise flow to consumers. 68. On the other hand, we are not persuaded by arguments of competitive LECs that the "impair" standard is met only once it is determined that a wholesale market exists for a particular element. We agree with the incumbent LECs that basing the "impair" standard on the existence of a wholesale market does not take into consideration self-provisioning as a viable substitute to the incumbent LECs' network elements. The Supreme Court decision in Iowa Utils. Bd. expressly faulted the Commission's analysis in the Local Competition First Report and Order for not comparing use of the incumbent LEC's element with "self-provision" or with "purchas[ed elements] from another provider." We find that, in order to thoroughly evaluate the availability of alternative elements outside of the incumbent LEC's network, we must consider elements available from all sources, including those elements available from third-party suppliers and through self-provisioning. 69. Several of the incumbent LECs argue that our standard should be based on an analysis similar to the one used by courts in determining whether, according to the essential facilities doctrine, a firm must share its facilities with competitors. We disagree. Although we acknowledge that the Supreme Court referred to the possibility of adopting a limiting standard based on the essential facilities doctrine, we find nothing in the legislative history or statutory language of the 1996 Act, or in the Court's decision that requires us to apply that doctrine in determining which network elements the incumbent LECs must unbundle. Indeed, the Court expressly declined to decide, as a matter of law, whether the essential facilities doctrine is mandated by section 251(d). Further, we believe that the standard under section 251(d) better reflects the overall goals of the Act. Accordingly, as discussed more fully below, we describe several factors that should be considered in determining whether a particular network element must be unbundled pursuant to section 251(c)(3). 70. As an initial matter, the legislative history and statutory language of the Act indicate that Congress did not intend to codify the essential facilities doctrine when it enacted section 251(d)(2). Specifically, the legislative history indicates that Congress was aware of antitrust principles and the essential facilities doctrine, in particular, when it considered the 1996 Act. At least since 1991, the Senate had considered telecommunications legislation that expressly referred to "essential facilities." Yet, in spite of its awareness of this doctrine, Congress did not adopt an essential facilities test for unbundling of network elements. Congress chose, instead, to adopt unbundling requirements that are based upon the "necessary" and "impair" standards of section 251(d)(2). Moreover, section 601(b)(1) of the Act expressly preserves the existing antitrust laws, indicating that Congress intended for the Act to augment, not replace, traditional antitrust rules. 71. The essential facilities doctrine is an antitrust doctrine that imposes an obligation on a firm that controls facilities that are essential for the existence of competition between itself and a competitor to share such facilities on non-discriminatory terms. The doctrine creates a narrow exception to the general antitrust presumption that a single firm may decline to deal with another firm. Under the essential facilities doctrine, a court may require a firm possessing monopoly control over an essential input to deal with a competitor, if it is shown that the monopolist is misusing control of an essential facility to foreclose competition in a downstream market. 72. Although we find that the essential facilities doctrine promotes the same economic and policy goals embodied in the 1996 Act, we find it to be of limited assistance in our analysis of the unbundling obligations of the Act because, as NTIA explains, the Act plainly imposes on incumbent LECs a broader duty to deal with competitors than does the essential facilities doctrine. In particular, the essential facilities doctrine differs from the analysis the Commission must undertake under section 251(d)(2) because Congress has already created an affirmative obligation for incumbent LECs to make their facilities available to competitors. Specifically, section 251(c)(3) imposes on incumbent LECs a general obligation to provide access on an unbundled basis to any network elements that the Commission identifies under section 251(d)(2). This obligation is not limited to situations in which the incumbent is misusing control of a unique facility to foreclose competition in a downstream market. Rather, section 251(d)(2) requires incumbents to share their facilities if competitors are merely "impaired" in their ability to provide services they seek to offer. In addition, sections 251(c)(3) and 251(d)(2) require incumbent LECs to make their facilities available at cost-based rates, whereas the essential facilities doctrine allows monopolists to continue charging monopoly rates for use of their facilities. 73. It is particularly notable that although the essential facilities doctrine is referenced in several Supreme Court rulings, the Supreme Court has never explicitly adopted the doctrine. Moreover, because antitrust jurisprudence has not clearly defined the contours of the essential facilities doctrine, the doctrine provides limited guidance in developing a limiting standard under section 251(d)(2). In order to establish liability under the essential facilities doctrine, a plaintiff must establish the existence of five elements: 1) a monopolist controls an essential facility; 2) the competitor is unable to practically or reasonably duplicate the essential facility; 3) the monopolist denies a competitor use of the facility; 4) the monopolist can feasibly provide the facility; and 5) there is no legitimate business justification for denying access to the facility the monopolist controls. Although the second prong of this test resembles the inquiry the Commission must undertake to evaluate the availability of alternative elements outside of the incumbent LEC's network, it does not establish a standard by which the Commission can measure the extent to which the cost of duplicating the element is economically infeasible, which, as described below, is a significant part of the our unbundling analysis. (i) Factors for Determining Availability of Alternative Network Elements 74. In order to respond to the Supreme Court's decision, we consider whether a requesting carrier's ability to provide the services that it seeks to offer would be materially diminished if it were required to use an alternative element available outside the incumbent LEC's network. We agree with those parties that argue that we must consider the totality of the circumstances to determine whether an alternative to the incumbent LEC's network element is available in such a manner that a requesting carrier can realistically be expected to actually provide service using the alternative. We therefore take into account alternatives that are available through both self-provisioning and from third-party suppliers, and we consider the extent to which these alternatives are available as a practical, economic, and operational matter. 75. We are not persuaded by the incumbents' argument that we must look at each element in isolation to determine whether or not that element independently satisfies section 251(d)(2). Such an analysis fails to reflect the manner in which carriers interconnect their networks, and ignores factors that would impair a requesting carrier's ability to actually provide service, which is the focus of section 251(d)(2)(B). Even if a particular element may be purchased outside of the incumbent LEC's network at reasonable prices, other factors, including the costs and delays associated with collocation arrangements, as well as additional costs and operational impediments associated with the manual processes used to interconnect certain network elements, may make it impossible as a practical, economic, and operational matter for a competitor to provide services in the local market quickly and on a wide-spread basis. 76. We acknowledge that some of the factors we consider in our analysis may implicate other proceedings or provisions of the statute. We therefore remain open to the possibility that issues that we address under our "impair" analysis, (e.g., collocation), could be addressed in other contexts, such as in enforcement proceedings. 77. Although we recognize that the factors of cost, timeliness, quality, ubiquity, and operational factors are only some of the factors that may influence a carrier's decision to enter a particular market, we agree with the California PUC that these factors are relevant to an inquiry of whether alternative sources of network elements are reasonably available from other sources, and thus, in many cases, whether requesting carriers are able to actually provide service using the alternative element. We also agree with the commenters that point out that we cannot evaluate the needs of every potential entrant for every network element on a carrier- by-carrier, market-by-market, week-by-week (or other time period) basis. We therefore will not analyze the availability of alternative elements, including those provided through self-provisioning, from the perspective of a carrier using any specific competitive strategy in a particular geographic market. 78. Although we find it reasonable to consider cost, time, quality, ubiquity, and other factors associated with self-provisioning or acquiring an element from a third-party provider, we do not base our decision on cost models or on the theoretical availability of alternatives from other sources. Rather, we find the marketplace to be the most persuasive evidence of the actual availability of alternatives as a practical, economic, and operational matter. As the Texas PUC stated, the Commission and the states should "base their decisions on marketplace information, while recognizing that minor increases in a competitor's costs must be weighed against other factors such as service quality, technological innovation, and the loss of efficiency in a rapidly changing marketplace." Discerning the practical, economic, and operational viability of self-provisioning or obtaining alternative elements from third-party providers is technical, complex, and subject to considerable uncertainty. We believe, however, that an examination of the factors we have identified provides the Commission with the ability to identify, through the exercise of its administrative judgment, discernable material differences between using the incumbent's unbundled network elements and those available from other sources that ultimately will affect a requesting carrier's ability to provide the services it seeks to offer. 79. We assign little weight in our "impair" analysis to the ability of a requesting carrier to use the incumbent LECs' resold or retail tariffed services as alternatives to unbundled network elements. In the Local Competition First Report and Order, the Commission expressly rejected the incumbent LECs' argument that requesting carriers are not impaired in their ability to provide service if they can provide their proposed service by purchasing the service at wholesale rates from the incumbent LEC. As the Commission concluded in that Order, allowing incumbent LECs to deny access to unbundled elements solely, or primarily, on the grounds that an element is equivalent to a service available at resale would lead to impractical results; incumbent LECs could completely avoid section 251(c)(3)'s unbundling obligations by offering unbundled elements to end users as retail services. In other words, denying access to unbundled elements on the grounds that an incumbent LEC offers an equivalent retail service could force requesting carriers to purchase, for example, an unbundled loop and switching out of an incumbent's retail tariff at a wholesale discount, subject to all of the associated tariff restrictions. US West maintains that it need not unbundle local transport because requesting carriers can purchase its tariffed special access services. In light of the little weight we assign to the availability of resold services in our analysis, we reject US West's argument. This argument would foreclose competitive LECs from taking advantage of the distinct opportunity Congress gave them, through section 251(c)(3), to use unbundled network elements. 80. As the Commission explained in the Local Competition First Report and Order, using unbundled network elements and resold services present different opportunities, risks, and costs, in connection with providing local telephone service. These differences influence the entry strategies of potential competitors. The Commission stated that carriers using unbundled elements will have greater opportunities to offer services that are different from those services offered by the incumbents. More specifically, carriers reselling incumbent LEC services are limited to offering the same service an incumbent LEC offers at retail. While competitive LECs using unbundled elements may have greater competitive opportunities than carriers offering services available for resale, they also face greater risks. A carrier purchasing unbundled elements must pay for the cost of the element, pursuant to terms and conditions agreed to in negotiations or ordered by states in arbitrations. Thus, the competitive LEC faces the risk that end-user customers will not demand a sufficient number of services to allow the competitive LEC to recoup the costs it incurs using the unbundled element; a carrier that resells the incumbent LEC's services does not face the same risk. The 1996 Act grants competitive LECs the option of using either the incumbent LEC's unbundled network elements or resold services, thereby allowing the competitors to balance the risks and opportunities associated with each. 81. In addition, even if we agreed with US West that an incumbent LEC's retail tariff provided competitive LECs with a viable alternative to the incumbent LEC's unbundled network element, competitors would have no assurance that the incumbent LEC would not change the tariff in such a manner that the competitive LEC could no longer rely on it to provide the services it seeks to offer. Most services that competitive LECs purchase for resale are contained in state tariffs, and are subject to the states' tariff approval process. Relying on these state-approved tariffs would compromise our ability to determine which network elements must be unbundled pursuant to section 251(d)(2) because we would not be able to evaluate each incumbent LEC retail tariff as a possible alternative for every network element. In addition to being administratively unworkable for us to evaluate every state tariff filed by the incumbent LECs, relying on these tariffs as alternatives to the incumbent LECs' unbundled network elements would create inconsistent unbundling rules among the states, a result that, as we explain further below, would not promote the development of competition for all consumers. 82. Moreover, we do not find the Supreme Court's decision requiring us to consider the availability of elements outside the incumbent LECs' network to be at all inconsistent with our decision to consider alternatives available through self- provisioning or from third-party suppliers. The Supreme Court required us to compare the use of unbundled network elements with "self-provision, or with purchase from another provider." If we were to construe the Supreme Court's opinion in the manner suggested by US West, we would have to consider whether an incumbent LEC's duty to unbundle an element would be limited by the existence of an alternative service that the incumbent LEC provides itself, whether or not there are other competitively-supplied alternatives. In other words, under US West's argument, the existence of its retail tariffs alone would be sufficient to eliminate its obligation to unbundle certain elements. The Supreme Court's opinion does not require us to ignore whether there are other non-incumbent LEC alternatives to the incumbent LEC's unbundled network elements, proposed by US West. 83. We believe that the "impair" standard we adopt in this Order will encourage the development of facilities-based competition. Specifically, as competitors acquire more customers, and the material differences in cost, time, quality, and operational impediments diminish, competitors will gradually reduce their reliance on the incumbent LECs' facilities. Competitors will also deploy more of their own facilities as it becomes practical to do so. As the material differences decrease, the Commission will be able to apply the same standard to remove elements from the national unbundling obligations. 84. Cost. In addition to the direct cost of purchasing the element, we consider all of the costs that requesting carriers would incur using an alternative element to provide the services it seeks to offer. Although not dispositive, the costs associated with self- provisioning or purchasing alternative elements from third-party suppliers are relevant to our determination of whether the element is a practical and economical alternative to the incumbent LEC's unbundled network element. 85. We believe that an "impair" standard based on cost is more appropriate than a standard based on profitability, because profit margins for both new and existing carriers will depend on the degree of competition that exists in the market. If the cost of the alternative element is materially greater than the cost of obtaining the corresponding element from the incumbent, the requesting carrier will not be able to provide service at prices that are competitive with the incumbent's prevailing retail prices. 86. In determining whether the cost of self-provisioning or purchasing an element from a third-party source is materially higher than using the incumbent LEC's unbundled network element, we evaluate the difference between the cost to the requesting carrier of obtaining the unbundled element from the incumbent LEC at forward- looking costs and the cost of an alternative element. Because the Commission's rules require that network elements be priced based on forward-looking economic costs, we believe that forward-looking costs are the appropriate costs to consider in our analysis. 87. In order to provide service using its own facilities, a competitor will incur the costs of purchasing, installing, and provisioning the equipment it needs to provide service using its own loop or by interconnecting with the incumbent's network. The record in this proceeding addresses several types of costs associated with using an alternative element. These include the direct costs of provisioning the element, including fixed and sunk costs, as well as other costs that are likely to materially affect the requesting carrier's ability to provide the services it seeks to offer. "Fixed costs" are costs that do not vary with the level of output. A "sunk cost," on the other hand, is a cost that, once incurred, cannot be recouped if the firm ceases production. To the extent that a competitive LEC incurs significant fixed costs or sunk costs when it uses its own facilities or acquires facilities from a third party, these costs can disadvantage the competitor relative to the incumbent. 88. Fixed costs are frequently associated with economies of scale. Specifically, where a firm faces both a fixed cost and a constant or declining variable cost, the firm's average unit cost will fall as output increases, and the firm's cost structure is said to exhibit economies of scale. For example, the cost a competitive LEC incurs to construct its own fiber transport ring would constitute a fixed cost, because, at least in the short run, this cost would not vary as the competitive LEC's output changed. If a competitive LEC incurs significant fixed costs when it uses a particular facility, in its early stages of development it would have a significantly higher average unit cost than the incumbent LEC, which has a significantly larger output and customer base over which to spread the fixed cost. Since the Commission's rules require unbundled transport to be priced based on forward-looking costs (a form of long-run average incremental cost), leasing the incumbent's unbundled transport facilities is likely to be significantly less costly than deploying one's own transport facilities when the competitor has a relatively small volume of traffic, and thus its output would be small relative to that of the incumbent. 89. Certain network facilities also involve sunk costs, because the facilities cannot be easily re-deployed or sold should the competitor decide to cease offering service over those facilities. For example, the cost of the loop serving a customer's home is largely a sunk cost because it cannot be recovered if the carrier ceases serving the customer. It is generally recognized that the need to incur sunk costs can constitute a barrier to entry. Specifically, where an incumbent has already deployed sunk facilities to serve all customers, a competitive LEC may be unwilling to sink the costs of duplicative facilities, either because it may be unable to lure customers away from the incumbent and generate enough revenue to recover these sunk costs, or because resulting competition between itself and the incumbent LEC would drive prices so low that, even if the competitive LEC won a significant number of customers, it would still be unable to recover its sunk costs. In such situations, the incumbent has a "first mover" advantage. 90. The non-recurring costs of collocating equipment in the incumbent's end offices, including the costs of connecting the incumbent LEC's unbundled loops to the competitor's switch, and the fees required to obtain rights-of-ways, also constitute sunk costs. Unlike the costs associated with purchasing portable equipment, such as multiplexers or switches, the non-recurring costs incurred to collocate equipment and connect network elements to the competitive LEC's collocated equipment in an incumbent's central office are sunk costs and cannot be recovered if, for whatever reason, the carrier exits that market. 91. Additional costs, such as the costs a competitive LEC incurs to connect its own facilities to the incumbent LEC's unbundled network elements, affect the extent to which an alternative element is available as a practical and economic matter, such that a requesting carrier can actually use the element to provide the service it seeks to offer. For example, when a competitive LEC deploys its own switch but purchases the customer's unbundled loop from the incumbent, the competitive LEC may incur significant costs to connect the customer's loop, located in the incumbent LEC's central office, to its own switch. When these cutover costs are added to the costs of collocation, a competitor's ability to provide service in an efficient manner, when using its own switch for unbundled switching, could be materially diminished. We thus look at all of the costs a competitor must incur when using alternatives to the incumbent LEC's network element. 92. We find that significant fixed and sunk costs associated with using alternatives outside the incumbent LEC's network contribute to a finding that lack of access to the incumbent's unbundled network elements impairs the requesting carrier's ability to provide the service it seeks to offer. This is particularly true for a new competitive LEC that has few customers from which it can recover these costs. Because the per-customer costs decrease as the number of subscribers served by the carrier increases, a carrier must acquire a sufficient customer base if it is to recover substantial costs associated with deploying its own facilities. It is reasonable, therefore, that a competitive LEC, at a minimum, would want to serve a substantial number of business and/or residential customers within a particular Metropolitan Statistical Area (MSA). If the competitor must collocate its own switches in multiple central offices throughout the MSA in order to serve those customers, the costs associated with collocation may impair the competitor's ability to provide the services it seeks to offer, even if the cost of purchasing the individual equipment hardware is not excessive. 93. In addition, we find that the type of customers that a competitive LEC seeks to serve is relevant to our analysis of whether the cost of self-provisioning or acquiring an element from a third-party supplier impairs the ability of a requesting carrier to provide the services it seeks to offer. Section 251(d)(2)(B) requires us to consider whether lack of access to the incumbent LEC's network elements would impair the ability of the carrier to provide the services it seeks to offer. Consistent with the Act, we define the term "services" as it is used in section 251(d)(2)(B), to mean "telecommunications service," as it is defined in section 153(46) of the Act. Section 251(c)(3) of the Act places an affirmative duty on the incumbent LEC to provide unbundled elements to "any requesting telecommunications carrier for the provision of a telecommunications service." Section 251(d)(2)(B), in turn, requires that a requesting carrier should not have access to unbundled elements unless it would be impaired in its ability to provide "the services that it seeks to offer." Different types of customers use different services (e.g., large business customers order different services than residential customers). We therefore conclude that it is appropriate for us to consider the particular types of customers that the carrier seeks to serve. 94. Competitive LECs generally seek to provide service to residential and small business customers and/or to large business customers. The different revenue- generating potential of these different customer groups will often determine whether or not a competitive LEC can afford to incur the costs of self-provisioning a facility or of acquiring it from a third-party supplier, to the extent that it is available from a third-party provider. For example, a model submitted by MCI WorldCom that compares the costs of serving residential customers using unbundled elements from the incumbent LEC with the costs of serving the customers using its own facilities indicates that, at low market penetration levels, the costs of collocation would impair a competitive LEC's ability to serve residential customers using its own facilities. The model further demonstrates, however, that using the incumbent LEC's unbundled network elements, the entrant would be able provide service, even at the same low market-penetration levels. 95. Although the model submitted by MCI WorldCom is clearly not dispositive, we note it to illustrate that a requesting carrier's ability to serve residential and small business customers may be materially diminished without access to the incumbent LEC's network. Larger business customers, on the other hand, may generate sufficient revenue to allow the requesting carrier to serve the customer using certain self-provisioned facilities or facilities acquired from third-party sources. 96. We also consider, as part of our analysis, the economies of scale and scope that the incumbents have due to their ubiquitous network. The record demonstrates that, although facilities-based competition has developed in particular markets (primarily for large business customers in high-density areas), incumbent LECs continue to enjoy significant economies of scale and density not enjoyed by competitive LECs. Because these economies lower the incumbent's per-customer costs of providing service, vis-…-vis their competitors, we find these economies relevant to our inquiry of the extent to which costs of using alternative elements impair a requesting carrier's ability to provide the services it seeks to provide. 97. We are not persuaded by the argument of BellSouth and other incumbent LECs that we should not consider the impact of the incumbents' economies of scale because competitors are capable of matching or exceeding the incumbent LECs' economies by building their own facilities. The Commission has concluded previously that an incumbent LEC's existing infrastructure generally enables it to serve new customers at a much lower cost than a requesting carrier that must install its own switches, trunking, and loops to serve its customers, and that Congress has addressed this problem by mandating that incumbent LECs share their economies of scale and density with competitors. 98. We continue to believe that one important purpose of the unbundling provisions of the Act is to permit competitive LECs to compete with the same economies as the incumbents, especially in the early stages of local competition, when their networks are limited in their reach, and their customer bases are necessarily small. The incumbent LECs still enjoy cost advantages and superiority of economies of scale, scope, and ubiquity as a result of their historic, government-sanctioned monopolies. These economies are now critical competitive attributes and would belong unquestionably to the incumbent LECs if they had "earned" them by superior competitive skills. These advantages of economies, however, were obtained by the incumbents by virtue of their status as government-sanctioned and protected monopolies. We believe that these government-sanctioned advantages remain barriers to the requesting carriers' ability to provide a range of services to a wide array of customers, and that their existence justifies placing a duty on the incumbent carriers to share their network facilities. Indeed, Congress, in section 259 of the Act, recognized expressly the benefits that the incumbent LECs have as a result of their economies of scale and scope. Section 259 requires the Commission to ensure that incumbent LECs make their infrastructure available to qualifying carriers on terms and conditions that permit the qualifying carriers to "fully benefit from the economies of scale and scope of such [incumbent] local exchange carrier." Although section 259 of the Act is different from section 251 in that qualifying carriers obtaining infrastructure from the incumbent LEC pursuant to a section 259 agreement may not use such infrastructure to compete with the incumbent LEC in its service territory, both sections make the incumbent LECs' broad economies of scale and scope available to other carriers by requiring them to grant other carriers access to their networks. 99. We do not agree with Ameritech that competitive LECs are not impaired in their ability to provide service because they have cost efficiencies which the incumbent LECs do not have. Although we agree that competitors may have certain cost advantages, we find that these advantages are likely to be outweighed by other costs that competitive LECs, but not incumbent LECs, incur to provide service. For example, many competitive LECs are likely to incur higher costs than the incumbent LECs to attract customers, because unlike the incumbent, many competitive LECs must establish a brand name and develop a reputation for service quality before they can overcome the incumbents' long-standing relationships with their customers. Similarly, competitive LECs must incur the initial costs of setting up their operations and developing their back-office systems. AT&T also points out that new entrants face a high level of risk when they enter the local market, because they enter without the incumbent LEC's knowledge of local operating costs (e.g., location and quality of outside plant facilities) and consumer demand (e.g., peak traffic volumes over certain facilities and demand growth). 100. We recognize that a new entrant in many industries will face disadvantages arising from economies of scale. We further recognize that, even after competition in local telecommunications markets is well-established, and the Commission can eliminate certain unbundling requirements, smaller competitors will be at a disadvantage to the extent that incumbent LECs continue to enjoy significant economies of scale in the provision of local telephone service. Nonetheless, we believe that the existence of economies of scale, as well as sunk costs, are relevant factors to consider in our assessment of whether failure to provide access to a particular unbundled network element will impair a requesting carriers' ability to provide the services it seeks to offer. Although we find economies of scale to be a relevant factor in our analysis, we note that we are not basing our determination of whether competitive LECs are "impaired" within the meaning of section 251(d)(2) solely on the existence of scale economies, nor do we assume that the incumbent LEC's scale economies are insurmountable in all circumstances. 101. Timeliness. We also conclude that the time associated with using alternative elements is relevant to a determination of whether a requesting carrier would be impaired in its ability to provide the services it seeks to offer. A thorough evaluation of the delays associated with using alternative elements requires an analysis of both the start-up time required for a competitor to enter a market and serve a substantial number of customers in an MSA, as well as the time it would take a competitor that has already entered the market to expand its operations to serve more customers. We conclude that delays caused by the unavailability of unbundled network elements that exceed six months to one year may, taken together with other factors, materially diminish the ability of competitive LECs to provide the services that they seek to offer. 102. We recognize that the deployment of alternative elements, whether through self- provisioning facilities or by acquiring them from third-party suppliers, will require a reasonable amount of time. The delays associated with using alternative network elements will exist whether the requesting carrier is either just beginning to provision service or whether it is deploying additional facilities to expand its operations to serve more customers. Commenters differ in their opinions as to what constitutes a reasonable time to self-provision facilities. There is considerable evidence in the record, however, that indicates that it takes between six months and one year to engineer, furnish, and install a switch, including the time needed to obtain collocation space in the incumbent LEC's central offices where the switch will be connected to unbundled loops. Also, NTIA argues that we should consider as nontrivial any delay in service provisioning in excess of six months as compared to the time it would take for a competitive LEC to begin provisioning a service using an incumbent LEC's network element. 103. Based on the record before us, we conclude that it is reasonable to expect that a competitive LEC will need between six months and one year to provide service using a self-provisioned facility or one acquired from an alternative source. The local telecommunications market grows at an extremely rapid pace for many products and services. Indeed, we have reported that the demand for certain services has increased significantly from year to year since the passage of the 1996 Act and that we expect this trend to continue, particularly for advanced services. We believe that any delay that a competitive LEC experiences in serving this fast-paced, high-growth market can impair its ability to provide its desired services. Although we cannot quantify precisely how much of a delay associated with an alternative network element will materially diminish the ability of a competitor to provide its desired services, we find that delays that exceed six months to one year may, taken together with other factors, materially diminish the ability of a competitive LEC to provide the services it seeks to offer because it prevents the competitive LEC from responding quickly to the demand for its services in a rapidly changing market. Moreover, we agree with NTIA that incumbent LECs can take advantage of delays caused by the unavailability of unbundled network elements by using their "unique access to most customers to gain a foothold in new markets, and, in markets where services may be offered pursuant to long term-contracts (e.g., DSL and other advanced data services), to 'lock-up' customers in advance of competitive entry." 104. We disagree with Ameritech that a competitor is not impaired in its ability to provide a service if it can deploy alternative facilities within two years of its decision to do so. Congress made unbundled elements available to competitive LECs to avoid the time it would take competitive LECs to duplicate the incumbents' networks, thereby promoting the rapid development of competition for all consumers. We believe that requiring consumers to wait up to two years to have access to a choice of competitive service offerings, while competitors attempt to provide service without access to unbundled elements, is unreasonable and inconsistent with the objectives of the Act. 105. We also disagree with US West's claim that we should not consider the amount of time required for a competitive LEC to self-provision an element or acquire it from a third-party supplier because there are always inherent provisioning delays associated with using alternative elements. We believe the amount of time it takes a competitive LEC to self-provision an element or acquire an alternative from a third-party supplier is highly relevant to its ability to provide the services it seeks to offer. In particular, we agree with commenters that in order to compete effectively, competitive LECs must be able to initiate service promptly upon the request of their customers. We also agree with NTIA that delays in the introduction of competitive services caused by the unavailability of unbundled elements from the incumbent LEC would give the incumbent valuable time to entrench itself with existing customers. 106. Although we agree with US West that self-provisioning or acquiring alternative network elements from third-party suppliers involves normal delays incurred when starting or expanding a business, we find that significant delays will materially diminish a requesting carrier's ability to provide the services it seeks to offer. In addition, we have accounted for the inherent provisioning period to which US West refers by determining that it will take competitors approximately six months to one year to provide service, and that delays that exceed that time period would materially diminish a requesting carrier's ability to provide a competitive service. 107. We disagree with US West that it would be too administratively complex to consider the differences between the time it would take a competitor to obtain an element from the incumbent LEC and the time it would take to self-provision an element. We do not find it to be too administratively complex to consider whether a delay associated with using an alternative network element exceeds the six month to one year timeframe we identified above. As we stated above, the "impair" standard does not mandate precise quantification; nor does it involve an analysis of the delay suffered by every carrier. It requires instead a consideration of whether, as a general matter, there is an identifiable difference in the amount of time required to provide service using an alternative element and the amount of time required to provide service using the incumbent LEC's element, such that the delay would materially diminish the competitor's ability to provide the services that it seeks to offer. 108. Quality. We also conclude that the quality of alternative network elements available to the competitive LEC is relevant to a determination of whether a requesting carrier's ability to provide service is impaired. We agree with the California PUC and other commenters that a material degradation in service quality associated with using an alternative element will materially diminish a competitor's ability to effectively provide service. Examples of diminished quality presented in the record include greater dialtone delay, higher blocking rates, elevated noise on a telephone line and increased failure rates. These types of quality problems, all of which are recognizable by the end-user customer may, taken together with other factors, materially diminish the ability of the competitor to provide the services that it seeks to offer. In addition, we believe that the type of service a competitor seeks to provide is also relevant to the quality factor. For example, end users may be much less tolerant of problems that affect data services, than they would be for voice service. 109. Ubiquity. We conclude that we should also consider the extent to which the competitive LEC can serve customers ubiquitously using its own facilities or those acquired from third-party suppliers. We agree with competitive LECs that they may be impaired if lack of access to an unbundled element materially restricts the number or geographic scope of the customers they can serve. For example, incumbent LECs own 98 percent of all access lines in Texas and have deployed 1538 switches throughout the state. According to the Texas PUC, if a competitive carrier seeks to provide local telephone service throughout the state, it would be impractical, if not impossible, for the carrier to replicate the incumbents' networks. 110. Although we acknowledge that not all competitive LECs will want to provide ubiquitous service across broad geographic areas, those that do will likely be disadvantaged vis-…-vis the incumbent, especially in the early stages of deployment, because the incumbent LECs still enjoy advantages of a ubiquitous network that provide them with economies of scale and the ability to reach all consumers in their service territories. It is reasonable to expect that, in many cases, competitors would want to provide ubiquitous service in order to achieve similar economies of scale that will allow them to spread the costs of construction, equipment, and marketing across as many customers as possible. It is also reasonable to expect that in some cases, the ability to serve ubiquitously will be necessary to meet consumer demand for competitive alternatives in broad geographic areas. It such cases, lack of access to the incumbent's unbundled network elements could significantly thwart the competitor's ability to respond to consumer demand. Denying access to the incumbent's unbundled network elements, when use of alternative sources would materially diminish the competitors' ability to serve their intended geographic area, would be inconsistent with the goal of the 1996 Act to bring competition to the greatest number of customers. Indeed, the inability to provide service ubiquitously may be especially important for competitive LECs seeking to serve residential and small business customers located throughout a state. 111. Impact on Network Operation. We find that we should also consider how self- provisioning a network element or obtaining it from a third-party supplier may affect the technical manner in which the competitor can operate its network. We agree with the Washington Utilities Commission that overall network performance is an important consideration in our "impair" analysis. In order to compete with the incumbent, competitive LECs must be able to connect alternative elements either to their own networks or to other incumbent LEC elements that they use to provide service. Thus, material operational or technical differences in functionality that arise from interconnecting alternative elements may also impair a requesting carrier's ability to provide its desired services. 112. As we stated above, the incumbent LECs' relative advantages regarding costs, ubiquity, timeliness, and quality comprise only a part of a determination of whether or not a competitive LEC's ability to provide a competitive service is impaired. Indeed, as stated above, competitive LECs may have reasons for not entering a particular market that have nothing to do with whether lack of access to the incumbent's network would or would not impair their ability to offer service in that market. For example, it is likely that not all competitive LECs intend to invest in their own facilities to serve residential customers. Congress, however, clearly intended for competition to develop in these markets, as well as in the business markets, and we see as one of the primary goals of section 251, to facilitate competition in these markets. Because the ground work for competition is still uncharted, and we have seen very limited competition in the residential market to date, we seek to remove economic and other barriers that may forestall the development of competition for these consumers. Accordingly, we unbundle elements in a manner that we believe will have the desired effect of promoting competition in all markets as quickly as possible. (i) Other Factors to Be Weighed in Our Unbundling Analysis 113. We conclude that, in addition to the necessary" and "impair" standards, section 251(d)(2) permits us to consider other factors that are consistent with the objectives of the Act in making our unbundling determination. Section 251(d)(2) states that, "[i]n determining what network elements should be made available for purposes of subsection 251(c)(3), the Commission shall consider, at a minimum" the "necessary" and "impair" standards." This language implies clearly that other factors may be considered as long as we consider the "necessary" and "impair" standards. Moreover, as the D.C. Circuit has held, when Congress requires an agency to "consider" several listed factors, it may also consider additional factors in making its decision. For example, in Central Vermont Railway, Inc. v. Interstate Commerce Commission, the D.C. Circuit found that the language of a statute addressing railroad mergers that directed the Interstate Commerce Commission to "consider at least the following [factors]," also allowed the agency to consider factors other than those specifically listed. In a later case that cited Central Vermont Railway, the court explained that an agency's duty to "consider' specific factors means only that it must "reach an 'express and considered conclusion' about the bearing of a factor, but is not required to give 'any specific weight' to the factor." 114. In the Local Competition First Report and Order, the Commission stated that it agreed with several incumbent LECs that the plain import of the "at a minimum" language in section 251(d)(2) requires the Commission to consider the standards enumerated there, "as well as other standards we believe are consistent with the objectives of the 1996 Act." The Supreme Court did not dispute this determination. In fact, it directed us to adopt "some limiting standard rationally related to the goals of the Act." We are therefore not persuaded by the argument of the incumbent LECs that we may now require unbundling only where the "necessary" or "impair" standards have been met. If Congress had intended to require the incumbent LECs to unbundle an element only when it was "necessary" to, or would "impair" the requesting carrier's ability to provide its desired service, Congress would not have used the discretionary phrase "consider at a minimum." Rather, Congress would have required the Commission to apply the "necessary" and "impair" standard, without consideration of any additional factors. 115. Accordingly, in addition to the "necessary" and "impair" standard, we conclude that we may consider several factors, set out below, that further the goals of the Act in accordance with the Supreme Court's directive. Two fundamental goals of the Act are to open the local exchange and exchange access markets to competition and to promote innovation and investment by all participants in the telecommunications marketplace. To further the goal of opening the local market to competition, we may consider how access to specific unbundled network elements will encourage the rapid introduction of local competition to the benefit of the greatest number of consumers. 116. We may also consider how the unbundling rules we adopt will promote facilities- based competition by competitive LECs. We believe that it is the development of facilities-based competition that will provide both incumbent and competitive LECs with the incentives to innovate and invest in new technologies. Such innovation and investment will bring greater choices of telecommunication services and lower prices to a greater number of consumers. We may also consider the extent to which we can reduce regulatory obligations to provide access to network elements as alternatives to the incumbent LECs' network elements become available in the future. 117. We may further consider whether unbundling particular network elements will provide certainty in the market so that competitive LECs can attract investment capital and execute their business plans. We may also take into account how we can make the unbundling rules administratively manageable for the Commission and the states to apply. The adoption of administratively workable unbundling rules will enable the Commission and the states to implement and enforce such rules, thereby facilitating the ability of competitive LECs to enter the market as quickly and efficiently as possible. 118. We do not give particular weight to any of the factors we identify. Rather, we consider the relationship among the factors we take into account for a particular network element, and determine whether the sum total of the effect of the factors require a finding that the element must be unbundled. Thus, we do not require that all of the factors be met before we decide whether or not to require incumbent LECs to unbundle a particular network element. Indeed, there may be circumstances in which there is significant evidence that competitors are impaired without unbundled access to a particular element, but that unbundling the element would not further the goals of the Act. In the final analysis, as we explain in more detail below, we consider the effect of these factors in order to develop unbundling obligations that are most consistent with Congressional intent. 119. Rapid Introduction of Competition in All Markets. Congress has emphasized that a major goal of the 1996 Act is to accelerate the development of local competition. Indeed, the preamble to the Act states that it provides a "pro-competitive, de- regulatory national policy framework designed to accelerate rapidly" deployment of advanced telecommunications technologies by opening all markets to competition. With regard to unbundled network elements, in particular, the Eighth Circuit Court of Appeals found that the use of unbundled elements promotes the prompt development of competition, as intended by the Act. The court stated that the Act "provides for unbundled access to incumbent LECs' network elements as a way to jumpstart competition in the local telecommunications industry." We therefore find that we may consider whether an unbundling obligation is likely to encourage requesting carriers to rapidly enter the local market and serve the greatest number of customers. Conversely, we may also consider whether the failure to require unbundling will cause any class of consumers to wait unnecessarily for competitive alternatives. 120. We also note that Congress specified certain network elements in the section 271 checklist that BOCs are required to unbundle before they obtain in-region interLATA relief. In particular, the checklist requires BOCs to demonstrate that they are providing loops, switching, transport, signaling and databases, and operator services/directory assistance. Accordingly, we may consider whether requiring all incumbent LECs to unbundle these same elements would promote the rapid introduction of competition on a nationwide basis. 121. We agree with NTIA that there is a common purpose between sections 251 and 271 of the Act of opening the incumbents' monopoly local exchange networks to competition. We believe that Congress intended section 251(c)(3) of the Act and the competitive checklist to contain similar, if not identical, obligations. Although we do not conclude that the checklist determines definitively that all incumbent LECs are required, pursuant to section 251, to unbundle the items enumerated in section 271, we find that section 271 sheds some light on what elements Congress believed should be unbundled in order to open local markets to competition. We may therefore consider whether an element is among the elements identified in the competitive checklist as we make our determination of which network elements incumbent LECs must provide on an unbundled basis. 122. Promotion of Facilities-Based Competition, Investment, and Innovation. A fundamental goal of the Act is to promote investment and innovation by all participants in the telecommunications marketplace, and, in particular, to encourage rapid deployment of new telecommunications technologies. As the Commission has stated, the construction of new local exchange networks "will not only lead to innovation by the new competitors, but should also spur [the incumbent LECs] to upgrade their systems and offer a broader array of desired service options to meet consumers' demands." By promoting innovation both by the incumbent LECs and competitive LECs, the Act enables these carriers to produce innovative new services for consumers. Specifically, consumers benefit when carriers invest in their own facilities because such carriers can exercise greater control over their networks, thereby promoting the availability of new products that differentiate their services in terms of price and quality. We may therefore consider the extent to which the unbundling obligations we adopt will advance the development of facilities-based competition and will encourage innovation by both incumbent and competitive LECs. 123. We seek to adopt unbundling requirements that are broad enough to provide requesting carriers with the elements they need to ramp up towards facilities deployment. At the same time, we remain cognizant of the Supreme Court's mandate against granting blanket access to the incumbents' network in a manner that is inconsistent with the "necessary" and "impair" standards of section 251(d)(2), or with the goals of the 1996 Act. We believe that the standards we articulate in this Order will strike the appropriate balance by unbundling only those network elements without which a competitive LEC's ability to provide service will be materially diminished. 124. We agree with the competitive LECs that argue that unbundled access to certain incumbents' network elements will accelerate initially competitors' development of alternative networks because it will allow them to acquire sufficient customers and the necessary market information to justify the construction of new facilities. Indeed, many commenters in this proceeding emphasize that they plan to deploy alternative facilities as soon as it is technically and economically possible to do so at a cost that is close to the incumbent LECs' prices for network elements. According to these commenters, competitive LECs prefer to use their own facilities or alternatives outside of the incumbent's network when they are able to do so, in order to reduce their reliance on a primary competitor. We find this explanation to be reasonable. Use of the incumbent LEC's network elements requires competitive LECs to disclose details about their customers to their chief competitor. Moreover, it is reasonable to expect that competitive LECs would prefer to have direct control of their networks to ensure the quality of their service and to offer products and pricing packages that differentiate their services from the perspective of end users. 125. Reduced Regulation. Another goal of the Act is to deregulate where market conditions warrant. We may therefore consider the extent to which we can reduce regulatory obligations to provide access to network elements as alternatives to the incumbent LECs' network elements become available in the future. 126. Certainty in the Market. Among other things, the Act seeks to promote competition by eliminating barriers to entry into the local market. We may therefore consider how the unbundling obligations we adopt in this Order facilitate competitive entry. Accordingly, we find that the unbundling requirements we adopt should typically provide the uniformity and predictability new entrants and fledgling competitors need to develop and implement national and regional business plans. In addition, uniform and predictable unbundling rules will provide financial markets with reasonable certainty so that competitive LECs can attract the investment capital they need to execute their business plans. Specifically, uniform and predictable unbundling rules reduce substantially competitive LECs' risk of underutilized investment or cash flow drain by providing financial markets with some certainty that the competitors will be able to execute their business plans. 127. We also find that we should, whenever possible, adopt unbundling obligations that can be included easily in interconnection agreements between the incumbents and the competitive LECs, with as little risk of subsequent litigation as possible. Litigation over the incumbents' unbundling obligations requires the parties to these agreements, and the state commissions that approve them, to expend vast amounts of time and resources, ultimately impairing the ability of competitive LECs to execute their business plans. 128. Administrative Practicality . We may also consider whether the unbundling rules we adopt are administratively practical to apply. Any rule adopted in an administrative proceeding runs the risk of being potentially overinclusive in some situations and under-inclusive in other situations. A rule of general applicability rarely will neatly fit all situations. Nonetheless, administrative agencies are entitled to proceed by rulemaking as well as by adjudication. In addition, the goal of administrative efficiency has widespread support from diverse segments of the industry, even where they disagree on the substantive outcome of the proceeding. We therefore seek to adopt unbundling rules that provide for administrative ease in addressing the incumbents' unbundling obligations today, as well as in the future, as alternatives to incumbent LEC network elements become available. We believe that adopting rules that are administratively practical to apply will also enhance certainty in the marketplace by allowing us to apply the rules efficiently to respond to changes in the marketplace. A. Adoption of a National List of Unbundled Network Elements 1. Background 129. In the Local Competition First Report and Order, the Commission concluded that identifying a specific list of network elements that must be unbundled, applicable in all states and territories, would best further the "national policy framework" Congress established to promote competition in local markets. In particular, the Commission found that a national list would: (1) allow requesting carriers, including small entities, to take advantage of economies of scale; (2) provide financial markets with greater certainty in assessing requesting carrier's business plans; (3) facilitate the states' ability to conduct arbitrations; and (4) reduce the likelihood of litigation regarding the requirements of section 251(c)(3). 130. In the Notice, we stated that we found nothing in the Supreme Court's decision that would require us to eliminate national unbundling requirements. We tentatively concluded that we should continue to identify a minimum set of network elements that must be unbundled on a nationwide basis, and sought comment on this conclusion. We also sought comment on whether the existence of geographic variations in the availability of elements outside of the incumbent LEC's network is relevant to a decision to impose minimum national unbundling requirements. 131. Nearly all of the state commissions commenting in this proceeding, and all of the competitive LECs, assert that we should adopt a national list of unbundled elements. The state commissions agree that the Commission has authority to adopt such a list, and that it should implement a process for the states to modify the list in the future, based on conditions that exist in a particular state. The New York Commission also proposes that, in establishing the national list, we should evaluate whether to exclude an element from the unbundling obligations in discrete market areas where commercially viable alternatives are available. The incumbent LECs argue, on the other hand, that the Supreme Court's decision in Iowa Utils. Bd. requires a geographic market-by-market analysis that will ultimately not result in a national list of unbundled elements. These carriers propose that the Commission adopt national standards to be applied by state commissions on a market-by-market basis. 1. Discussion 132. We adopt our tentative conclusion to identify a minimum list of network elements that should be unbundled on a national basis. Similar to New York's proposal, we also conclude, as explained below, that we must apply discrete geographic and product market exceptions to the incumbent's duty to unbundle the elements on the national list, where appropriate. We conclude that the Commission has the legal authority to adopt a national list of network elements that must be made available on an unbundled basis, and that the other factors we identify above, such as rapid introduction of competition, certainty in the marketplace, administrative practicality, and promotion of facilities-based competition, can only be furthered by adoption of a national list. a. Legal Authority 133. The Supreme Court decision in Iowa Utils. Bd., the statutory language of section 251(d)(2), and the legislative history of the 1996 Act support our authority to develop a national list of unbundled elements. In particular, the Supreme Court upheld explicitly the Commission's jurisdiction to adopt minimum national rules to implement each subsection of the 1996 Act. Consistent with the language in the statute, the Supreme Court stated that section 251(d)(2) ". . . requires the Commission to determine on a rational basis which network elements must be made available, taking into account the objectives of the Act and giving some substance to the 'necessary' and 'impair' requirements." The Court stated that some of the national unbundling rules the Commission adopted originally in the Local Competition First Report and Order might have been supported by the standard required by section 251(d)(2). The Court stated however, that because the standard was not consistently applied, it was forced to vacate Rule 319. As explained above, we have adopted a limiting standard that we believe responds to the Supreme Court's concerns. We have also applied the standard consistently to derive a list of network elements that must be made available on an unbundled basis nationwide. 134. In addition, we do not find that the Supreme Court decision in Iowa Utils. Bd. requires us to determine, on a localized state-by-state or market-by-market basis which unbundled elements are to be made available. The Commission examined the conditions in the nation as a whole to determine, in the Local Competition First Report and Order, that the incumbent LECs must make available a minimum list of elements. The Commission also concluded that it would not adopt an exhaustive list of elements, but that the states would identify additional unbundling obligations based on local market conditions. The Supreme Court did not take issue with this determination. The Court held that the Commission must determine on a rational basis which network elements must be made available, taking into account "the objectives of the Act and giving some substance to the 'necessary' and 'impair' requirements of section 251(d)(2)." Although this language permits the Commission to undertake a market-by-market assessment of alternatives, it plainly does not mandate such an approach. Rather, it provides the Commission with the discretion to look at the nation as a whole and to identify differences in the availability of alternatives outside of the incumbent's network that may exist in discrete geographic areas. 135. However the Commission chooses to limit the incumbent LEC's duty to unbundle in accordance with the Supreme Court's opinion, Congress has charged the Commission in section 251(d)(2) with "determining what network elements should be made available for purposes of subsection [251](c)(3)." We thus have the authority to identify a minimum list of network elements that must be unbundled on a nationwide basis. In addition, the legislative history indicates that Congress specifically contemplated that the Commission would open the last monopoly bottleneck strongholds in telecommunications by requiring incumbents to share their local exchange facilities, including "the equipment with capabilities of routing and signaling calls, network capacity, and network standards." This legislative history indicates that Congress expected the Commission would identify a national list of unbundled network elements that would include, at a minimum, these basic network elements. a. Goals of the Act 136. We find that adoption of a national list of unbundled network elements furthers the statutory purpose and design of section 251(d)(2) to provide competitive LECs with access to unbundled network elements that will allow them to provide the services they seek to offer. Moreover, we find that adoption of a national list is supported by the factors we identify above as being important to further the fundamental goals of the Act. 137. Rapid Introduction of Competition. We find that a national list of unbundled elements will encourage the rapid introduction of competition in the greatest number of markets because it will provide competitive LECs with certainty regarding the availability of network elements. In fact, the record reflects that many competitive LECs are poised to begin providing service using unbundled elements, particularly for residential and small business customers, as soon as the elements are available with a reasonable degree of certainty. Thus, we believe that the certainty that adoption of a national list will bring to the market will benefit the greatest number of consumers, particularly residential and small business customers. 138. We agree with AT&T that the lack of nationwide access to unbundled elements will hinder mass market competition during the time it would take competitive LECs to construct alternative networks capable of serving all residential customers and most business customers in a community. Even in areas where competitors are able to provide facilities-based service in specific wire centers, their ability to provide service on an MSA, LATA, or state-wide basis, for all classes of customers, is impaired without access to the incumbent's elements on a broader basis. A national list of unbundled elements will allow requesting carriers to enter local markets in a manner that will allow them to approach the incumbent LECs' historic economies of scale, scope, and ubiquity, thereby promoting rapid competition for all customers, including residential and small business customers, in all areas of the country. 139. According to the FCC Local Competition Report, competitors provide only about 1.8 percent of local services to end users. The record in this proceeding indicates that requesting carriers have not yet been able to obtain unbundled elements on a wide-spread basis nationwide, which may have prevented competitive LECs from serving a greater number of end users. For example, only recently has unbundled switching been made available in combination with other unbundled network elements in certain states. MCI WorldCom observes that, with the availability of unbundled switching in New York, it has been able to provide local service to upwards of 60,000 residential customers since January, 1999. We believe that by re-establishing a national list, with certain geographic and product market exceptions that are consistent with the standards of section 251(d)(2), we will best promote efficient, rapid, and widespread entry by carriers using unbundled network elements. Competitive market entry and service expansion on a widespread basis is a necessary precondition to construction of self-provisioned facilities. 140. Moreover, as the Illinois Commerce Commission; California PUC, and Connecticut Department of Public Utility Control all assert, a national list will allow competition to proceed quickly because it will reduce the number of issues that the states must address in upcoming arbitrations under section 252(b) of the Act. This is significant because many states will be conducting arbitrations and reviewing interconnection agreements as the initial agreements that they approved in 1996 and 1997 begin to expire. 141. We are not persuaded by Ameritech's argument that adoption of national standards containing bright-line tests, as opposed to a national list of unbundled elements, would facilitate arbitrations. Using the bright-line test proposed by Ameritech is inappropriate because the test does not allow us to consider the totality of the circumstances to determine whether alternative elements are actually available as a practical, economic, and operational matter. Moreover, the resources and time that requesting carriers would be required to devote to individual regulatory proceedings designed to determine if the bright-line criteria had been met in every market would delay the introduction of competition. The outcomes of each proceeding would likely vary across the country, thereby making it more difficult for competing carriers to execute reasonably uniform national or regional business plans. We believe that a national list of elements will better allow carriers to enter the market and to expand their businesses as rapidly as possible. 142. As explained below, we will revisit our unbundling rules in three years. Although we recognize that due to changes in the market and new technologies, the national list will likely be modified over time, we do not find that we should delay the onset of meaningful competition while we require the incumbent LECs and the competitors to produce voluminous amounts of data and participate in multiple proceedings to determine whether alternatives to the incumbent's network are available and being used in every market. We believe that a national list (that accounts for discrete geographic and product market exceptions) that can be applied at this time, with the least amount of regulatory involvement, will allow carriers to deploy resources to provide service to the greatest number of consumers instead of conducting regulatory proceedings. 143. We note that we established recently collocation-based triggers to determine when it would be appropriate to grant incumbent LECs pricing flexibility for certain interstate access services based on the existence of competition for those services. In the Pricing Flexibility Order, we stated that the triggers we adopted were policy determinations based on our agency expertise and our interpretation of the record before us in that proceeding. We acknowledged, however, that the use of triggers to measure competition precisely is not an exact science, particularly because we lack verifiable data from competitors concerning the deployment of their facilities. Given this constraint, and our desire not to impose heavy administrative burdens on the industry or conduct protracted proceedings to determine the extent of competition, we devised pricing flexibility triggers based on "objectively measurable criteria," such as the number of collocation arrangements in a given wire center. We found that it is appropriate to give incumbent LECs pricing flexibility when competitors have made an irreversible, sunk investment in facilities, and that collocation by competitors in incumbent LEC wire centers is a reliable indication of sunk investment by competitors. Specifically, to obtain pricing flexibility, we required incumbent LECs to show that "at least one competitor relies on transport facilities provided by a transport provider other than the incumbent at each wire center listed in the incumbent's pricing flexibility petition as the site of an operational collocation arrangement." 144. It is not appropriate to use these types of triggers to determine whether alternative sources of network elements are actually available as a practical, economic, and operational matter. As we explain above, the ability of one competitor to serve certain customers in a particular market is not indicative of whether, without unbundled access to the incumbent LEC's facilities, competitive LECs could provide service to other customers in the same market or to customers in other markets. While the triggers we adopted in the Pricing Flexibility Order allow us to determine when an incumbent LEC can re-price its services to respond to competition, they do not allow us to evaluate whether the incumbent LEC can withhold access to the inputs that requesting carriers need to provide competitive services in the first place. In order to undertake this evaluation, we must consider the cost, timeliness, quality, ubiquity and operational characteristics of alternative elements. As we explain above, discerning the practical, economic, and operational viability of these alternatives is technical, complex, and subject to considerable uncertainty. Based on the record before us, we do not believe that we can develop reliable triggers based on objectively measurable criteria to make this determination. In particular, the administrative difficulty associated with developing triggers that capture the cost, timeliness, quality, ubiquity, and operational factors of alternatives in every wire center throughout an incumbent LEC's service territory requires us to reject such an approach. Indeed, the Commission chose precisely to adopt triggers in the Pricing Flexibility Order, because we found that they were administratively easy to apply. Conversely, it would not be administratively easy to apply triggers to determine which network elements the incumbent LECs must unbundle. Moreover, the use of triggers also does not allow us to evaluate whether the unbundling obligations we adopt are consistent with the goals of the Act, as the Supreme Court has required us to do. 145. Moreover, a national list of unbundled network elements will facilitate the introduction of rapid competition by eliminating needless litigation that would result from unbundling requirements that differ in every market. Such litigation would require incumbent LECs, competitive LECs, and the state commissions to expend considerable time and resources to litigate issues surrounding whether a particular unbundled network element should be available to individual carriers seeking to serve specific customers or specific areas of the state. Although there has been significant litigation over the past three years regarding the incumbent's duty to unbundle elements under section 251(c)(3), we believe that re-establishing a national list, subject to the Supreme Court's mandate to include a rational limiting standard, will reduce the likelihood of further litigation and its accompanying delays and costs, in all fifty states. 146. Promotion of Facilities-Based Competition, Investment, and Innovation. We find that adoption of a national list will facilitate the deployment by competitors of their own facilities. Permitting competitors to obtain access to unbundled elements on a broad basis will allow these carriers to acquire sufficient customers and essential market information to enable them to determine whether construction of new facilities is justified. We believe that it is through self-provisioning their own facilities that competitive LECs will have a greater ability to serve all classes of customers. 147. Ameritech claims that the Commission "dismissed outright" the principal goal of the 1996 Act to encourage new investment and innovation by all competitors in the market when it adopted national unbundling rules. According to Ameritech, the national unbundling rules adopted in 1996 protected inefficient competitors and discouraged efficient entrants from investing and innovating in telecommunications services as the Act intended. Based on the incumbents' own evidence, we find this argument lacking in credibility. 148. The incumbent LECs have submitted a market study in this proceeding, the USTA UNE Report, that details the competitive LECs' investment in their own facilities on an element-by-element basis since the passage of the 1996 Act, and during the time that the Commission's national unbundling rules have been in effect. Although the Commission's unbundling rules have been the subject of extensive litigation, none of the parties dispute that competitors have used unbundled elements, particularly unbundled loops and transport, where these elements have been made available. Yet, the incumbents' UNE Report shows that competitors have built nearly 30,000 miles of fiber within the top 50 MSAs, serving nearly 15 percent of all commercial office buildings. 149. The USTA UNE Report also states that competitors have deployed approximately 700 switches to serve medium and large business customers. The report indicates that these carriers have deployed fixed wireless connections to extend their fiber networks out to many more customers. The incumbents also assert that many competitors are providing advanced services by attaching their own facilities to the incumbent LEC's unbundled copper loops. Overall, the incumbents estimate that competitive LECs are offering service over approximately 2.5 million facilities- based lines in the incumbents' service territories. As explained more fully below, these facilities are still not available broadly enough to prevent competitive LECs, in most cases, from being impaired in their ability to provide service without access to the incumbent's network. Nonetheless, the data presented by the incumbents shows significant and growing investment by the competitive LECs. Accordingly, we find no merit in the claim made by Ameritech and other incumbent LECs that unbundling elements will impede the Act's goal of encouraging new investment and innovation in telecommunications services. 150. The incumbents also claim that national unbundling requirements will discourage them from investing and innovating, particularly if they have to unbundle elements for the provision of advanced services. While we desire to do nothing to discourage investment and innovation by all carriers, we note that the Commission's national unbundling policy has clearly not discouraged incumbent LECs from seeking to serve new markets. Although in the Local Competition First Report and Order, the Commission did not order unbundling of certain equipment used in providing advanced services, it made clear that the states could extend the incumbents' unbundling obligations as necessary to account for changes in technology and to address local conditions. Incumbent LECs have therefore known since 1996 that they might eventually be required to unbundle elements used to provide advanced services. Moreover, last year, in the Advanced Services Order and NPRM, we sought specifically comment on whether to unbundle facilities used to provide advanced services. Notwithstanding the fact that the incumbents have been on notice that they could be required to unbundle facilities used to provide advanced services, the incumbents have announced aggressive rollout plans for xDSL service. In fact, a recent financial analyst's report indicates that advanced data services currently comprise an average of 9.9 percent of the revenues of the BOCs and GTE. Although the incumbents claim that competitors have deployed more advanced services equipment than the incumbents have deployed, they nevertheless acknowledge that the incumbent LECs are offering advanced services in 7 of the 10 largest MSAs and in 22 of the top 50 MSAs. We find these statistics to be significant because they demonstrate that the development of competition, and the threat of losing revenue and customers to carriers offering advanced services, provides a powerful incentive for carriers to invest. 151. We therefore conclude, as the Commission did in the Local Competition First Report and Order, that by adopting a national list of elements, and by giving the states the flexibility to add elements as technology and local market conditions change, we will not discourage incumbent LECs from investing and deploying innovative services. The incumbent LECs will have an increased incentive to reduce their operating and capital costs and to introduce new and innovative services that will increase the overall usage level of their networks as they face competition for all of their services. Moreover, the Commission's pricing methodology includes a risk-adjusted return on capital and economic depreciation for the incumbent as part of the forward-looking rate. As we indicated above, we are also adopting a "necessary" standard that fully protects the incumbents' intellectual property associated with proprietary network elements when those elements are used by the incumbent to differentiate its products from those of its competitors. We therefore do not find merit in arguments that the adoption of a list of network elements that must be unbundled nationwide will discourage innovation and investment by incumbent or competitive LECs. 152. Certainty in the Marketplace. We find that a national list of unbundled elements will provide uniformity and predictability that will facilitate the development and implementation of national and regional business plans by competitive LECs, thereby extending the benefits of competition for the greatest number of consumers. We agree with the California PUC that a national list will allow multi-state competitors to create a national business plan with the knowledge that a set of network elements will be available in all states. Indeed, we find that the unavailability of elements on a nationwide basis would jeopardize the usefulness of unbundled elements as a method of serving the maximum number of consumers. 153. We also continue to believe that national unbundling requirements will provide financial markets with greater certainty regarding the elements that are available to competitive LECs. Such certainty should reduce the risk of entry, thereby making more capital available at less cost to new entrants and fledgling competitors. We do not agree with Ameritech that a national list would perpetuate uncertainty in capital markets because carriers would challenge the list regardless of what elements it contains. As stated above, we believe that a national list will actually reduce the risk of litigation. 154. Administrative Practicality. We find that a national list of unbundled elements is administratively easier for the Commission, the states, and the industry to apply than a list that varies on a state-by-state or market-by-market basis. As we stated in the Notice, application of the "necessary" and "impair" standard is fact-intensive. Determining the availability of practical alternatives to the incumbents' network elements on a market-by-market basis, even through the use of bright-line tests as proposed by the incumbent LECs, would potentially require the Commission or the states to analyze the availability of alternatives in almost every wire center. In addition to creating uncertainty in the market, such a proposal would consume enormous amounts of resources and time, thereby undermining the goal of the Act to bring the benefits of rapid competition to all consumers. Such an approach would also require a new analysis each time a new carrier sought to initiate service in a particular market, and would likely lead to additional litigation by adversely affected carriers. We do not believe that Congress or the Supreme Court had in mind the adoption of a procedure that would impose such an undue and unworkable administrative burden on the Commission, the states, or the industry. 155. Reduced Regulation. We believe that a national list of elements that contains discrete geographic and product market exceptions will result immediately in reduced regulation. Moreover, a national framework under which elements can be removed from the national list is consistent with the deregulatory goals of the Act. Reduced regulation will occur as we remove elements from the list as requesting carriers are no longer impaired without access to those elements, and it otherwise does not further the goals of the Act to continue requiring incumbent LECs to unbundle them. A. Modification of the National List 1. Background 156. In the Local Competition Order First Report and Order, the Commission acknowledged that the rapid pace and ever-changing nature of technological advancement in the telecommunications industry made it essential that the Commission retain the ability to revise the rules as circumstances change. The Commission noted that, absent such ability, its rules might impede technological change and frustrate the 1996 Act's overriding goal of bringing the benefits of competition to consumers of local phone service. Accordingly, the Commission determined that, in addition to identifying unbundled network elements that incumbent LECs were required to make available at the time the original rules were adopted, it had the authority to identify additional or different unbundling requirements that would apply to incumbent LECs in the future. 157. In the Local Competition Order First Report and Order, the Commission also determined that state commissions could impose additional unbundling requirements, as long as the requirements were consistent with the 1996 Act and our regulations. The Commission codified this grant of authority in section 51.317 of its rules. The Commission believed that the states' authority to impose additional requirements, combined with its ability to modify the national unbundling rules, provided the necessary flexibility to accommodate any truly unique conditions that might exist. The Commission, however, did not address the issue of whether states could remove elements from the national list. 158. In the Notice we sought comment on whether the Commission should adopt an approach that would allow sunset or modification of the section 253(c)(3) unbundling obligations as technology and market conditions evolve over time. We noted that, under our rules, states have the authority to impose additional unbundling requirements. We sought comment on whether section 251(d)(2), or any other provision of the Act, provides the Commission with the authority to delegate to the states the responsibility of removing network elements from any national requirement. We sought comment on proposals for a mechanism for removal, including which party should bear the burden of proof. We asked whether the Commission should consider a phase-out period for network elements removed from the national list. Further, we asked whether we should institute a period of time during which incumbents could not seek removal of network elements from our new unbundling rules. We also asked whether we could adopt a "sunset" provision. 159. Several of the state commissions argue that they have the authority to add and subtract elements from the national list, while the Vermont and Illinois state commissions argue that the Commission should establish a set of unbundling obligations to which the state may add additional unbundling obligations. BellSouth argues that states should be able to add or remove unbundled elements in a particular zone. SBC and GTE oppose allowing the states to add or subtract elements. US West argues that states should be able to determine whether network elements no longer need to be unbundled, but that they not be allowed to add network elements. The vast majority of competitive LECs that commented in this proceeding, as well as NTIA and ALTS, argue that the states should be allowed to add, but not to remove, elements from the national list. 1. Discussion a. Modification of the National List by the Commission 160. As discussed above, section 251(d)(2) grants the Commission authority to establish a national list of network elements that are subject to the unbundling requirements of the Act. Given the rapid changes in technology, competition, and the economic conditions of the telecommunications market, we expect that the list of unbundled network elements that meets the standards of section 251(d)(2) will change over time. We therefore agree with commenters that we will need to reevaluate our national rules periodically. 161. The need to reassess periodically the availability of elements outside the incumbent's network is borne out by the changes that have taken place since we first adopted our unbundling rules three years ago. For example, the evidence in this proceeding indicates that competition is developing in some geographic markets for certain customer groups, (e.g., medium and large businesses in major metropolitan areas). Only by periodically reevaluating the availability of alternative network elements outside the incumbent's network can we truly determine whether the incumbent's network should be unbundled in order to meet the requirements of section 251 and the goals of the Act. We therefore conclude that as market conditions change and new technologies develop, we will periodically revisit the issue of what elements are subject to the unbundling obligations of the Act. 162. Although we will periodically revisit our unbundling rules, we believe that it would be inconsistent with our overall policy goals to consider petitions to remove elements from the national list immediately upon adoption of this order. Specifically, as discussed above, the rules we adopt today seek to provide a measure of certainty to ensure that new entrants and fledgling competitors can design networks, attract investment capital, and have sufficient time to attempt to implement their business plans. Entertaining, on an ad hoc basis, numerous petitions to remove elements from the list, either generally or in particular circumstances, would threaten the certainty that we believe is necessary to bring rapid competition to the greatest number of consumers. In addition, entertaining numerous petitions on an ad hoc basis would undermine the goal of implementing unbundling rules that are administratively practical to apply. 163. We expect to reexamine our national list of network elements that are subject to the unbundling obligations of the Act every three years. We note that many of the first interconnection agreements negotiated in 1996 are now approaching expiration of their typical three-year terms and will be eligible for renewal. We expect parties to implement the requirements of this Order as they negotiate new interconnection agreements. We find that a similar three-year time frame for reevaluating the unbundling obligations is warranted to provide competitors with reasonable certainty for a period of time that is sufficient time to implement their plans. Revisiting our rules in three years should provide sufficient certainty to the carriers and capital markets and should provide carriers with sufficient time to implement their plans. 164. We decline to adopt a rule mandating that elements will not be subject to unbundling after a date certain in the future. Several parties have suggested that it would be extremely difficult for us to predict a date at which a particular network element would no longer meet the "necessary" and "impair" standards of section 251(d)(2). As noted by the Illinois Commission, in the three years since the Act was implemented, no BOC has demonstrated that it satisfies the competitive checklist in section 271. In 1996, few would have expected that three years later BOCs would not have qualified for section 271 approval. This suggests that it would be similarly very difficult for us to predict, at this time, the date at which incumbent network elements would no longer be subject to unbundling obligations under section 251. Moreover, we note that we find no basis in the record before us to make predictive judgments about when an unbundling standard will no longer be met for particular network elements. Thus, at this point in time, we do not have enough information and experience to determine what events would lead to an automatic sunset of one of our unbundling requirements. Accordingly, at this time, we decline to adopt a sunset provision for removing network elements from the national list adopted in this Order. a. Modification of the National List by the States 165. We agree with commenters that section 251(d)(3) provides state commissions with the ability to establish additional unbundling obligations, as long as the obligations comply with subsections 251(d)(3)(B) and (C). Section 251(d)(3) states that: In prescribing and enforcing regulations to implement the requirements of this section, the Commission shall not preclude the enforcement of any regulation, order, or policy of a State commission that (A) establishes access and interconnection obligations of local exchange carriers; (B) is consistent with the requirements of this section; and (C) does not substantially prevent implementation of the requirements of this section and the purposes of this part. 166. This section of the statute allows state commissions to establish access obligations of local exchange carriers that are consistent with our rules implementing section 251. We believe that section 251(d)(3) grants state commissions the authority to impose additional obligations upon incumbent LECs beyond those imposed by the national list, as long as they meet the requirements of section 251 and the national policy framework instituted in this Order. As explained below however, we find that state-by-state removal of elements from the national list would substantially prevent implementation of the requirements and purposes of this section of the Act. 167. Section 51.317 of the Commission's rules codifies the standards state commissions must apply to add elements to the national list of network elements we adopt in this Order. In its current form, Rule 51.317 reflects the Commission's interpretation of the necessary and impair standards adopted in the Local Competition First Report and Order. Inasmuch as we have modified the "necessary" and "impair" standard to respond to the Supreme Court's directive, we must also amend Rule 51.317 to reflect the new standards. Accordingly, we modify Rule 51.317, to bring it into compliance with our new standards and the Supreme Court's decision. Modification of this rule will enable state commissions to add additional unbundling obligations consistent with sections 251(d)(3)(B) and(C) of the Act. 168. We agree with the California PUC that states have the authority to remove network elements added by the states. Thus, if a state commission, pursuant to section 251(d)(3), adds a network element to the list of network elements an incumbent LEC must provide, state commissions also have the authority subsequently to remove those elements they add. As discussed above, section 251(d)(3)(A) allows state commissions to impose additional unbundling obligations as long as they comply with subsections 251(d)(3)(B) and (C). If a state commission determines that the additional unbundling obligations it imposed no longer comply with section 251, it must remove those obligations pursuant to section 251(d)(3). Beyond ensuring that removal of those state-imposed obligations are consistent with sections 251 and 253 of the Act, the Commission has no authority to prevent a state from removing a state-imposed unbundling obligation. Furthermore, state commissions that have imposed additional unbundling requirements, pursuant to section 51.317 of our rules, will need to periodically revisit such decisions to determine whether such decisions continue to comply with the standards articulated in this Order. 169. We conclude that, at this time, removing network elements from the unbundling obligations established in this Order on a state-by-state basis would not be consistent with the goals of the 1996 Act. Specifically, in this proceeding, we have examined each network element identified previously by the Commission or by the parties, and we have made an affirmative finding as to whether or not the particular element now satisfies the unbundling standards of the Act as clarified by the Supreme Court. Moreover, we have considered how unbundling these elements will affect the development of competition in the local markets as contemplated by Congress, and whether unbundling particular elements will further the goals of the Act. Indeed, we have found that unbundling particular network elements is necessary to further the goals of the Act. Consequently, at this time, state decisions to remove these network elements from the national unbundling obligations would "substantially prevent implementation of the requirements of section 251," as prohibited by subsection 251(d)(3)(C). 170. Furthermore, we find that there are compelling policy reasons for not removing elements from the national list on a state-by-state basis at this time. Unbundling obligations that vary from state to state in the near future would substantially undermine the reasons discussed above for implementing a national list in the first instance. We agree with commenters that argue that state-by-state removal of network elements from the national list, at least in the near future, would lead to greater uncertainty in the market and would hinder the development of competition. As discussed above, we have determined that national unbundling rules promote competition in telecommunications market by guaranteeing that a specific set of network elements will be available nationwide for a minimum amount of time. 171. We agree with the California PUC and other state commissions that having a guaranteed list of network elements provides enough certainty to allow competitive LECs to develop and implement regional and national business plans. Creating certainty and predictability in the market will also benefit competition by enabling competitors to raise capital at lower cost to create and enhance their networks. If each state could remove immediately the unbundling obligations established in this Order, competitors would not have the benefit of knowing how long an element would be available on an unbundled basis in any given locale. The resulting uncertainty would frustrate the ability of carriers to plan and implement competitive entry strategies developed to serve customers on a regional or national basis. 172. We also agree with commenters that state-by-state removal of network elements from the national list would complicate negotiation of interconnection agreements and would most likely lead to increased litigation. Indeed, it could force competitive LECs, each time they seek to enter into an interconnection agreement, to demonstrate that the identified elements continue to meet the standards of the Act. Once an incumbent LEC is able to convince a state commission that the element no longer meets our unbundling standard, the ruling would likely set a precedent for other LECs. In addition, the possibility that a state decision in one interconnection proceeding could affect all interconnection agreements would require competitive LECs to monitor the status of these arbitrations even if they are not participants in the arbitration. We therefore agree with the Illinois Commission that having only this Commission remove elements from the national list makes it easier for the states to resolve disputed issues during inter-carrier negotiations and arbitrations. 173. We believe that incumbent LECs have more of an incentive than competitive LECs to challenge the unbundling obligations set forth in this Order. In addition to the delay and uncertainty created by litigating the unbundling obligations of incumbent LECs, state commissions, as well as incumbent LECs and competitors, would be faced with the additional costs of litigation. Many state commissions and small carriers have limited resources and would be unduly burdened if they were have to finance on-going litigation of the unbundling rules. Moreover, as several state commenters and NARUC note, section 252(e)(6) appears to limit review of state commission decisions to federal district court. Thus, each state decision could eventually lead to litigation in the federal courts, creating even more uncertainty and further delaying the benefits of competition to consumers. CLXXIV. APPLICATION OF THE STANDARD TO INDIVIDUAL NETWORK ELEMENTS A. Loops 1. Background 175. In the Local Competition First Report and Order, the Commission found that incumbent LECs must provide local loops on an unbundled basis to requesting carriers. The Commission concluded that such access was technically feasible and would promote competition in the local exchange market. The Commission, at that time, did not require subloop unbundling, or specify whether "dark fiber" fell within the definition of the loop. The Local Competition First Report and Order also did not address the status of "inside wire" (wiring located inside the customer premises but owned by the incumbent). 176. In the Notice, we stated that it was our strong expectation that, under any reasonable interpretation of the "necessary" and "impair" standards of section 251(d)(2), loops would be subject to the section 251(c)(3) unbundling obligations. The Notice also requested that parties discuss specific costs and analyze the availability of alternative sources of the loop facilities. 177. In general, incumbent LECs contend that the definition of the loop should not include high-capacity loops that serve large business customers, dark fiber, inside wire, and loop conditioning. State regulatory commissions and competitive LECs argue that loops should be unbundled. The state commissions disagree among themselves as to whether or not competitive providers are impaired without access to dark fiber. They also disagree as to whether dark fiber should be included within the loop and transport unbundled network elements definitions or be unbundled as a separate network element. 1. Discussion 178. We conclude that LECs must provide access to unbundled loops, including high- capacity loops, nationwide. We find that requesting carriers are impaired without access to loops, and that loops include high-capacity lines, dark fiber, line conditioning, and certain inside wire. Requiring carriers to obtain loops from alternative sources would materially raise entry costs, delay broad-based entry, and limit the scope and timeliness of the competitor's service offerings. As described below, we conclude that neither self-provisioning loops nor obtaining loops from third-party sources is a sufficient substitute that would justify excluding loops from an incumbent LEC's unbundling obligation under section 251(c)(3). a. Definition 179. In the Local Competition First Report and Order, the Commission defined the loop as "a transmission facility between a distribution frame, or its equivalent, in an incumbent LEC central office, and the network interface device at the customer premises." The Commission also stated that the definition included, for example, two-wire and four-wire analog voice-grade loops, and two-wire and four-wire loops that are conditioned to transmit digital signals, such as xDSL. The Commission did not, however, specify whether "dark fiber" fell within the definition of the loop. 180. We modify the definition of the loop network element to include all features, functions, and capabilities of the transmission facilities, including dark fiber and attached electronics (except those used for the provision of advanced services, such as DSLAMs) owned by the incumbent LEC, between an incumbent LEC's central office and the loop demarcation point at the customer premises. In order to secure access to the loop's full functions and capabilities, we require incumbent LECs to condition loops. This broad approach accords with section 3(29) of the Act, which defines network elements to include their "features, functions and capabilities." Our intention is to ensure that the loop definition will apply to new as well as current technologies, and to ensure that competitors will continue to be able to access loops as an unbundled network element as long as that access is required pursuant to section 251(d)(2) standards. 181. Termination of the Loop. The loop definition the Commission adopted in the Local Competition First Report and Order defined the loop as terminating at the network interface device (NID) at the customer premises. We find the demarcation point preferable to the NID in defining the termination point of the loop because, in some cases, the NID does not mark the end of the incumbent's control of the loop facility. Where incumbents maintain ownership and control over a portion of the loop beyond the NID, the definition of the loop as set forth by the Commission in the Local Competition First Report and Order may not provide the competitor with actual access to the subscriber. 182. Section 68.3 of our rules defines the demarcation point as that point on the loop where the telephone company's control of the wire ceases, and the subscriber's control (or, in the case of some multiunit premises, the landlord's control) of the wire begins. Thus, the demarcation point is defined by control; it is not a fixed location on the network, but rather a point where an incumbent's and a property owner's responsibilities meet. The demarcation point is often, but not always, located at the minimum point of entry (MPOE), which is the closest practicable point to where the wire crosses a property line or enters a building. In multiunit premises, there may be either a single demarcation point for the entire building or separate demarcation points for each tenant, located at any of several locations, depending on the date the inside wire was installed, the local carrier's reasonable and nondiscriminatory practices, and the property owner's preferences. Thus, depending on the circumstances, the demarcation point may be located either at the NID, outside the NID, or inside the NID. 183. Although inside wire typically consists of junction and utility boxes, riser cable, and horizontal distribution wiring within an apartment building, it can also include the loop facility within a campus, a commercial park, or a garden apartment complex. We note that Teligent prefers the term "intra-building wiring," to emphasize that the plant in question is not always inside the customer premises, but may, especially in multiunit buildings, exist primarily within the landlord's, rather that the subscriber's, premises. Yet even the term "intra-building wire" may suggest limitations that do not apply in some situations, because "inside" wire is often out- of-doors, as is the case at garden apartments and campuses, among other places. Thus, although we refer to "inside wire" and "customer premises," for the sake of convenience, we acknowledge that the wire may be out-of-doors, and the "customer" may be a subscriber, a landlord, a condominium, a university, and so on. 184. Defining the loop to terminate at the same point as the incumbent LEC's control over facilities that it owns, will ensure that the competitor will be able to gain access to the entire loop, including inside wire. We note that, in our Access to Competitive Networks proceeding, we are seeking additional comment on the legal and technical issues arising from unbundled inside wiring and premises facilities. We also note that Section 251(d)(2) imposes obligations only on incumbent local exchange carriers and not, for instance, on third parties (such as the owners of multi- tenant buildings). Thus, the rules adopted in this Order are not intended to give competitive service providers any additional legal rights vis-a-vis such third parties, including access to a multi-unit building over the objection of the property owner. Those issues are being addressed in other proceedings before the Commission. 185. Conditioned Loops. We clarify that incumbent LECs are required to condition loops so as to allow requesting carriers to offer advanced services. The terms "conditioned," "clean copper," "xDSL-capable" and "basic" loops all describe copper loops from which bridge taps, low-pass filters, range extenders, and similar devices have been removed. Incumbent LECs add these devices to the basic copper loop to gain architectural flexibility and improve voice transmission capability. Such devices, however, diminish the loop's capacity to deliver advanced services, and thus preclude the requesting carrier from gaining full use of the loop's capabilities. Loop conditioning requires the incumbent LEC to remove these devices, paring down the loop to its basic form. 186. GTE contends that the Eighth Circuit, in the Iowa Utils. Bd. v. FCC decision, overturned the rules established in the Local Competition First Report and Order that required incumbents to provide competing carriers with conditioned loops capable of supporting advanced services even where the incumbent is not itself providing advanced services to those customers. We disagree. Although the Eighth Circuit overturned certain rules to the extent those rules required incumbent LECs to provide access to unbundled network elements at levels of quality superior to those the incumbent LECs provide themselves, the court also expressly affirmed the Commission's determination that section 251(c)(3) requires incumbent LECs to provide modifications to their facilities to the extent necessary to accommodate access to network elements. We find that loop conditioning, rather than providing a "superior quality" loop, in fact enables a requesting carrier to use the basic loop. Because competitors cannot access the loop with all its native "features, functions, and capabilities" unless it has been stripped of accreted devices, we conclude that loop conditioning falls within the definition of the loop network element, and is also consistent with the Eighth Circuit opinion. 187. Dark Fiber. We also modify the loop definition to specify that the loop facility includes dark fiber. Dark fiber is fiber that has not been activated through connection to the electronics that "light" it, and thereby render it capable of carrying communications services. Because it is in place and easily called into service, we find that dark fiber is analogous to "dead count" or "vacant" copper wire that carriers keep dormant but ready for service. Thus, we disagree with GTE's argument that, unlike vacant copper, dark fiber does not qualify as loop plant. GTE maintains that extra "copper cable is installed to provide optimum flexibility" and contrasts this copper to dark fiber, which GTE terms "unused inventory." GTE clarifies that "[t]hese fibers remain dark until they are needed." We find this to be a distinction without a difference, and conclude that both copper and fiber alike represent unused loop capacity. We find, therefore, that dark fiber and extra copper both fall within the loop network element's "facilities, functions, and capabilities." 188. Attached Electronics. We conclude that, with the exception of Digital Subscriber Line Access Multiplexers (DSLAMs), the loop includes attached electronics, including multiplexing equipment used to derive the loop transmission capacity. The definition of a network element is not limited to facilities, but includes features, functions, and capabilities as well. Some loops, such as integrated digital loop carrier (IDLC), are equipped with multiplexing devices, without which they cannot be used to provide service to end users. Because excluding such equipment from the definition of the loop would limit the functionality of the loop, we include the attached electronics (with the exception of DSLAMs) within the loop definition. By contrast, and as we discuss below, we find that the DSLAM is a component of the packet switch network element. 189. High-Capacity Loops. We disagree with incumbent LECs that high-capacity loops should be excluded from the definition of the loop. High-capacity loops retain the essential characteristic of the loop: they transmit a signal from the central office to the subscriber, or vice versa. In a DS1 loop, for example, the attached electronics boost the wire's capacity, but the wire facility used for transmission of the traffic is indistinguishable from any other copper wire. Although it may be more profitable to serve customers over higher capacity lines, such differences do not support a modification of the loop definition to exclude high-capacity lines. Whether the Commission should refrain from unbundling high-capacity loops is another matter, which we discuss below in our unbundling analysis. 190. For similar reasons, we reject US West's argument that we should exclude from the definition the loop facilities that underlie private line and special access interconnection, because providing these services to competitors at lower-than- tariffed rates would "promote regulatory arbitrage and serve no valid statutory or public purpose." The Commission has not previously found that the requirements of section 251(c)(3) are limited to any particular kind of service. Moreover, section 251(d)(2) of the Act refers to a " . . . carrier seeking access to provide the services that it seeks to offer." We find no basis for placing a restriction on what services a carrier may offer using the loop network element. Indeed, the prospect of competition among carriers to provide services over the loop at prices that more closely reflect the provider's costs seems to us to accord fully with Congress's intent in passing the 1996 Act. We do not now decide whether or not this analysis may extend to the enhanced extended loop (EEL), but rather seek comment on that issue in the Further Notice of Proposed Rulemaking, below. 191. Cross Connects. In the Local Competition First Report and Order, the Commission concluded that incumbent LECs must provide cross connect facilities between an unbundled loop and a requesting carrier's collocated equipment. The Commission emphasized this requirement because of its concern that incumbent LECs might have imposed unreasonable rates, terms, and conditions for such cross connect facilities in the past. Nothing in this Order disturbs the Commission's findings regarding cross connect facilities. In particular, we continue our policy that incumbent LECs may recover the cost of providing such facilities in accordance with our rules governing the costs of interconnection and unbundling. Charges for cross connect facilities must meet the cost-based standard provided in section 252(d)(1), and the terms and conditions of providing cross connect facilities must be reasonable and nondiscriminatory under section 251(c)(3). 192. Because we agree with the Commission's analysis of cross connect facilities in the Local Competition First Report and Order, we decline to include cross connect facilities within the definition of the loop network element. We continue to view the cross connect as a means of interconnection with a network element, rather than as part of the network element. We require, however, that incumbents provide cross connect facilities according to sections 252(d)(1) and 251(c)(3) at any technically feasible point that a requesting carrier seeks access to the loop. We conclude that such a requirement is needed wherever a competitor seeks access to the loop, because cross connection offers a potential bottleneck, and incumbents may have the incentive to impose unreasonable rates, terms, and conditions for cross-connect facilities. a. Proprietary Concerns Associated with the Loop 193. In the Local Competition First Report and Order, the Commission concluded that the technology associated with the loop is not proprietary in nature. Parties in this proceeding have not identified any proprietary concerns associated with unbundled loops, and we find none. We therefore apply the "impair" standard of section 251(d)(2), rather than the "necessary" standard, to determine whether loops are subject to the unbundling obligations of the Act. a. Unbundling Analysis for the Loop in General 194. We require incumbent LECs to provide unbundled access to loops nationwide. The record demonstrates that lack of access to unbundled loops impairs a carrier's ability to provide the services it seeks to offer because requiring carriers to self-provision loops would materially raise entry costs, delay broad-based entry, and limit the scope and quality of the competitor's offerings. We conclude that neither self- provisioning loops nor obtaining loops from third-party sources is an adequate alternative for loops that a carrier can obtain from an incumbent LEC under the section 251(c) unbundling obligation. We analyze the obligation to unbundle separable elements of the loop, such as inside wire, when we discuss subloop unbundling, below. We defer a decision on whether to unbundle the high frequency portion of the loop to a further proceeding. 195. Cost and Timeliness. We agree with the argument that self-provisioning is not a viable alternative to the incumbent's unbundled loops because replicating an incumbent's vast and ubiquitous network would be prohibitively expensive and delay competitive entry. We find the reasons for unbundling the loop that the Commission articulated in the Local Competition First Report and Order are still valid three years later. In that order, the Commission recognized that, without access to unbundled loops, competitors would need to invest immediately in duplicative facilities in order to compete for most customers, and that such investment and construction would likely delay, if not prohibit, market entry and postpone, perhaps indefinitely, the benefits of telephone competition for consumers. Moreover, the Commission found that without access to unbundled loops, competitive LECs would be required to sink a large initial investment in loop facilities before they had a customer base large enough to justify such an expenditure, thereby increasing the risk of entry and raising the competitive LEC's cost of capital. By contrast, permitting a competitor to purchase unbundled loops from the incumbent LEC allows the competitive LEC to build facilities gradually, and to deploy loops for its customers where it is efficient to do so. 196. Nothing in the record of this proceeding leads us to a different conclusion. To the contrary, we find that, as a practical matter, building loop plant continues to be, in most cases, prohibitively expensive and time-consuming. Because of the size of their networks, incumbent LECs enjoy advantages of scope that competitors cannot replicate. We find that it would be unreasonable to expect a competitive LEC to invest the large sums of capital needed to build out ubiquitous loop plant before the competitive LEC has established a substantial and secure customer base. Unlike switches, which can be scaled to need, relocated if the business fails to develop, and which can accommodate a fluctuating customer base, much of the loop is often dedicated to a particular location. In addition, if the competitive LEC loses the customer back to the incumbent or to another competitor, the competitive LEC would probably bear the full loss of its sunk investment in the redundant loop. 197. We disagree with incumbents' assertions that we should not unbundle high-capacity loops because competitive LECs have successfully self-provisioned loops to certain large business customers. According to these commenters, the call concentration and revenue potential of "high-capacity" lines (DS1 and higher) make self- provisioning high-capacity lines an economically viable alternative to the incumbent LECs' unbundled high-capacity loops. Building out any loop is expensive and time-consuming, regardless of its capacity. That some competitive LECs, in certain instances, have found it economical to serve certain customers using their own loops suggests to us only that carriers are unimpaired in their ability to serve those particular customers. This evidence tells us nothing about the customer the competitor would like to serve but cannot because the cost of building a loop from the customer premises to the competitive LEC's switch is prohibitive. 198. For similar reasons, we reject BellSouth's proposal that we not require incumbent LECs to unbundle larger business loops in Special Access Pricing zones 1 and 2. Because of the expense inherent in building loops, we find that it would be extremely difficult for competitive LECs to overbuild the ubiquitous loop plant that the incumbents have built up over decades, even to serve businesses in urban districts. The enormous sunk investment required to install loops would inevitably lead to competition in patches, rather than the seamless competitive service of a fully competitive market. Moreover, we find that using Special Access Pricing zones, as recently modified by the Commission, would provide incumbent LECs with discretion to define their own loop unbundling obligations. We agree with MCI WorldCom that the Special Access Pricing zone approach would grant incumbent LECs latitude to "change their methodologies for defining zones to upset their competitor's business plans." We find that premising an incumbent LEC's loop unbundling obligation on a geographic boundary defined, to a large degree, by the incumbent LEC itself could allow an incumbent LEC to minimize its unbundling obligation, and would not respond to a requesting carrier's need for access to unbundled loops. 199. In addition to the large costs of building loop plant, we agree with commenters in this phase of the proceeding that overbuilding the incumbent LEC's loops would embroil the competitor in lengthy rights-of-way disputes, and would require the unnecessary digging up of streets. Thus, we find that even if competitors were able to finance the replication of the incumbents' loop plant, construction of new facilities would at the least materially delay competitors' ability to bring their services to consumers. Such delays would frustrate the competitor's ability to offer timely service to prospective customers. Although competitive LECs have successfully constructed loops in some circumstances, we find that the cost, risk, disruption, and delay of self-provisioning loop plant would, for many consumers, foreclose the benefits of competition. 200. Moreover, in the Local Competition First Report and Order, the Commission specified that the definition of the loop includes various grades of loops to allow transmission of digital signals needed to provide multiple services and DS1-level signals. The Commission reasoned that the ability to offer various functions in competition with incumbent LECs could benefit small entities serving niche markets. We continue to believe that access to these high-capacity lines is necessary for ubiquitous deployment of high-speed services, including high-speed Internet access. We therefore agree with competitive LECs that failing to assure access to high-capacity loops would impair their ability to provide the services that they seek to offer in broadband service markets. 201. Ubiquity and Quality. We disagree with parties that argue that mobile telephones and fixed wireless offer an alternative to the incumbent's loop, and that loops therefore should not be unbundled. Although we find these technologies promising, we conclude that they are not yet viable alternatives to the incumbent's wireline loop facilities. In particular, we find that alternative loop technologies are not as widespread as the incumbent's ubiquitous network. These alternatives do not offer the same functionality as wireline service, and the data capabilities of these mobile services are generally inferior to wireline loops' data transmission capabilities. Cellular and PCS telephone footprints, though expanding, are not ubiquitous. Indeed, millions of Americans are not yet served by mobile wireless carriers. Moreover, the sound quality of cellular and PCS service is not always equal to wireline service. Fixed wireless cannot yet offer more than four lines, or high-speed Internet connection. 202. We also disagree with the incumbent LECs' argument that cable television service offers a viable alternative to the incumbent's unbundled loop. Cable service is largely restricted to residential subscribers, and generally supports only one-way service, not the two-way communications telephony requires. Moreover, we conclude that declining to unbundle loops in areas where cable telephony is available would be inconsistent with the Act's goal of encouraging entry by multiple providers. Given that neither mobile nor fixed wireless can yet replace wireline service, if we were to take the incumbents' approach, consumers might be left to a choose between only the cable company and the incumbent LEC. 203. Loops Capable of Providing High-Speed Data Services. We conclude that permitting incumbents to deny access to basic loops stripped of accreted devices, i.e., "conditioned" loops, would preclude the ability of competitors to offer high- speed data services. Such unencumbered copper wire is necessary for requesting carriers to provide most types of xDSL service. While some "flavors" of xDSL can be provided over loops with a limited number of impediments, as a general rule the quality of such service particularly the speed is significantly diminished, compared to the service provided over unencumbered wires. DSL-capable loops provide end users with broadband data transmission, which allows rapid access to the Internet. Unbundling basic loops, with their full capacity preserved, allows competitors to provide xDSL services. This in turn will foster investment, innovation, and competition in the local telecommunications marketplace. Without access to these loops, competitors would be at a significant disadvantage, and the incumbent LEC, rather than the marketplace, would dictate the pace of the deployment of advanced services. We also note that the availability of conditioned loops enables competitors to deploy xDSL service beyond the major metropolitan areas. Finally, we note our obligation under section 706 to encourage the deployment of advanced services by, among other means, promoting competition in the telecommunications market. 204. As the Commission stated in the Local Competition First Report and Order, requiring incumbents to provide conditioned loops will, in some instances, require the incumbent LEC to take affirmative steps to enable requesting carriers to provide services that the incumbent does not currently provide. We now clarify that we require the incumbent to provide loops with all their capabilities intact, that is, to provide conditioned loops, wherever a competitor requests, even if the incumbent is not itself offering xDSL to the end-user customer on that loop. Thus, incumbent LECs cannot refuse a competitive LEC's request for conditioned loops on the grounds that they themselves are not planning to offer xDSL to that customer. 205. In the Local Competition First Report and Order, the Commission also stated that requesting carriers would compensate the incumbent LECs for the cost of conditioning the loop. Covad and Rhythms argue that, because loops under 18,000 feet generally should not require devices to enhance voice-transmission, the requesting party should not be required to compensate the incumbent for removing such devices on lines of that length or shorter. 206. We agree that networks built today normally should not require voice-transmission enhancing devices on loops of 18,000 feet or shorter. Nevertheless, the devices are sometimes present on such loops, and the incumbent LEC may incur costs in removing them. Thus, under our rules, the incumbent should be able to charge for conditioning such loops. 207. We recognize, however, that the charges incumbent LECs impose to condition loops represent sunk costs to the competitive LEC, and that these costs may constitute a barrier to offering xDSL services. We also recognize that incumbent LECs may have an incentive to inflate the charge for line conditioning by including additional common and overhead costs, as well as profits. We defer to the states to ensure that the costs incumbents impose on competitors for line conditioning are in compliance with our pricing rules for nonrecurring costs. 208. In addition, we agree with commenters that argue that incumbent LECs must provide "trouble reports" to the competitive LECs for any function or capability of the accessed loop element, and that the incumbent may not limit such reports to voice-transmission trouble only. Not knowing whether or not the accessed line is functioning properly impairs a competitive LEC's ability to provide service, because subscribers may tend to blame the new competitor, rather than the familiar incumbent, for any lapse or degradation of service. Thus, we conclude that, in so far as it is technically feasible, the incumbent must test and report trouble on conditioned lines, if requested by the competitor, for all of the line's features, functions, and capabilities, and may not restrict its testing to voice-transmission only. 209. Dark Fiber. We agree with commenters that argue that, because dark fiber provides high transmission capabilities at relatively low cost, unbundling dark fiber is essential for competition in the provision of advanced services. We reject the incumbents' reasoning that, because competitive LECs have installed lit fiber to certain high-volume customers, they could also install their own dark fiber, and therefore are not impaired without access to the incumbent's dark fiber. As with other loops, we decline to infer from competitive LEC self-provisioning in certain circumstances that, as a general matter, the expense and delay involved in laying fiber do not impair the ability of entrants and other competitive LECs to provide the services they seek to offer. We see no reason to distinguish dark fiber from our general unbundling analysis for loops. 210. US West argues that competitors do not need the incumbent LECs' fiber because a wholesale market exists in loop fiber. We find, however, that the nascent wholesale market in fiber loop facilities is not yet extensive enough for us to conclude that competitors are not impaired without access to incumbent LECs' unbundled dark fiber loops. We also agree with the argument that unbundled loops, including fiber, allow competitive LECs to build out their networks gradually. By supplementing their own facilities with unbundled fiber loops, a competitive LEC can offer advanced services ubiquitously and not limit its service offering to small areas of concentrated demand. 211. Because fiber is currently a more significant component of interoffice transport than the loop network element, we discuss aspects of dark fiber common to both elements when we discuss interoffice transport below. We note here, however, that GTE raises concerns that incumbents, because of their carrier-of-last-resort obligations, have a special need for fiber reserves. As we explain in greater detail below, we find these concerns exaggerated, because the capacity of fiber can be increased many fold simply by increasing the power of the electronics that light it. We find, therefore, that a shortage of fiber capacity caused by unbundling is highly unlikely. 212. In addition, GTE and the Telecommunications Industry Association argue that requiring incumbent LECs to unbundle fiber will reduce their incentive to build fiber loops in the first place. We remain skeptical that this is the case, because incumbents face loop unbundling obligations no matter which technology they deploy. We note, however, that the Texas commission has already established moderate restrictions governing the availability dark fiber. We do not wish to disturb the reasonable limitations and technical parameters for dark fiber unbundling that Texas or other states may have in place. If incumbent LECs are able to demonstrate to the state commission that unlimited access to unbundled dark fiber threatens their ability to provide service as a carrier of last resort, state commissions retain the flexibility to establish reasonable limitations governing access to dark fiber loops in their states. 213. Goals of the Act. We conclude that access to the full capabilities of incumbent LECs' loop plant nationwide will further the goals of the Act. Requiring access to unbundled loops will promote the rapid development of competition and bring the benefits of competition to greater numbers of consumers. Access to unbundled loops will also encourage competition to provide broadband services. We are convinced that greater, not fewer, options for procuring loops will facilitate entry by competitors, and that Congress intended for competitors to have these options available." We find that the benefits of uniform loop unbundling outweigh the costs of creating a patchwork regime in which incumbents will seek to litigate whether particular loops should be unbundled or where an alternative to the incumbent LEC's loop is arguably substitutable. For these reasons, incumbent LECs must provide unbundled access to their loop network element nationwide. 214. Spectrum Unbundling. A number of parties request that the Commission identify loop spectrum as a separate unbundled network element. In particular, they argue that requesting carriers need access to the high-frequency loop spectrum on an unbundled basis in order to provide advanced telecommunications services, including xDSL. We decline, at this time, to identify loop spectrum as a separate unbundled network element. In the Advanced Services First Report and Order and FNPRM, we will consider whether the high-frequency spectrum of the loop qualifies as an unbundled network element and the operational issues associated with such unbundling. We believe that the record developed in that proceeding more fully addresses the issues associated with spectrum unbundling, and we therefore decline to address those issues in this proceeding. A. The Subloop 1. Background 215. In the Local Competition First Report and Order, the Commission declined to identify the feeder, feeder/distribution interface (FDI), and distribution components of the loops as individual network elements. The Commission noted, however, that subloop unbundling could provide competitors flexibility in deploying some portions of loop facilities, while elsewhere relying on the incumbent LEC's facilities. In addition, the Commission noted that carriers would need access at points along the loop closer to the customer premises to provide some high bandwidth services, such as ADSL. The Commission also found that, although the record presented evidence mainly of logistical, rather than technical, impediments to subloop unbundling, proponents of subloop unbundling did not address technical issues raised by incumbent LECs. The Commission stated that it would revisit subloop unbundling when the record on the issue had been more fully developed 216. In the Notice, we sought comment on whether, due to technological changes, we should require subloop unbundling at the remote terminal or at other points within the incumbent LEC's network. We sought comment on whether to unbundle incumbent-owned facilities on the end-user side of the NID. We asked commenters to apply the "necessary" and "impair" standards and to discuss costs and availability on an element-by-element basis. We also asked those commenters requesting further unbundling of the local loop to discuss possible alternatives. 217. Competitive LECs argue generally that they need unbundled access to subloop elements in order to: (1) connect their own facilities to the incumbent's inside wire; (2) access loops that an incumbent LEC provides over integrated digital loop carrier (IDLC) technology; and (3) provide advanced services over xDSL. These commenters argue that they are also financially burdened if they must pay for an entire loop when they need to use only a portion of it. Incumbents argue generally that competitors are not impaired without access to subloops; that technical and logistical impediments prevent subloop unbundling; and that network architectures differ too broadly to adopt an unbundling rule that applies nationwide. Several state commissions argue that subloop unbundling requires a case-by-case analysis that the states are in the best position to perform. For example, Texas states that subloop unbundling meets the "impair" standard of section 251(d)(2) and requires subloop unbundling at the remote terminal. 1. Discussion 218. We find that lack of access to unbundled subloops materially diminishes a requesting carrier's ability to provide services that it seeks to offer. We also conclude that access to subloop elements is likely to be the catalyst that will allow competitors, over time, to deploy their own complementary subloop facilities, and eventually to develop competitive loops. Lack of access to subloops discourages competitive LECs from attempting to combine their own feeder plant with the incumbent's distribution plant to minimize their reliance on the incumbents' facilities. We also find that lack of unbundled access to the incumbent's subloops would preclude competitors from offering some broadband services. Accordingly, we conclude that incumbent LECs must provide unbundled access to subloops nationwide, where technically feasible. a. Definition of the Subloop 219. We define subloops as portions of the loop that can be accessed at terminals in the incumbent's outside plant. An accessible terminal is a point on the loop where technicians can access the wire or fiber within the cable without removing a splice case to reach the wire or fiber within. These would include a technically feasible point near the customer premises, such as the pole or pedestal, the NID (which we discuss below), or the minimum point of entry to the customer premises (MPOE). Another point of access would be the feeder distribution interface (FDI), which is where the trunk line, or "feeder," leading back to the central office, and the "distribution" plant, branching out to the subscribers, meet, and "interface." The FDI might be located in the utility room in a multi-dwelling unit, in a remote terminal, or in a controlled environment vault (CEV). We acknowledge that some FDIs are more accessible than others; utility rooms are generally more spacious than vaults. A third point of access is, of course, the main distribution frame in the incumbent's central office. 220. We believe that a broad definition of the subloop that allows requesting carriers maximum flexibility to interconnect their own facilities at these points where technically feasible will best promote the goals of the Act. Access to portions of the loop element at these points, i.e., access to the subloop, will facilitate rapid development of competition, encourage facilities-based competition, and promote the deployment of advanced services. Our intention is to ensure that the subloop definition will apply to new as well as current technologies, and to ensure that competitors will continue to be able to access subloop unbundled network elements as long as that access is required pursuant to section 251(d)(2) standards. a. Proprietary Concerns Associated with Subloops 221. The record does not indicate, nor do commenters argue, that subloops are proprietary. Moreover, we do not discern any copyright, patent, or trademark or trade secrecy implications to subloop unbundling. We therefore apply the "impair" standard of section of 251(d)(2)(B) to determine whether subloops are subject to the unbundling requirements of the Act. a. Unbundling Analysis for Subloops 222. We conclude that incumbent LECs must provide unbundled access to subloops. Applying our unbundling analysis, we conclude that lack of access to unbundled subloops at technically feasible points throughout the incumbent's loop plant will impair a competitor's ability to provide services that it seeks to offer. We agree with commenters that self-provisioning subloop elements, like the loop itself, would materially raise entry costs, delay broad-based entry, and limit the scope and quality of the competitive LEC's service offerings. In addition, we find that access to subloop elements promotes self-provisioning of part of the loop, and thus will encourage competitors, over time, to deploy their own loop facilities and eventually to develop competitive loops where it is cost efficient to do so. 223. We clarify that "technically feasible points" would include a point near the customer premises, such as the point of interconnection between the drop and the distribution cable, the NID, or the MPOE. Such access would give competitors unbundled access to the inside wire subloop element, in cases where the incumbent owns and controls wire inside the customer premises. It would also include any FDI, whether the FDI is located at a cabinet, CEV, remote terminal, utility room in a multi- dwelling unit, or any other accessible terminal. 224. Cost and Timeliness. We agree with commenters that loop facilities, including subloop elements, are the most time-consuming and expensive network element to duplicate on a pervasive scale, and that the cost of self-provisioning subloops can be prohibitively expensive. Self-provisioning subloops would require requesting carriers to incur significant sunk costs prior to offering services to end users. Requiring competitors to expend such sums would, at a minimum, delay entry and thus postpone the benefits of competition for consumers. 225. We are not persuaded by GTE's argument that, because the whole loop is an acceptable substitute, a competitor is not impaired without access to the subloop. First, as we explain below, the undivided loop does not always afford competitors access to subscribers, as is the case with IDLC loops. Also, as a rule, requesting carriers that supply their own facilities cannot afford to pay twice first for the facilities they self-provision, and again for the incumbent's loop, including the portion that they do not utilize. We agree with the Illinois Commission that unbundling subloops provides greater efficiency for the requesting carrier because the carrier will not have to buy the entire loop to interconnect its own facilities with wiring on the customer premises. If competing carriers that need only a portion of the loop must either pay for the entire loop or forego access to that loop altogether, many consumers will be denied the benefits of competition. 226. GTE contends that possible rights-of-way, zoning, power supply, and similar alleged impediments should prevent us from requiring the incumbent to provide loop sub-elements on an unbundled basis. We assume that GTE is referring to potential obstacles that the requesting carrier may encounter from cities, counties, electric power companies, and similar third parties when it seeks to interconnect its equipment at subloop access points. We find that such obstacles, however, to the extent they develop, are for the competitive LEC to resolve with the municipality or utility. Such obstacles are not relevant to our determination of whether the competitor is impaired without unbundled access to the incumbent's subloop elements, and do not absolve the incumbent from its obligation to provide unbundled access to those elements. 227. Impact on Network Operations. In order to encourage the development of facilities- based competition, requesting carriers must be able to interconnect their networks with the incumbent's network facilities that are designed to provide similar services. 228. First, if those competitors that are attempting to rely primarily on their own facilities are unable to interconnect near the customer premises, the end users those competitors target would have to forego the benefits of competition and new technology those competitors offer. We agree with several state regulatory commissions that argue that, to the extent that requesting carriers are denied flexibility in connecting their facilities to the local loop, these carriers are impaired from developing their own network infrastructure. In those instances where competitive carriers are able to self-provision a portion of the loop, lack of access to the part of the incumbent's loop they need could impede the competitors' ability to develop their own network architecture and provide new service offerings. On the other hand, the gradual self-provisioning that such access encourages could lead, in time, to conditions that would permit the eventual elimination of the loop element from the unbundling obligations of the Act. 229. For example, wireless providers may require only the final leg of loop distribution plant before the wire passes to customer control at the demarcation point. In particular, a facilities-based provider's ability to offer service in a multi-unit building or campus may be severely impaired if it must install duplicative inside wiring. We agree with the argument that requiring competitive LECs to convince landlords and customers to permit the construction of redundant inside wiring would substantially impede market entry and competition. Even if permission were obtained, over-building inside wire might be sufficiently expensive and time- consuming to deter potential competitors. Thus, we conclude that access to these subloop elements at technically-feasible interconnection points is necessary for successful competition by facilities-based competitors. 230. Second, carriers need unbundled subloops to serve subscribers currently served by IDLC loops. IDLC technology allows a carrier to "multiplex" and "de-multiplex" (combine and separate) traffic at a remote concentration point, or remote terminal, and to deliver the combined traffic directly into the switch, without first separating the traffic from the individual lines. In such cases, competitors generally cannot access IDLC loops at the incumbent's central office. In order to reach subscribers served by the incumbent's IDLC loops, a requesting carrier usually must have access to those loops before the point where the traffic is multiplexed. That is where the end-user's distribution subloop can be diverted to the competitive LEC's feeder, before the signal is mixed with the traffic from the incumbent LEC's other distribution subloops for transport through the incumbent's IDLC feeder. Accordingly, we find that denying access at this point may preclude a requesting carrier from competing to provide service to customers served by the incumbent's IDLC facilities. This would particularly affect consumers in rural areas, where incumbent LECs use the greatest proportion of DLC loops. 231. Third, competitors seeking to offer services using xDSL technology need to access the copper wire portion of the loop. In cases where the incumbent multiplexes its copper loops at a remote terminal to transport the traffic to the central office over fiber DLC facilities, a requesting carrier's ability to offer xDSL service to customers served over those facilities will be precluded, unless the competitor can gain access to the customer's copper loop before the traffic on that loop is multiplexed. Thus, we note that the remote terminal has, to a substantial degree, assumed the role and significance traditionally associated with the central office. In addition, in order to use its own facilities to provide xDSL service to a customer, a carrier must locate its DSLAM within a reasonable distance of the customer premises, usually less than 18,000 feet. In both of these situations, a requesting carrier needs access to copper wire relatively close to the subscriber in order to serve the incumbent's customer. 232. Goals of the Act. Access to unbundled subloop elements allows competitive LECs to self-provision part of the loop, and thus, over time, to deploy their own loop facilities, and eventually to develop competitive loops. If requesting carriers can reduce their reliance on the incumbent by interconnecting their own facilities closer to the customer, their ability to provide service using their own facilities will be greatly enhanced, thereby furthering the goal of the 1996 Act to promote facilities- based competition. Failure to unbundle the subloop would cause residential and small business consumers to wait unnecessarily for competitive alternatives. We also find that the availability of unbundled subloops will accelerate the development of alternative networks, because it will allow requesting carriers efficiently to connect their facilities with the incumbent's loop plant. Thus, our decision to unbundle subloops is consistent with the 1996 Act's goals of rapid introduction of competition and the promotion of facilities-based entry. a. Technical Feasibility 233. We note that parties commenting in this proceeding disagree as to the technical feasibility of accessing various points of the loop. For example, SBC contends that incumbents should not be required to unbundle subloops at the CEV because the CEV is a small, protected environment that is not designed for access by multiple parties. SBC also argues that unbundling at the cabinet will jeopardize network security. Competitive LECs, on the other hand, argue that SBC exaggerates these impediments, which they maintain are not insurmountable. 234. MGC asserts, and we agree, that our collocation rules, which we recently clarified in the Advanced Services First Report and Order, apply to collocation at any technically feasible point, from the largest central office to the most compact FDI. This is because our collocation rules concern methods and standards of obtaining interconnection and access to unbundled network elements under section 251 of the Act, and thus are not directed to any one type of facility. Although we intend to make collocation available at all accessible terminals on the loop, we acknowledge that the incumbent's network was not designed to house additional equipment of competitors. Our rules do not require incumbents to build additional space. Nor do our rules, however, preclude requesting carriers from constructing their own facilities adjacent to the incumbent's equipment. Moreover, in some cases, technicians may not need to enter the cabinet or vault at all because virtual collocation arrangements will satisfy the needs of all parties. We note that, prior to adoption of rules requiring incumbent LECs to offer collocation to competitors, incumbent LECs raised similar doubts as to whether collocation would be feasible at central offices. As indicated by the number of collocation arrangements in place today, these doubts were not well-founded. 235. The record indicates that the space available for collocating and interconnecting at various subloop access points will vary depending on the incumbent's existing plant at a particular location. For example, the feeder/distribution interface for a business park could be located in a room that contains a mini-MDF, racks of equipment, and enough unoccupied space to accommodate easily the requesting carrier's equipment. In other situations, such as at a remote terminal in a cabinet, the FDI may be housed in a facility that has no spare space at all. We note that Texas supports unbundling the subloop, but has not ordered unbundling at the FDI due to technical problems that, according to Texas, would threaten the integrity of the network. Ohio states that copper loops are still the dominant technology in its state, and that it has not seen evidence to suggest that it is technically feasible to unbundle copper subloops. Ohio also points out that the technical feasibility of unbundling subloops at particular points on the network may change with the introduction of new technologies. 236. As we explain above, however, we conclude that the goals of the Act are best served by determining unbundling rules that apply to network elements nationwide. In adopting a rule that requires incumbents to unbundle subloops at the points identified above, we seek to provide requesting carriers maximum flexibility to interconnect with the incumbent's network at technically feasible points in order to allow competitors to serve customers efficiently. Accordingly, we establish a rebuttable presumption that the subloop can be unbundled at any accessible terminal in the outside loop plant. If the parties are unable to reach an agreement pursuant to voluntary negotiations about the availability of space or the technical feasibility of unbundling the subloop at one of the points identified above, the incumbent will have the burden of demonstrating to the state, in the context of a section 252 arbitration proceeding, that there is no space available or that it is not technically feasible to unbundle the subloop at these points. 237. Our approach to subloop unbundling permits evaluation of the technical feasibility of subloop unbundling on a case-by-case basis, and takes into account the different loop plant that has been deployed in different states. We find that the questions of technical feasibility, including the question of whether or not sufficient space exists to make interconnection feasible at assorted huts, vaults, and terminals, and whether such interconnection would pose a significant threat to the operation of the network, are fact specific. Such issues of technical feasibility are best determined by state commissions, because state commissions can examine the incumbent's specific architecture and the particular technology used over the loop, and thus determine whether, in reality, it is technically feasible to unbundle the subloop where a competing carrier requests. We also note we are considering legal issues regarding access to premises in the Access to Competitive Networks proceeding. 238. We further note that SBC proposes to avoid difficulties associated with competing carriers serving multi-unit premises by eliminating multiple demarcation points in favor of a single demarcation point, which, according to SBC, would remedy competitive LECs' concerns. OpTel similarly suggests that the incumbent should provide a single point of interconnection at or near the property line of multi-unit premises. OpTel further maintains that the cost of any network reconfiguration required to create a point of interconnection that would be accessible to multiple carriers should be shared by all the carriers concerned. 239. Although we do not amend our rules governing the demarcation point in the context of this proceeding, we agree that the availability of a single point of interconnection will promote competition. To the extent there is not currently a single point of interconnection that can be feasibly accessed by a requesting carrier, we encourage parties to cooperate in any reconfiguration of the network necessary to create one. If parties are unable to negotiate a reconfigured single point of interconnection at multi-unit premises, we require the incumbent to construct a single point of interconnection that will be fully accessible and suitable for use by multiple carriers. Any disputes regarding the implementation of this requirement, including the provision of compensation to the incumbent LEC under forward- looking pricing principles, shall be subject to the usual dispute resolution process under section 252. We emphasize that this principle in no way diminishes a carrier's right to access the loop at any technically feasible point, including other points at or near the customer premises. We also note that unbundling inside wire, and access to premises facilities in general, present specific technical issues, and that we have sought additional comment on these issues in our Access to Competitive Networks proceeding. If the record developed in that proceeding demonstrates the need for additional federal guidance on legal or technical feasibility issues related to subloop unbundling, we will provide such additional guidance, consistent with the policies established in this Order. 240. Our approach to subloop unbundling reflects the network as it exists today. Technology may develop, however, in ways that would render this approach too limiting. For that reason, we establish a further rebuttable presumption that, once one state has determined that it is technically feasible to unbundle subloops at a designated point, it will be presumed that it is technically feasible for any incumbent LEC, in any other state, to unbundle the loop at the same point everywhere. If the conditions surrounding a request for unbundling at a similar point differ to such an extent that it is not technically feasible for the incumbent to provide unbundled access to that subloop element, the incumbent will have the burden of demonstrating in a section 252 arbitration proceeding that such an arrangement is indeed not technically feasible under those different conditions. For example, Texas requires subloop unbundling at the remote terminal. If a competitive LEC seeks unbundled access to a subloop at the remote terminal from an incumbent LEC in New York, the burden rests with the New York incumbent LEC to prove that its own situation differs to such an extent that the Texas arrangement is not technically feasible. We believe that this "best practices" approach insures that incumbent LECs do not limit access to subloops based on unforeseeable technological and infrastructure developments. 241. In addition to arguing that remote terminals will in some cases be inaccessible, SBC also argues that, by separating feeder plant from distribution plant, the ability to perform mechanized testing and monitoring of the loop from the incumbent's switch would be lost. We do not believe that this technical issue precludes us from establishing unbundling obligations for subloops. Once the competitor has acquired the customer from the incumbent, the competitor will have the incentive to ensure that there is a method by which the customer's loop can be tested. The technical method by which this testing is accomplished is a matter for the parties to decide through negotiations. If the incumbent can demonstrate to the satisfaction of the state regulatory commission that it would incur increased expenses associated with testing the subloop network element, we presume such expenses would be included in the forward-looking price of the element. For similar reasons, we reject the argument that subloop unbundling is not feasible because it may create additional administrative costs. 242. Accordingly, we are not persuaded by incumbents' arguments that technical feasibility issues require us to find that subloops are not subject to the unbundling obligations of the Act. We note that incumbent LECs advanced similar arguments against collocation at central offices; we continue to reject those arguments in the subloop context as well. To the extent disputes arise over the feasibility of interconnecting at various points on the loop, states will address these issues as part of the arbitration process under section 252. A. Network Interface Devices (NIDs) 1. Background 243. In the Local Competition First Report and Order, the Commission concluded that incumbent LECs must offer unbundled access to the network interface device (NID). It defined the NID network element as a cross-connect device used to connect loop facilities to inside wiring. In that order, the Commission noted that a competitor deploying its own loops must be able to connect those loops to customers' inside wiring in order to provide service, especially to customers in multi-tenant buildings. The Commission also concluded that a requesting carrier is entitled to connect its loops, via its own NID, to the incumbent LEC's NID. 244. In the Notice, we sought comment on application of the "necessary" and "impair" standards of section 251(d)(2) to the network elements previously identified in the Local Competition First Report and Order, including the NID. Incumbent LECs argue that NIDs are off-the-shelf devices that can be purchased inexpensively. Competitive LECs argue that self-provisioning the NID is economically impracticable at the level of ubiquity needed to deploy service on a widespread basis. 1. Discussion 245. We conclude that lack of unbundled access to the incumbent's NID impairs the ability of requesting carriers to provide the services that they seek to offer. As described below, we conclude that the competitor's ability to self-provision NIDs does not constitute a viable alternative to unbundled access to the incumbent's NID element. Although the physical structure of the NID is widely available, it is access to the function, rather than the hardware itself, that competitors rely upon. The record indicates that requiring a requesting carrier to self-provision NIDs for all customers it seeks to serve would materially raise the cost of entry, delay broad facilities-based market entry, and materially limit the scope and quality of the competitor's service offerings. Accordingly, we require incumbent LECs to provide unbundled access to NIDs nationwide. a. Definition of the NID 246. In the Local Competition First Report and Order, the Commission defined the NID as a cross-connect device used to connect loop facilities to inside wiring. We modify that definition of the NID to include all features, functions, and capabilities of the facilities used to connect the loop distribution plant to the customer premises wiring, regardless of the particular design of the NID mechanism. Specifically, we define the NID to include any means of interconnection of customer premises wiring to the incumbent LEC's distribution plant, such as a cross-connect device used for that purpose. 247. We conclude that the NID definition, for the purposes of our unbundling analysis, should be flexible and technology-neutral. The Commission's rules permit considerable variation in the interconnection facilities between carrier and customer- controlled facilities. Furthermore, evolution in network design and technology will likely cause additional design variations among the hardware interfaces between carrier and customer premises facilities. Accordingly, we define the NID broadly to ensure that competitors will be able to obtain access to any of these facilities as an unbundled network element. Our intention is to ensure that the NID definition will apply to new technologies, as well as current technologies, and to ensure that competitors will continue to be able to access customer premises facilities as an unbundled network element, as long as that access is required pursuant to section 251(d)(2) standards. 248. We decline to adopt parties' proposals to include the NID in the definition of the loop. Similarly, we reject arguments that we should include inside wiring in the definition of the NID in order to permit facilities-based competitors access to inside wiring. Although competitors may choose to access the inside wire via the NID, in some circumstances they may choose to access the inside wire at another point, such as the minimum point of entry. By continuing to identify the NID as an independent unbundled network element, we underscore the need for the competitive LEC to have flexibility in choosing where best to access the loop. Competitors purchasing a subloop at the NID, however, will acquire the functionality of the NID for the subloop portion they purchase. We therefore find no need to include inside wiring in the definition of the NID, or to include the NID as part of any other subloop element. a. Proprietary Concerns Associated with the NID 249. In the Local Competition First Report and Order, the Commission did not identify any proprietary concerns associated with NIDs. No parties in this proceeding identify any proprietary concerns associated with the NID, and we find none. We therefore apply the "impair" standard of section 251(d)(2)(A) to determine whether NIDs are subject to the unbundling obligations of the Act. a. Unbundling Analysis 250. We find that lack of access to the NID would materially diminish a competitor's ability to provide the services it seeks to offer. In particular, we find that requesting carriers would be impaired without access to NIDs because self-provisioning NIDs would materially raise entry costs, delay broad facilities-based entry, and materially limit the scope and quality of the competitor's service offerings. Accordingly, we require incumbent carriers to provide unbundled access to their NIDs nationwide. Specifically, an incumbent LEC must permit a requesting carrier to connect its own loop facilities to the inside wire of the premises through the incumbent LEC's network interface device, or at any other technically feasible point, to access the inside wire subloop network element. 251. Cost and Timeliness. We agree with those commenters that maintain that there are no economic or practical alternatives to the NID that would otherwise enable requesting carriers to provide service. NIDs are individually dedicated to specific customer premises, and are often difficult to replace. Requesting carriers' ability to provide service to their customers would be materially diminished if they had to self provision NIDs because of the significant labor and construction costs involved in visiting the premises of each customer and installing the device. This is true for all customers, but is particularly evident for residential and small business markets because of the greater number of NIDs required to provide service to each customer. We therefore conclude that requiring competitors to install numerous, redundant NIDs at the interface to customer premises wiring would constitute a substantial economic and practical barrier to market entry, and a needless waste of carrier resources. 252. Ubiquity. We conclude that self-provisioning NIDs is not economically practical at the level of ubiquity at which incumbent LECs' NIDs are currently deployed. We disagree with GTE's argument that the NID should not be unbundled because the hardware is inexpensive and available from a multitude of non-incumbent LEC sources. Specifically, GTE claims that the NID hardware costs between $25 and $40, and that requesting carriers can purchase NIDs from the same sources that incumbent LECs use. We do not find that the cost and availability of NID hardware is dispositive of the need to unbundle access to incumbent LEC-installed NIDs. As with other network elements, in conducting our unbundling analysis under section 251(d)(2), we do not consider the cost and availability of network elements in isolation. Rather, we examine whether, after applying the factors we explained in the unbundling standard above, a requesting carrier is able, as a practical, economic, and operational matter, to use alternatives for the incumbent's network elements. Although the record indicates that NID hardware may be available from alternative sources and that NIDs are affordable individually, it is the aggregate cost and difficulty of installing duplicate NIDs at every potential customer location that substantially impairs a requesting carrier from offering services. 253. Goals of the Act. Access to unbundled NIDs furthers the Act's goals of promoting innovation, the rapid introduction of competition, and the development of facilities- based competition. If requesting carriers can reduce their reliance on the incumbent by interconnecting their own facilities closer to the customer, their ability to provide services using their own facilities will be greatly enhanced, thereby furthering the goal of the 1996 Act to promote facilities-based competition. We find that the availability of unbundled NIDs will accelerate the development of alternative networks, because it will allow requesting carriers efficiently to connect their facilities with the incumbent's loop plant. Thus, our decision to unbundle NIDs is consistent with the 1996 Act's goals of rapid introduction of competition and the promotion of facilities-based entry. We recognize that there may be situations where a competitive LEC could successfully self-provision NIDs. We find, however, that the benefits of unbundling the NID on a nationwide basis outweigh the costs of creating a patchwork regime in which incumbents will seek to litigate whether particular NIDs should be unbundled or whether an alternative to the incumbent LEC's NID is arguably available as a practical, economic, and operational matter. A. Local Switching 1. Local Circuit Switching a. Background 254. In the Local Competition First Report and Order, the Commission concluded that incumbent LECs must provide local circuit switching as an unbundled network element. The Commission found that denying access to the local circuit switching element would "substantially impair the ability of many competing carriers to provide switched telecommunications services." 255. In the Notice, we sought comment on the application of the "necessary" and "impair" standards to previously identified unbundled network elements, including the switch. The Notice requested that parties include specific costs and an analysis of the availability of alternative sources of switching. 256. Incumbent LECs argue that a market-by-market analysis of the availability of local circuit switching requires a finding that requesting carriers are not impaired without access to unbundled switching in certain areas. Conversely, a number of requesting carriers argue that they are impaired without unbundled local circuit switching nationwide primarily because of the operational impairment associated with obtaining collocation and coordinated hot cuts. We note also that at least nine of the eleven states participating in this proceeding agree that the Commission should unbundle local circuit switching. a. Discussion (i) Definition of Local Circuit Switching 257. In the Local Competition First Report and Order, the Commission defined local circuit switching as including the basic function of connecting lines and trunks. In addition to line-side and trunk-side facilities, the definition of the local switching element encompasses all the features, functions and capabilities of the switch. With the exception of MCI WorldCom, no commenter proposes that we modify the current definition of local switching. We disagree with MCI WorldCom, and find no reason to alter our current definition of local circuit switching. 258. MCI WorldCom proposes that we modify our definition of local circuit switching to omit the line-side/trunk-side distinction in favor of a technologically-neutral definition that connects "loop access points" and "transport access points" to the "switching facility," regardless of whether a given switch has equipment that could be identified as line port cards or trunk port cards. MCI WorldCom suggests that we should take into account the increasing use of switches to connect to facilities other than home run copper loops, including DLCs. We cannot find, on the basis of the record before us, that incumbent LEC circuit switching technologies have changed in such a way as to warrant modification of our circuit switching definition. Furthermore, adopting MCI WorldCom's proposed changes could require state commissions to re-evaluate their current pricing analysis of unbundled circuit switching. Replacing the existing definition of switching with "loop access points" or the "switching facility" could lead to uncertainty and different cost determinations in state pricing proceedings. We find no procompetitive basis on which to require states to modify their settled state proceedings that have addressed forward-looking pricing for unbundled switching. Accordingly, we decline to modify our definition of local circuit switching. (i) Proprietary Concerns Associated With Local Circuit Switching 259. We conclude that incumbent LECs may not withhold access to switch routing tables as part of the unbundled local circuit switching element under section 251(d)(2)(A). With the exception of Ameritech, no commenter identifies any proprietary concerns associated with local circuit switching. Ameritech argues that, if we conclude that local switching qualifies as an unbundled network element, we should decline to require incumbent LECs to make their switch routing tables available to requesting carriers because these tables are "proprietary," within the meaning of section 251(d)(2)(A). According to Ameritech, routing tables are "part of the computer software that instructs a switch how to route network traffic," and contain "extremely valuable information" that is not "necessary" to a requesting carrier under section 251(d)(2)(A). Ameritech further argues that its routing tables meet the legal requirements for trade secret protection from unauthorized disclosure. 260. Ameritech contends that, because it has invested resources in creating economic value in its routing tables, and it takes reasonable steps to protect its routing tables from unauthorized disclosure, switch routing tables meet the general requirements for trade secret protection. No opposing party addresses whether Ameritech's routing table meets the legal requirements for trade secret protection. CompTel avers that switch routing tables merely perform a function that allows a switch to operate in a network, and as such, switch routing tables are not proprietary and should be included within the unbundled switching element. On the basis of Ameritech's uncontested assertion that its routing tables qualify for trade secret protection, we find that Ameritech's routing table may qualify for trade secret protection. Thus, it appears that the routing aspect of the local circuit switching element may be proprietary. 261. The "Necessary" Standard. As previously discussed, there are several circumstances which, if they exist with respect to information or functionalities that the incumbent LEC claims are proprietary, will permit us to order unbundling of the proprietary information or functionality even if such unbundling is not strictly "necessary." Access to the incumbent LEC's routing tables may be necessary for some carriers because they would be precluded, as an economic matter, from providing service to certain classes of customers if they were forced to self- provision their own routing tables. Requesting carriers would be economically precluded from providing service because of the costs associated with developing their own routing tables and the additional non-recurring and administrative costs of substituting dedicated transport unbundled network elements for shared transport. As we found in the Local Competition Third Reconsideration Order, the high costs that requesting carriers would incur without access to shared transport would deter entry and impose significant costs on new entrants without any corresponding, direct benefits. 262. Even if it is not strictly necessary for all carriers to have access to the incumbent LECs' routing tables, we find that they should be required to unbundle them because two of the three circumstances that we identified previously exist, and because requesting carriers would be impaired without access to routing tables as part of the unbundled local circuit switching element. 263. Specifically, we find that it is unlikely that Ameritech will compete for end-user customers based on the ability to send a call to an appropriate destination, or that its routing tables allow it to differentiate its services from its competitors' services. As we stated above, information or functionalities that do not distinguish an incumbent LEC's service from that of its competitor's services are unlikely to be the focus of an incumbent LEC's efforts to innovate, and therefore do not require the higher level of protection normally afforded to proprietary elements under the "necessary" standard. 264. Moreover, we find that incumbent LECs may not withhold access to switch routing tables as part of the unbundled local switching element because doing so would jeopardize the goal of the 1996 Act to bring rapid competition to the greatest number of customers. One of the most essential functions a switch performs is to provide routing information that sends a call to the appropriate destination. Requiring requesting carriers to engage in the potentially lengthy process of compiling traffic studies and populating routing tables with data in the incumbent LEC's unbundled switch would frustrate a requesting carrier's ability to use unbundled local circuit switching to serve customers quickly. 265. As described below, we conclude that carriers would be impaired without access to routing tables as part of the unbundled local circuit switching element. Requesting carriers have not generally deployed self-provisioned local circuit switches to serve the mass market. We conclude that requesting carriers are impaired without access to unbundled local circuit switching to serve certain customer classes in discrete geographic areas. We therefore order incumbent LECs to provide access to unbundled local circuit switching in these circumstances. We note that requesting carriers will request unbundled access to local circuit switching when, in the judgment of the requesting carrier, the costs and delays associated with self- provisioning switching do not warrant purchase and installation of a local circuit switch. Accordingly, because of the circumstances identified above, and because requesting carriers would be impaired without access to routing tables as part of the local circuit switching element, we find that incumbent LECs may not, pursuant to section 251(d)(2)(A), withhold access to switch routing tables. (i) General Unbundling Analysis for Local Circuit Switching 266. We conclude that, as a general matter, unbundled local circuit switching meets the "impair" standard set forth in section 251(d)(2). Accordingly, we require incumbent LECs to provide local switching as an unbundled network element. Based on the record, we find that, in general, lack of access to unbundled local switching materially raises entry costs, delays broad-based entry, and limits the scope and quality of the new entrant's service offerings. As discussed in detail below, our unbundling analysis focuses upon the ability of a requesting carrier to self-supply switching because the record does not support a finding that requesting carriers, as a general matter, can obtain switching from carriers other than the incumbent LEC. We find, however, that an exception to this rule is required under certain market circumstances. We find that, where incumbent LECs have provided nondiscriminatory, cost-based access to combinations of loop and transport unbundled network elements, known as the enhanced extended link (EEL), requesting carriers are not impaired without access to unbundled switching for end users with four or more lines within density zone 1 in the top 50 metropolitan statistical areas (MSAs). 267. Alternatives Outside the Incumbent's Network. As of March 1999, approximately 167 different competitors have deployed approximately 700 switches throughout the country. According to USTA, approximately 320 cities are served by at least one competitive switch. SBC, using a methodology that tracks requesting carriers' switches by examining migration of lines using ported numbers, contends that within the 50 largest MSAs, competitors' switches currently serve approximately 75 percent of all BOC and GTE rate exchange areas. Although certain requesting carriers argue that incumbent LEC statistics are not precise, the record indicates that a significant number of competitive switches have been deployed. Our examination of switching investments in the market shows that requesting carriers have self-provisioned a significant number of switches, but that this investment represents only a small fraction of the number of switches deployed by the incumbent LECs. 268. Since the Commission adopted the Local Competition First Report and Order, competition has continued to develop in certain geographic markets, particularly for large business customers or other users with substantial telecommunications needs. The pattern of switch deployment by competitors suggests that the costs and operational delays of self-provisioning switching do not preclude requesting carriers from serving certain customer classes in certain geographic markets. In general, however, we conclude that requesting carriers are impaired in their ability to provide service in most markets, primarily because of the costs of self- provisioning switching in those markets. We find that section 251(d)(2)(B) requires consideration not simply of whether denial of access to unbundled switching would impair a competitor's ability to serve the high-volume business market that many requesting carriers are already serving, but whether the requesting carrier is impaired in its ability to provide the "services that it seeks to offer," including services to residential and small business markets. Although the groundwork for residential local competition is evolving, and competition, to date, has focused upon users with substantial telecommunications needs, we do have some evidence that some requesting carriers will seek to offer residential phone service to the mass market where unbundled switching is available. Accordingly, we find that our unbundling analysis should take into account the possibility that carriers will offer residential service. We find that, taking into account the cost, quality, ubiquity and timeliness factors in our "impair" standard as well as the goals of the Act, lack of access to unbundled switching as a general matter, impairs the ability of a requesting carrier to provide service to consumers. 269. Incumbent LECs use the geographic dispersion of deployed local circuit switches to argue for a geographic or market-specific approach to circuit switch unbundling. Certain incumbent LECs further argue that the presence of one competitor's switch and collocation in a given market is dispositive of whether requesting carriers generally will be impaired without access to unbundled switching. We reject this argument. Just as the Supreme Court made clear that the "impair" standard is not triggered by any increase in cost or decrease in quality, we find that switch unbundling cannot turn on whether a single carrier has self-provisioned switching. The fact that a single carrier is collocated in a particular central office and is not using unbundled switching does not conclusively demonstrate that a variety of carriers can self-provision switches without significant cost or other impediments that diminish a collocating carrier's ability to provide the services it seeks to offer. Indeed, based on financial analysts' reports of competitive LECs' operations, a significant number of requesting carriers currently self-provisioning switches are not generating net income (i.e., profits). Thus, it is too early to know whether self- provisioning is economically viable in the long run, although capital markets appear to be supplying requesting carriers with access to capital in the absence of demonstrated profitability. 270. Incumbent LECs have provided business case analyses that purport to demonstrate that a requesting carrier could expect to earn profits upon entry using self-provided switching by comparing the revenues that could be expected from self-provisioning switching with the full costs of entry. As discussed in Section IV above, we favor an analytical approach that considers the totality of the circumstances a requesting carrier will face, rather than a specific business case analysis, to determine whether lack of access to particular network elements materially diminishes a requesting carrier's ability to provide the services it seeks to offer. Adopting a business case approach would require the Commission to conduct a detailed analysis of the profitability of entry for a representative firm using various business strategies in each possible market. Such an approach would also require the Commission to make specific assumptions regarding the competitor's business model, including which technology a competitor would choose to deploy, which market a competitor would choose to enter (e.g., business and/or residential), and what services a competitor would choose to offer. In addition, a business case approach would require the Commission to forecast revenues that a requesting carrier would earn if it entered a particular market. Whereas the actual costs of network elements such as switches are quantifiable, revenues may fluctuate according to evolving competitive conditions in the local telecommunications market. 271. Although we do not adopt an approach that is based on a business case analysis for determining whether a particular entity is impaired, we do make several general observations regarding the direct cost that bear on a requesting carrier's decision to self-provision a switch. Fixed costs are the largest portion of the cost of a switch. The average cost of providing service to customers decreases as the number of customers served increases. As a general rule, we find that scale economies are more pronounced when switches operate at full utilization. Because incumbent LEC switches serve the majority of customers for local exchange service, they are likely to be able to take advantage of substantially greater economies of scale than the competitor would using its own switches. We find however, that facilities- based competitors need not deploy switches in exactly the same network configuration as an incumbent, thus allowing competitors to achieve their own unique and competitive efficiencies by deploying their own switches. 272. Cost. We find, as a general matter, that the total costs of self-provisioning a switch impose on the requesting carrier a significant cost disadvantage relative to the incumbent LEC, particularly in its early stages of entry. We emphasize that cost is only one factor we examine in our "impair" analysis. The evidence of circuit switching direct costs submitted in the record varies significantly. For example, incumbent LECs provide evidence that the direct costs to competitors of self- provisioning switches is between $100,000 and $814,000 and that the incremental cost is between $110 and $146 per line. AT&T counters that incumbent LEC models exaggerate the efficiencies associated with requesting carrier switches. Independent sources, however, estimate the fixed cost per host switch to be $447,000 and the per-line cost to be $83. The disparity in switching costs contained in the record appear to depend on the technical attributes of the switch at issue. The more critical aspect of our "impair" analysis is not the costs of purchasing a local circuit switch, but rather the economies of scale that may characterize local circuit switching and the additional costs that requesting carriers incur when placing their self-provisioned switches into operation. 273. We find that incumbent LECs retain material scale advantages with regard to provisioning and operating local circuit switches. Requesting carriers therefore will encounter generally greater direct costs per subscriber when provisioning their own switches, particularly in the early stages of entry when requesting carriers may not have the large number of customers that is necessary to increase their switch utilization rates significantly. When we examine the market as a whole, we find that requesting carriers incur higher costs due to their inability to realize economies of scale using circuit switching equipment. We find that the scalability of a switch mitigates but does not eliminate the incumbent LEC's scale advantages and reduces but does not eliminate competitor's sunk costs and entry barriers. For example, it is generally less expensive to purchase a 20,000 line switch rather than four increments of 5000 lines. Furthermore, the advantages of incumbent LEC scale economies are more pronounced when requesting carriers provide switch-based service to a relatively small number of customers through a self-provisioned switch. For example, competitor's switching costs per minute at a 10% penetration level are slightly more than twice the cost of an incumbent LEC serving the remaining 90% of the market with its own switch. We find that, as a general proposition, requesting carriers will incur a materially greater cost when self- provisioning switching at low penetration levels. As a requesting carrier's switch utilization rates increase, the difference between the switching costs incurred by competitive and incumbent LECs decreases, but the impact of this difference does not become irrelevant in the impair analysis until incumbent LEC and competitor's switch utilization levels are more comparable. Market facts show that that competitors have made inroads into the local telecommunications markets, but they have garnered only between 2.6 percent to 5 percent of the market for switched telecommunications services. A significant portion of these figures represent service to medium and large business customers, rather than to the mass market. Accordingly, we find that as a general matter, requesting carriers have not gained sufficient market share to generate switch utilization rates and economies of scale comparable to the incumbent LEC, particularly to serve the mass market. 274. We recognize that switches deployed by competitive LECs may be able to serve a larger geographic area than switches deployed by the incumbent LEC, thereby reducing the direct, fixed cost of purchasing circuit switching capacity and allowing requesting carriers to create their own switching efficiencies. If a competitor uses a single switch to serve a rate area consisting of 10-15 incumbent LEC switches, the average utilization of the competitor's one switch can be as high, or higher, than many, or even all, of the incumbent LEC switches. This dynamic mitigates, to a varying degree, incumbent LEC advantages of scale, but does not enable competitive LECs to achieve comparable scale economies, particularly in the early stages of entry. Incumbent LECs contend that once a requesting carrier incurs the costs to deploy a switch, it can economically extend the reach of the switch to serve broader markets. We find, however, that switch capacity, distance-sensitive transport costs, and collocation costs significantly impair a requesting carrier from fully exploiting this market entry strategy. We note that, for smaller carriers, an inability to achieve switching scale economies may have greater effect upon their ability to offer service than it does for larger carriers. For example, TRA contends that, without access to unbundled switching, smaller requesting carriers with targeted entry plans deploy their own switch to serve approximately 3,000 lines will incur a direct additional cost of $300,000 annually without access to unbundled local switching. We find that utilizing unbundled switching is likely to mitigate this early-stage entry barrier and is consistent with Congress' intention that requesting carriers use unbundled network elements as a transitional market entry strategy. 275. We find, as a general matter, that the costs of self-provisioning switching also materially diminish a requesting carrier's ability to provide the services it seeks to offer. Our standard recognizes that the full costs of using self-provisioned circuit switching must include the costs incurred by a competitor to substitute its local circuit switch for that of the incumbent LEC. These costs include the costs of collocating in an incumbent LEC's central office from which the requesting carrier accesses unbundled loops to serve its end-user customers. Requesting carriers require collocation because they have not yet duplicated the incumbent LEC's loop plant to provide "last mile" connectivity to end users. Obtaining unbundled loops and connecting these loops to collocated equipment is therefore the only reasonable and economically rational manner by which requesting carriers can provide connectivity to their end users. 276. We agree with parties that argue that collocation imposes materially greater costs on requesting carriers than use of the incumbent LEC's switching. Based on the record, it appears that the current range for non-recurring charges for obtaining physical collocation is between $15,000 and $508,000 for each central office from which a competitor serves customers using the incumbent LEC's unbundled loops. This additional cost increases the costs of the equipment installed in the cage by between 15 to 20 percent. We find that due in part to these non-recurring charges, a requesting carrier's decision to collocate presumes significant market penetration, even in dense wire centers. For example, data submitted by MCI WorldCom for New York City suggests that collocation in dense wire centers is not profitable until a requesting carrier's market penetration exceeds 8-15 percent. We agree that the costs associated with collocation and the revenue opportunities associated with a given wire center may not justify establishing a collocation arrangement with the incumbent LEC in many central offices. 277. Accordingly, we find that as a general matter, collocation costs materially diminish the ability of a requesting carrier to offer service using self-provisioned switching. Although the collocation rules we adopted in our Advanced Services First Report and Order and FNPRM are intended, and expected, to reduce the costs and operational delays associated with collocation, our revised rules do not eliminate altogether the cost and delay associated with collocation. We recognize that incumbent LECs still have an incentive and the ability to raise a requesting carrier's cost of collocation, and thus raise the total cost of self-provisioning switching. 278. In addition to the costs of establishing collocation arrangements with the incumbent LEC, requesting carriers incur additional costs to extend unbundled loops to their collocation cage. The manual work of extending a loop to a requesting carrier's collocation cage is known as a coordinated loop cutover. A coordinated loop cutover requires incumbent LEC technicians to disconnect the subscriber's loop from the incumbent LEC's main distribution frame and rapidly cross-connect it to the competitor's facilities. From the time the technician disconnects the subscriber's loop until the competitor re-establishes service, the subscriber is without service. Simultaneously, incumbent LEC and competitor technicians must coordinate to ensure that the subscriber's telephone number is "ported" to the competitor's switch so that inbound calls are properly routed to the requesting carrier's switch. 279. The coordinated cutover process imposes a non-recurring cost on competitive carriers that connect their own switches to unbundled loops. For example, AT&T contends that the non-recurring, per-line charge for a coordinated cutover is approximately $45 in New York. CompTel argues that a manual loop and switching port migration costs between $59.91 and $218.62 per unbundled loop. We acknowledge that incumbent LECs may incur coordinated cutover costs when they win customers from competitive LECs. The record does not demonstrate, however, that incumbent LECs are incurring coordinated cutover costs in the same or substantially similar volumes as competitive LECs. We find that the additional cost of coordinated loop cutovers, when added to the costs of collocation, materially diminishes a competitor's ability to substitute its own switch for unbundled switching. Although this per-line non-recurring cost is likely to vary between incumbent LECs, it represents a significant cost to those requesting carriers seeking to provide service to the mass market due to the large number of individual loop cutovers that are necessary to serve this market. 280. Ubiquity and Timeliness. In addition to the costs associated with accessing individual unbundled loops in multiple end offices, we find that collocation and the coordinated loop cutover process imposes a material delay on competitive LECs that offer services using self-provisioned switches, and materially limits the scope of customers a requesting carrier may serve quickly. The delay includes the total amount of time required to purchase, install, turn up a switch, and obtain collocation, as well as the amount of time needed for incumbent LECs to complete coordinated loop cutovers. 281. In order to self-provision a switch, a requesting carrier must order, test the switch, and integrate it into its network and internal operations support systems. Incumbent LECs claim that a switch can be fully provisioned in as little as 40 days. Although this may be theoretically possible, there is evidence in the record that the time frame for provisioning a switch is significantly longer. Furthermore, incumbent LECs focus their analysis on the time to purchase a switch rather than the time required to put a switch into operation. Actual delivery of a switch is only one part of the process of self-provisioning switching. Requesting carriers assert that it takes approximately six months to one year to engineer, furnish and install a switch. 282. Related to the time required to provision a switch is the time required to provision collocation. Incumbents and requesting carriers offer conflicting assertions regarding the time required to provision collocation in incumbent LEC central offices. We are troubled by anecdotal evidence that collocation imposes a delay of six, nine or twelve months on the provision of ubiquitous service. MCI WorldCom for example, argues that collocating on a broad scale to provide ubiquitous service results in lengthy collocation delays. NorthPoint maintains that some incumbent LECs have imposed "governors" on the number of collocation applications they will accept, thereby delaying ubiquitous rollout of services. Incumbent LECs counter that they have, and will continue to, provision collocation on a broad scale and in a timely fashion. Ameritech specifically contends that competitive LECs have established collocation arrangements in rate centers in which 70 percent of Ameritech's access lines are located. The presence of one collocator, in and of itself, however, does not establish how long it will take to accommodate subsequent collocators. We find nothing in the record to demonstrate conclusively that incumbent LECs have committed to and satisfied a collocation provisioning interval of less than six months. 283. We therefore find incumbent LEC arguments that requesting carriers do not experience collocation delays contradicted by the actual experiences of requesting carriers. Incumbent LECs do not appear to include such things as the collocation application process in their analysis of collocation delays. We are persuaded by those commenters that assert that collocation, examined from the time a requesting carrier initiates the collocation process until a collocation arrangement is delivered, generally imposes a delay of approximately six months on the provision of service. We conclude that, although the delays associated with provisioning collocation arrangements will vary from incumbent LEC to incumbent LEC and by requesting carrier, as a general matter, collocation delays materially diminish the ability of a requesting carrier to provide the services it seeks to offer. As discussed in Section IV above, although we cannot quantify precisely how much of a delay associated with collocation and self-provisioning switching will materially diminish the ability of a competitor to provide the services it seeks to offer, we find that delays that exceed six months to one year materially diminish the ability of a competitive LEC to provide the services it seeks to offer because such delay prevents the competitive LEC from responding quickly to the demand for its services in a rapidly changing market. 284. As noted above, requesting carriers must also wait for coordinated cutovers before providing service with their own switch. We disagree with BellSouth, GTE, Ameritech, and other commenters that argue that the Commission should not consider coordinated cutover delays and service-quality issues in its impair analysis. Without coordinated loop cutovers, requesting carriers self-provisioning switching and accessing unbundled loops cannot provide the services they seek to offer. To date, incumbent LECs have provisioned relatively small volumes of coordinated loop cutovers compared to anticipated demands. Incumbent LECs counter that they have instituted procedures to provide timely coordinated cutovers to requesting carriers. Where incumbent LECs have undergone comprehensive testing of their loop provisioning processes, however, independent auditors have found difficulties regarding coordinated loop cutover performance. Furthermore, because broad-based residential competition is at best nascent, incumbent LECs generally have not successfully provisioned coordinated loop cutovers in the volumes necessary for requesting carriers to serve the mass market. We therefore find incumbent LEC promises of future hot cut performance insufficient to support a Commission finding that the coordinated loop cutover process does not impair the ability of a requesting carrier to provide the service it seeks to offer without unbundled circuit switching. We recognize that the hot cut process requires manual processing, which likely creates delays between the time a requesting carrier wins a customer and the time it can provide service to that customer. Accordingly, we find that the coordinated loop cutover process impairs the ability of a requesting carrier to provide timely service. 285. Goals of the 1996 Act. As noted above, our unbundling analysis takes into account whether unbundling a particular network element is consistent with the goals of the 1996 Act. We find our decision to unbundle local circuit switching is consistent with the 1996 Act's goals of rapid introduction of competition and the promotion of facilities-based entry. 286. Our unbundling analysis considers how the switch unbundling obligation we adopt will encourage requesting carriers to rapidly enter the local market in order to serve the greatest number of customers, and whether the failure to require unbundling will cause any class of consumers to wait unnecessarily for competitive alternatives. Failure to unbundle local circuit switching would cause residential and small business consumers to wait unnecessarily for competitive alternatives. As discussed above, the costs and operational delays associated with collocating in multiple end offices and provisioning delays caused by the inability of a requesting carrier to gain access to unbundled local circuit switching will cause residential and small business customers to wait for service. Requiring incumbent LECs to provide access to unbundled switching, and to use unbundled switching in combination with other network elements, will allow requesting carriers to serve the greatest number of customers, without incurring collocation and switch provisioning delays. Where unbundled switching has been made available, requesting carriers have gained market share in the residential and small business markets. Accordingly, we find that requiring incumbent LECs to provide access to unbundled switching will allow requesting carriers to rapidly enter local markets. 287. We also find that the availability of unbundled switching will also accelerate the development of alternative networks because it will allow requesting carriers to generate revenues to justify the construction of new switching facilities. As noted above, many carriers emphasize that they plan to deploy alternative facilities as soon as it is technically and economically possible to do so at a cost close to the incumbent LECs' prices for network elements. Granting requesting carriers access to unbundled switching will allow these carriers to serve customers in areas where traffic volumes and customer densities make it difficult initially to justify deploying a switch. Furthermore, allowing requesting carriers to purchase unbundled switching will allow new entrants to test market demand for circuit switched services before deploying their own facilities. As requesting carriers obtain customers using unbundled switching, we expect that the revenues generated from this activity will enable requesting carriers to extend the reach of their existing switching capabilities or deploy switching capability to serve the residential and small business market. 288. On balance, we conclude that local circuit switching should be unbundled nationwide. We now consider whether it would be appropriate to establish an exception to the national unbundling requirement. (i) Exception to National Unbundling Requirement 289. As discussed in section IV above, we do not limit our unbundling analysis to the cost, timeliness, ubiquity and quality factors described above. Rather, we look at the totality of the circumstances and marketplace developments when considering whether a requesting carrier is impaired without access to unbundled local circuit switching. In addition to examining where requesting carriers have deployed switches, we look to the marketplace to see which customers are receiving service from facilities-based competitors. To the extent the market shows that requesting carriers are not serving a market segment with self-provisioned switches, we find that this fact is probative evidence that for a discrete market segment requesting carriers are impaired without access to unbundled local circuit switching. Conversely, to the extent that the market shows that requesting carriers are generally providing service in particular situations with their own switches, we find this fact to be probative evidence that requesting carriers are not impaired without access to unbundled local circuit switching. The task before us is to develop an administratively simple rule that reflects marketplace developments and provides certainty to market participants. We seek to adopt a rule that serves as a reasonable proxy for when competitors are indeed impaired in their ability to provide the services they seek to offer. 290. In their initial and reply comments in this proceeding, the parties take sharply diverging positions regarding the circumstances and geographic areas where local circuit switches should be unbundled, if at all. Incumbent LECs generally support elimination of their obligation to unbundle local circuit switches in a geographic area where one requesting carrier has deployed a single local circuit switch. Competitive LECs oppose the incumbent LEC proposals for elimination of the circuit switch unbundling obligation and argue that local circuit switching should be unbundled on a national basis. In several ex parte presentations after the record closed, a number of parties softened their initial positions and propose a more narrowly tailored rule for determining when circuit switching need not be unbundled. A number of other parties respond to these fall-back positions in subsequent ex parte presentations. 291. Despite our conclusion that, in general, requesting carriers are impaired without access to unbundled switching, we conclude that it is appropriate to establish a more narrowly tailored rule to reflect significant marketplace developments. As described more fully below, we find that requesting carriers are not impaired without access to unbundled local circuit switching when they serve customers with four or more lines in density zone 1 in the top 50 metropolitan statistical areas (MSAs), as set forth in Appendix B, where incumbent LECs have provided nondiscriminatory, cost-based access to the enhanced extended link (EEL) throughout density zone 1. 292. Top 50 MSAs. We conclude that it is appropriate to create an exception to the switching unbundling obligation in certain circumstances in the top 50 MSAs, as they are defined by the Office of Management and Budget. We thus respond to various suggestions in the record that an exception from the switching unbundling obligation should encompass the top 29, top 35 and top 100 MSAs in the United States. 293. As previously noted, as of March, 1999, approximately 167 different competitors have deployed approximately 700 switches throughout the country. When we analyze where requesting carriers have deployed these switches, we find that most of these switches have been deployed within the confines of the top 50 MSAs. According to USTA's data, which relies on the Local Exchange Routing Guide, approximately 61 percent of all requesting carrier switches nationwide have been deployed in the top 50 MSAs. More significantly, the vast majority of these MSAs contain multiple switches owned by competitors. In particular, four or more competitive switches have been deployed in 96 percent of the top 50 MSAs. According to USTA's data, only two MSAs in the top 50 -- Cincinnati and Las Vegas -- have less than three requesting carrier switches serving an incumbent LEC rate exchange area within the MSA. 294. Based on the evidence in the record, we conclude that exempting incumbent LECs from unbundling local circuit switching in certain circumstances in the top 50 MSAs is reasonable because nearly all of the top 50 MSAs contain a significant number of competitive switches. In contrast, MSAs below the top 50 typically contain fewer competitive switches. For example, in US WEST's territory, no MSA between 50 and 150 contains more than three competitive switches. In the top 100 MSAs in Ameritech's territory, only six percent of Ameritech's wire centers are served by four or more competitive switches. 295. We recognize that drawing the line at the top 50 MSAs means that incumbent LECs serving more rural territories, which have fewer MSAs that are in the top 50 MSAs, will continue to be subject to an unbundled switching obligation. We nonetheless believe that this is a reasonable exercise of our administrative discretion. Extending an incumbent LEC's switch unbundling exemption to include more than the top 50 MSAs would require us to find that requesting carriers are not impaired without unbundled access to local circuit switching in these MSAs. We have no basis in the record before us to make such a finding because there are relatively few competitive switches outside of the top 50 MSAs. 296. We note that collocation costs and delay, as compared to revenue potential, may contribute to the relative lack of robust competitive switch deployment in areas outside of the top 50 MSAs. As discussed above, the total costs of a competitor using self-provisioned local circuit switching on an MSA basis include the costs incurred in providing service to every customer that the competitor seeks to serve. We concluded above that collocation imposes indirect costs on carriers installing their own switches. We also found that the amount of collocation cost are likely to vary according to individual requesting carriers. We believe that the revenue potential of serving less dense markets outside the top 50 MSAs is unlikely to outweigh the costs of collocating in these markets, and accordingly, competitors are impaired without access to unbundled local switching. 297. Density Zone 1. When we examine the deployment of switches by competitors at a more granular level, we find that, based on the record before us, requesting carriers have deployed greater numbers of switches in areas of high customer density. Several incumbent LECs argue that switching should not be unbundled in dense wire centers, but each proffers its own geographic market definition for our local circuit switch unbundling analysis. BellSouth proposes, and other incumbent LECs support, the use of density zones 1 and 2 to capture the areas in which competitors have deployed switches and where incumbent LECs need not unbundle switching. 298. We conclude that it is appropriate to create an exception to the local circuit switching unbundling obligation only in density zone 1, within the top 50 MSAs. The exception applies to density zone 1 as it was defined on January 1, 1999. Based on the limited evidence in the record, we believe that density zone 1 closely reflects the wire centers where competitive LEC switches are located. In particular, of the seven markets in the top 50 MSAs served by BellSouth, each MSA contains at least one density zone 1 where approximately 97 percent of all competitive LEC switches have been deployed. We recognize that only one commenter, BellSouth, provided detailed data to describe where requesting carriers have deployed switches in density zone 1. The record does not contain similar data for other incumbent LECs. Given the record before us and the need to provide a measure of certainty to the market, we believe that drawing a line at density zone 1 within the top 50 MSAs represents a reasonable approximation of where requesting carriers are not impaired without access to unbundled local circuit switching. 299. In order to prevent incumbent LECs from modifying their density zones to limit their unbundling obligation for local circuit switching, we freeze, for unbundling purposes, the incumbent LECs' density zone 1 as it was defined on January 1, 1999. Otherwise, incumbent LECs would retain significant discretion to define their density zone boundaries in the future. The Commission reviews incumbent LEC zone density pricing plans under a "reasonableness" standard. For example, our rules allow incumbent LECs to define zone boundaries upon a showing that "the assignment of central offices to each of the zones reflects cost-related characteristics, such as traffic density or some measure of traffic through each office." MCI WorldCom argues that using the zone approach would allow incumbent LECs to "redefine breakpoints to put more central offices into zones in which the incumbent LECs were not required to provide switching as an unbundled network element" and would allow incumbent LECs to "change their methodologies for defining zones to upset their competitor's business plans." To address the possibility that incumbent LECs, going forward, could amend their density zones to minimize their unbundling obligations, we create an exception to the unbundling obligation in the density zones as they existed on January 1, 1999. We believe that freezing the zones as of January 1, 1999, for purposes of section 251 unbundling obligations, addresses MCI WorldCom's concerns. 300. As discussed in our unbundling analysis above, as requesting carriers' switch utilization rates increase, the difference between the switching costs incurred by competitive and incumbent LECs decreases, and the per line switching costs will decrease as a requesting carrier's customer base grows. Because of increased demand for telecommunications services and the enhanced revenue opportunities associated with serving customers in high-density areas, such as density zone 1, we find that requesting carriers serving these dense areas are able to make more efficient use of their switching facilities, and can thus counter incumbent LEC scale economies. We therefore find that the cost of purchasing a circuit switch does not impair a requesting carrier's ability to provide the services it seeks to offer in density zone 1, in certain circumstances. 301. Need for Enhanced Extended Link. Our conclusion that competitors are not impaired in certain circumstances without access to unbundled switching in density zone 1 in the top 50 MSAs also is predicated upon the availability of the enhanced extended link (EEL). As noted in section VI(B) above, the EEL allows requesting carriers to serve a customer by extending a customer's loop from the end office serving that customer to a different end office in which the competitor is already collocated. The EEL therefore allows requesting carriers to aggregate loops at fewer collocation locations and increase their efficiencies by transporting aggregated loops over efficient-high capacity facilities to their central switching location. Thus, the cost of collocation can be diminished through the use of the EEL. We agree with ALTS that, if requesting carriers can obtain nondiscriminatory, cost-based access to the enhanced extended link, their collocation costs would decrease, and they would need to collocate in as few as one incumbent LEC central office in an MSA to provide service. 302. We are not persuaded by arguments that use of the EEL produces only a short term advantage over collocation. Although we agree with SBC that distance-sensitive EEL costs can exceed the costs associated with collocation over time, we find that the ability of a requesting carrier to provision EELs more quickly than collocation arrangements, without the substantial upfront costs of establishing collocation in multiple central offices, can reduce significantly the costs of self-provisioning a switch in the initial phase of an entry strategy. When projected EEL costs exceed projected collocation costs, competitive LECs may reconfigure their networks to ensure the continued efficiency of their networks. We conclude that requesting carriers, reacting to marketplace demands and their own network topologies, are better able to weigh the costs and benefits of EELs compared to collocation and adjust their plans accordingly. Where a requesting carrier chooses the EEL, we find that it reduces a requesting carrier's reliance on collocation. 303. Customers with Four or More Lines. Our analysis of an incumbent LEC's local circuit switching obligation has focused primarily upon the geographic areas where competitive carriers have deployed switches. We now consider whether, within these geographic areas, market facts demonstrate that requesting carriers are not impaired without access to local circuit switching for discrete market segments or customer classes. 304. We conclude that without access to unbundled local circuit switching, requesting carriers are impaired in their ability to serve the mass market. As discussed above, our unbundling analysis takes into account market conditions to determine whether a requesting carrier is impaired without access to unbundled local circuit switching. Since the Commission adopted the Local Competition First Report and Order, competition has continued to develop, primarily for business customers or users with substantial telecommunications needs. Commenters in this proceeding generally argue that requesting carriers have deployed switches to serve medium and large business customers and are not yet serving mass market customers, which largely are residential customers. No party in this proceeding, however, identifies the characteristics that distinguish medium and large business customers from the mass market. 305. There are several methods we could use to distinguish between the mass market and the medium and large business market for purposes of our unbundling analysis. For example, we could use revenues, number of employees, number of lines, or some other factor to distinguish between the mass market and the medium and large business market. 306. We find, however, that a rule that provides access to unbundled local switching for requesting carriers when they serve customers with three lines or less captures a significant portion of the mass market. First, virtually all residential customers would be captured by such a rule. While an increasing number of American homes are served by second lines, we believe it is a rare case in which residences have three lines, and even more unusual for a home to have four or more lines. Second, any business that has three or fewer lines is likely to share more characteristics of the mass market customer than a medium and large business. In particular, small businesses are likely to use the same number of lines as many residential subscribers and purchase similar volumes and types of telecommunications services. 307. We recognize that a rule that removes unbundling obligations based on line count will be marginally overinclusive or underinclusive given individual factual circumstances. We find, however, that in our expert judgment, a rule that distinguishes customers with four lines or more from those with three lines or less reasonably captures the division between the mass market where competition is nascent and the medium and large business market where competition is beginning to broaden. 308. Our decision to examine mass market and larger business markets separately is consistent with the Commission's merger review analysis and the Commission's reform of the interstate access charge regime. In the MCI-WorldCom merger, we identified two distinct product markets residential and small business, which we described as one market, and medium and large business customers, which we described as the larger business market. In the Access Reform proceeding, the Commission distinguished between primary residences and single line businesses which constitute a large portion of the mass market, and multi-line business customers which constitute the medium and large business markets. We therefore conclude that it is appropriate to make a similar distinction between mass market customers and larger business customers in creating an exception to the unbundling obligation for local circuit switching. 309. As discussed above, a requesting carrier is materially diminished in its ability to offer service to mass market customers without access to unbundled switching because it will face materially greater costs, materially greater delay, and will lack the same ubiquitous reach as the incumbent LEC's network. In addition to the costs of establishing a collocation arrangement with the incumbent LEC, we noted above that requesting carriers incur additional costs and face service quality impediments when extending a customer's loop to their collocation cages. 310. In contrast, marketplace developments suggest that competitors are not impaired in their ability to serve certain high-volume customers in the densest areas. We believe that the coordinated cutover process will not necessarily impair the ability of a requesting carrier to serve an end user in density zone 1. Medium and large business customers are often sophisticated users of telecommunications services that are able to order their operations in a manner that minimizes disruptions that may be caused by coordinated cutovers. For example, requesting carriers seeking to provide service to medium and large business customers may engage in direct outbound marketing in such a way as to control coordinated cutover order flows to the incumbent LEC. In addition, to the extent that incumbent LECs provide requesting carriers with unbundled switching to serve the mass market, requesting carriers will require fewer coordinated loop cutovers in the aggregate and incumbent LECs can focus their efforts on coordinated cutovers for customers not served with unbundled local circuit switching. Finally, because business customers generate comparably greater revenues than residential customers, requesting carriers may be more willing to incur the provisioning difficulties that may be present in the coordinated cutover process. 311. We conclude that carriers will not be impaired in their ability to serve high volume users only when the EEL is provided throughout density zone 1. While some customers in this area already are being served by facilities-based carriers without the EEL, the availability of the EEL will ensure that requesting carriers are able to serve customers ubiquitously throughout the area. If the EEL is available and a requesting carrier seeks to serve a high volume business, the incumbent LEC can provision the high capacity loop and connect directly to a requesting carrier's collocation cage. In this scenario, the requesting carrier need not initiate a coordinated loop cutover. Moreover, the availability of the EEL substantially reduces the delay a requesting carrier would experience before it is able to actually provide service. 312. Goals of the 1996 Act. As noted above, our unbundling analysis considers how the switching unbundling obligation we adopt will encourage requesting carriers to rapidly enter the local market and whether the failure to require unbundling will cause any class of consumers to wait unnecessarily for competitive alternatives. Our decision to relieve incumbent LECs of their unbundling obligations in the circumstances described above will not require medium and large businesses to wait unnecessarily for competitive alternatives. We find that requesting carriers have deployed a large number of switches to serve medium and large business customers in the densest areas of the top 50 MSAs, and these medium and large business customers by and large, have a choice in their local service provider. Accordingly, we find that relieving incumbent LECs of their unbundled switching obligation, as set forth herein, will not require medium and small business consumers to wait unnecessarily for competitive alternatives because they are largely available today. Furthermore, eliminating an incumbent LEC's local circuit switching obligation in these circumstances is consistent with our goal to reduce regulation when possible. Our decision also provides requesting carriers with access to the elements they need to ramp up towards continued deployment of self- provisioned switches and is therefore consistent with our policies of encouraging facilities-based competition and encouraging innovation. 1. Packet Switching a. Background 313. In the Local Competition First Report and Order, the Commission declined to find that incumbent LEC packet switches should be identified as unbundled network elements because the Commission did not have an adequate record to support such a conclusion. In the Notice, we sought comment on whether "packet switches should be unbundled pursuant to section 251(c)(3), and whether there is "any basis for treating network elements used in the provisioning of packet-switched advanced services any differently than those used in the provisioning of circuit-switched voice services." Incumbent LECs argue that they generally trail in the deployment of packet switches, and therefore should not be subject to unbundling requirements that might eliminate their incentives to invest in equipment used to provide advanced services. Several competitors argue in favor of unbundling packet switching to encourage the broad-based deployment of advanced services. 314. We are aware, however, that US WEST has argued that section 251(c)(3) does not apply to any network elements, such as packet switches, used to provide advanced services, such as xDSL. We note that the Commission has requested, and has received, a remand from the United States Court of Appeals for the District of Columbia Circuit to address US WEST's argument that the Commission is without statutory authority to require incumbent LECs to provide access to unbundled elements used in the provision of advanced services. After receiving a more complete administrative record, we intend to fully address US WEST's arguments in the Advanced Services Memorandum Opinion and Order and NPRM remand proceeding. In remanding back to the agency, the court declined to vacate portions of the Advanced Services Memorandum Opinion and Order and NPRM challenged by US WEST. Accordingly, our decision in that Order that xDSL services are "either" telephone exchange service or exchange access service remains in effect during the pendency of the Advanced Services Memorandum Opinion and Order and NPRM remand proceeding. We therefore may consider whether packet switching should be unbundled under the framework established in this proceeding. a. Discussion (i) Definition of Packet Switching 315. As a threshold matter, we must define the functionality of the packet switching unbundled network element. In packet-switched networks, messages between network users are divided into units, commonly referred to as packets, frames, or cells. These individual units are then routed between network users. The switches that provide this routing function are "packet switches," and the function of routing individual data units based on address or other routing information contained in the units is "packet switching." 316. We find that a component of the packet switching functionality, and included in our definition of packet switching is the Digital Subscriber Line Access Multiplexer (DSLAM). The DSLAM splits voice (low band) and data (high band) signals carried over a copper twisted pair. DSLAM equipment sometimes includes a splitter. If not, a separate splitter device separates voice and data traffic. The voice signal is transmitted toward a circuit switch, and the data from multiple lines is combined in packet or cell format and is transmitted to a packet switch, typically ATM or IP. The DSLAM combines: (1) the ability to terminate copper customer loops (which includes both a low-band voice channel and a high-band data channel, or solely a data channel); (2) the ability to forward the voice channels, if present, to a circuit switch or multiple circuit switches; (3) the ability to extract data units from the data channels on the loops; and (4) the ability to combine data units from multiple loops onto one or more trunks that connect to a packet switch or packet switches. 317. We define packet switching as the function of routing individual data units, or "packets," based on address or other routing information contained in the packets. The packet switching network element includes the necessary electronics (e.g., routers and DSLAMs). We find that packet switching qualifies as a network element because it includes "all features, functions and capabilities . . . sufficient . . . for transmission, routing or other provision of a telecommunications service." Because packet switching and DSLAMs are used to provide telecommunications services, packet switching qualifies as a network element. We adopt a definition of packet switching that does not favor or disadvantage one packet switching technology over another. Our intention is to define packet switching in such a way as to capture the functionality of packet networks, without regard to a particular "packetizing" technology that an incumbent LEC has deployed in its network. Several parties propose definitions of packet switching which elaborate on the Commission's existing circuit switching definition. We decline to adopt proposed definitions of packet switching that exclude DSLAMs from the packet switching functionality. We further decline to adopt equipment-specific packet switching network elements, as proposed by Intermedia and e.spire. We find that with today's technology, packetizing is an integral function of the DSLAM. Accordingly, we include the DSLAM functionality, with the routing and addressing functions of packet switches, in our functional definition of packet switching. (i) Proprietary Concerns Associated With Packet Switching 318. No party alleged that packet switching was proprietary within the meaning of section 251(d)(2). We find that the record provides no basis for withholding packet switching from competitors based on proprietary considerations or subjecting packet switching to the more demanding "necessary" standard set forth in section 251(d)(2)(A). Instead we examine packet switching under the "impair" standard of section 251(d)(2)(B). (i) Unbundling Analysis for Packet Switching 319. We decline at this time to unbundle the packet switching functionality, except in limited circumstances. Among other potential factors, we recognize that the presence of multiple requesting carriers providing service with their own packet switches is probative of whether they are impaired without access to unbundled packet switching. The record demonstrates that competitors are actively deploying facilities used to provide advanced services to serve certain segments of the market namely, medium and large business and hence they cannot be said to be impaired in their ability to offer service, at least to these segments without access to the incumbent's facilities. In other segments of the market, namely, residential and small business, we conclude that competitors may be impaired in their ability to offer service without access to incumbent LEC facilities due, in part, to the cost and delay of obtaining collocation in every central office where the requesting carrier provides service using unbundled loops. We conclude, however, that given the nascent nature of the advanced services marketplace, we will not order unbundling of the packet switching functionality as a general matter. 320. Both the record in this proceeding, and our findings in the 706 Report, establish that advanced services providers are actively deploying facilities to offer advanced services such as xDSL across the country. Competitive LECs and cable companies appear to be leading the incumbent LECs in their deployment of advanced services. For example, in 1999, Rhythms expects to roll out xDSL services in 1,000 end offices nation wide. Covad's planned network deployment is expected to reach 51 MSAs by the end of 1999. In the past year, NorthPoint deployed facilities capable of transmitting xDSL signals in 17 metropolitan markets. NorthPoint plans to expand its DSL-based local networks from 25 major markets, representing 37 metropolitan statistical areas (MSAs), to 28 markets, or 61 MSAs, by the end of 1999. Qwest announced in August 1999, that it is now providing DSL service in 13 U.S. markets and plans to expand to more than 30 major markets by the end of 1999. In addition, EarthLink has partnered with Sprint to offer nationwide xDSL service. KMC Telecom Inc. announced aggressive rollout of DSL services with plans to introduce additional broadband applications by year-end. Marketplace developments like the ones described above suggest that requesting carriers have been able to secure the necessary inputs to provide advanced services to end users in accordance with their business plans. This evidence indicates that carriers are deploying advanced services to the business market initially as well as the residential and small business markets. 321. Several parties, in addition to the incumbent LECs, argue that the Commission should not unbundle packet switching or DSLAMS generally. We recognize that equipment needed to provide advanced services, such as DSLAMS and packet switches, are available on the open market at comparable prices to incumbents and requesting carriers alike. Incumbent LECs and their competitors are both in the early stages of packet switch deployment, and thus face relatively similar utilization rates of their packet switching capacity. Packet switching utilization rates will differ from circuit switching utilization rates because of the incumbent LEC's monopoly position as carrier of last resort. Incumbent LEC circuit switches, because they serve upwards of 90 percent of the circuit switched market, may achieve higher utilization rates than the circuit switches of requesting carriers. Because the incumbent LEC does not retain a monopoly position in the advanced services market, packet switch utilization rates are likely to be more equal as between requesting carriers and incumbent LECs. It therefore does not appear that incumbent LECs possess significant economies of scale in their packet switches compared to the requesting carriers. 322. Collocating in incumbent LEC central offices imposes material costs and delays on a requesting carrier and materially diminishes a requesting carrier's ability to provide the services it seeks to offer. As discussed above, we identified the costs and delays associated with collocation as factors that impair a requesting carrier's ability to self-provision circuit switches to serve residential and small business market. We see no reason to distinguish a requesting carrier's collocation-related costs and delays to provide circuit-switched service from those collocation costs and delays incurred by requesting carriers to provide packet-switched services. These costs and delays lead us to find that competitors are impaired in their ability to offer advanced services without access to incumbent LEC facilities. As discussed in more detail below, that conclusion is not dispositive of whether unbundling is appropriate at this time under section 251(d)(2). As discussed in section IV above, in addition to the "impair" standard we consider whether unbundling will open local markets to competition and how access to a given network element will encourage the rapid introduction of local competition to the benefit of the greatest number of customers. 323. NorthPoint argues that an additional impediment it faces when providing advanced services using xDSL technologies is the absence of line sharing. Currently, many incumbent LECs offer advanced services over the high-frequency range of the same loops they use to offer voice services. Although the incumbent LEC may use a single copper pair to provide xDSL services, in the absence of line sharing, requesting carriers providing xDSL services must purchase an additional unbundled loop to serve their customers, thereby incurring additional non-trivial costs. In light of the substantial number of packet switches deployed by competitive LECs, even in comparison to incumbent LEC deployment, we conclude that these non-trivial costs are substantial enough to impair the requesting carrier's ability to provide the services it seeks to offer within the meaning of section 251(d)(2). Unlike circuit switching services, however, requesting carriers providing data services do not face the operational impediment of obtaining a coordinated cutover of the loop on a timely basis, because they typically are providing service over a second line. Because such carriers purchase an additional unbundled copper loop to serve the customer, the customer's voice service is never disconnected, and the requesting carrier faces none of the timing and quality impediments associated with the "hot cut" process. 324. We further decline to unbundle specific packet switching technologies incumbent LECs may have deployed in their networks. E.spire/Intermedia request that we require incumbent LECs to unbundle: (1) the ports on their data switches or routers; and (2) the connectivity, including the switching fabric and associated software functions, between such ports at capacities ranging from DS0 to DS3. E.spire/Intermedia focus their request upon a particular packet-switching technology -- frame relay. E.spire/Intermedia argue that they are impaired without access to these data unbundled network elements to complete "virtual circuits" because they lack the incumbent LEC's economies of density and the ability to statistically multiplex data traffic to make efficient use of transport facilities. 325. We reject e.spire/Intermedia's request for a packet switching or frame relay unbundled network element. First, as discussed above, we will define unbundled network elements, to the extent practicable, in a technologically neutral manner so as to not favor one particular packet switching technology over another. Defining an unbundled network element according to a particular packet switching technology, such as frame relay, violates this principle of technological neutrality. Furthermore, defining packet switching elements according to a specific technology creates the possibility that as innovative packet switching technologies are deployed, they may or may not fall within our definition of packet switching. Second, e.spire/Intermedia have not provided any specific information to support a finding that requesting carriers are impaired without access to unbundled frame relay. We note, however, that e.spire/Intermedia are free to demonstrate to a state commission that lack of unbundled access to the incumbent's frame relay network element impairs their ability to provide the services they seeks to offer. A state commission is empowered to require incumbent LECs to unbundle specific network elements used to provide frame relay service, consistent with the principles set forth in this order. 326. We do find, however, one limited exception to our decision to decline to unbundle packet switching. Access to packetized services to provide xDSL service requires "clean" copper loops without bridge taps or other impediments. Furthermore, xDSL services generally may not be provisioned over fiber facilities. In locations where the incumbent has deployed digital loop carrier (DLC) systems, an uninterrupted copper loop is replaced with a fiber segment or shared copper in the distribution section of the loop. In this situation, and where no spare copper facilities are available, competitors are effectively precluded altogether from offering xDSL service if they do not have access to unbundled packet switching. Moreover, if there are spare copper facilities available, these facilities may not meet the necessary technical requirements for the provision of certain advanced services. For example, if the loop length exceeds 18,000 feet, the provision of ADSL service is technically infeasible. When an incumbent has deployed DLC systems, requesting carriers must install DSLAMs at the remote terminal instead of at the central office in order to provide advanced services. We agree that, if a requesting carrier is unable to install its DSLAM at the remote terminal or obtain spare copper loops necessary to offer the same level of quality for advanced services, the incumbent LEC can effectively deny competitors entry into the packet switching market. We find that in this limited situation, requesting carriers are impaired without access to unbundled packet switching. Accordingly, incumbent LECs must provide requesting carriers with access to unbundled packet switching in situations in which the incumbent has placed its DSLAM in a remote terminal. This obligation exists as of the effective date of the rules adopted in this Order. The incumbent will be relieved of this unbundling obligation only if it permits a requesting carrier to collocate its DSLAM in the incumbent's remote terminal, on the same terms and conditions that apply to its own DSLAM. Incumbents may not unreasonably limit the deployment of alternative technologies when requesting carriers seek to collocate their own DSLAMs in the remote terminal. 327. Policy Goals. Incumbent LECs argue in this proceeding that their incentive to invest and innovate in new technologies capable of providing advanced services will be curtailed if we mandate unbundling. We note that investments in facilities used to provide service to nascent markets are inherently more risky than investments in well established markets. Customer demand for advanced services is also more difficult to predict accurately than is the demand for well established services, such as traditional plain old telephone service (POTS). 328. We acknowledge that the incumbent LEC argument that unbundling may adversely affect innovation is consistent with economic theory, but events in the marketplace suggest that other factors may be driving incumbent LECs to invest in xDSL technologies, notwithstanding the economic theory. For example, in January 1998, U S WEST announced a rollout of ADSL service to 40 in-region metropolitan areas. In October 1998, BellSouth announced its plans to offer ADSL service to 1.7 million customers in 30 markets by the end of 1998, and 23 additional markets in 1999. In January 1998, SBC announced a "massive rollout" of ADSL, "targeting more than 500 central offices and 9.5 million residential and business customers by year-end." In January 1999, Bell Atlantic announced plans to rollout ADSL service in several states and entered into a marketing alliance with America On-Line in which Bell Atlantic hopes, by the end of 1999 to make ADSL available to seven million subscribers. Combined, Bell Atlantic and GTE have stated that the number of xDSL capable-lines available in region will be 17 million and they will have ADSL capability in 550 central offices, allowing them to serve as many as 6.1 million xDSL customers. Such investments have been planned and undertaken notwithstanding the fact that we sought comment in August 1998 on whether facilities used to provide advanced services must be unbundled pursuant to section 251. 329. Despite the encouraging signs of investment in facilities used to provide advanced services described above, we are mindful that regulatory action should not alter the successful deployment of advanced services that has occurred to date. Our decision to decline to unbundle packet switching therefore reflects our concern that we not stifle burgeoning competition in the advanced service market. We are mindful that, in such a dynamic and evolving market, regulatory restraint on our part may be the most prudent course of action in order to further the Act's goal of encouraging facilities-based investment and innovation. 330. Our overriding objective, consistent with the congressional directive in section 706, is to ensure that advanced services are deployed on a timely basis to all Americans so that consumers across America have the full benefits of the "Information Age." The advanced services marketplace is a nascent one. Although some investment has occurred to date, much more investment in the future is necessary in order to ensure that all Americans will have access to these services. We remain concerned about the lack of deployment in rural areas. We note that we will carefully monitor the deployment of broadband services to ensure that the objectives of section 706 and the Act are being met. We decline to unbundle packet switching at this time, except for the limited exception described above. A. Interoffice Transmission Facilities 1. Background 331. In the Local Competition First Report and Order, the Commission concluded that incumbent LECs must provide interoffice transmission facilities on an unbundled basis to requesting carriers. In particular, the Commission required incumbent LECs to provide dedicated and shared transport as an unbundled network element pursuant to section 251(c)(3). The Commission found that such access was technically feasible and would promote competition in the local exchange market. In that order, however, the Commission declined to address the unbundling of incumbent LEC dark fiber because the record provided insufficient evidence to decide that issue. 332. In the Notice, we sought comment on the application of the "necessary" and "impair" standards to previously identified unbundled network elements, including interoffice transport facilities. The Notice requested that parties include specific costs and an analysis of the availability of alternative sources of transport supply. We also sought comment on whether, in light of technological advances or experience in the marketplace since adoption of the Local Competition First Report and Order, we should modify the definition of any of the previously identified network elements including, for example, the definition of "transport," to include dark fiber. 333. Incumbent LECs generally argue that interoffice transport should not be unbundled where a single alternative source of transport is available. Competitive LECs argue that because alternative sources of transport supply are largely unavailable, requesting carriers are impaired without access to unbundled transport. Most of the state commissions addressing this issue agree that transport should remain an unbundled network element. 1. Discussion 334. We find that requesting carriers are impaired without access to unbundled dedicated and shared transport network element. In particular, self-provisioning ubiquitous interoffice transmission facilities, or acquiring these facilities from non-incumbent LEC sources, materially increases a requesting carrier's costs of entering a market or of expanding the scope of its service, delays broad-based entry, and materially limits the scope and quality of a requesting carrier's service offerings. Although the record indicates that competitive LECs have deployed transport facilities along certain point-to-point routes, the record also demonstrates that self-provisioned transport, or transport from non-incumbent LEC sources, is not sufficiently available as a practical, economic, and operational matter to warrant exclusion of interoffice transport from an incumbent LEC's unbundling obligations at this time. Accordingly, we conclude that incumbent LECs must offer unbundled access to their interoffice transmission facilities nationwide. a. Dedicated Transport (i) Definition 335. In the Local Competition First Report and Order, the Commission defined dedicated interoffice transmission facilities as "incumbent LEC transmission facilities dedicated to a particular customer or carrier that provide telecommunications between wire centers owned by incumbent LECs or requesting telecommunications carriers, or between switches owned by incumbent LECs or requesting telecommunications carriers." The Commission further concluded that incumbent LECs must provide all technically feasible capacity-related services such as DS1-DS3 and OC3-OC96 services. 336. High-Capacity Transmission. We reaffirm that the definition of dedicated transport set forth in the Local Competition First Report and Order includes all technically feasible capacity-related services such as DS1-DS3 and OC3-OC96 dedicated transport services. We clarify that this definition includes all technically feasible capacity-related services, including those provided by electronics that are necessary components of the functionality of capacity-related services and are used to originate and terminate telecommunications services. We find that unbundling high-capacity dedicated transport offerings will encourage competition and facilitate the deployment of advanced services. Unbundling high-capacity dedicated transport offerings also addresses claims by CompTel and other parties that non-incumbent LEC facilities cannot provision sufficient bandwidth for data-intensive services. Accordingly, we modify section 319(d)(ii) of our rules to clarify that incumbent LEC must unbundle DS1 through OC192 dedicated transport offerings and such higher capacities as evolve over time. Our intention is to ensure that the definition of interoffice transmission will apply to new, as well as current technologies, and to ensure that competitors will continue to be able to access these facilities as unbundled network elements as long as that access is required pursuant to section 251(d)(2). 337. Notwithstanding the fact that we require incumbents to unbundle high-capacity transmission facilities, we reject Sprint's proposal to require incumbent LECs to provide unbundled access to SONET rings. In the Local Competition First Report and Order, the Commission limited an incumbent LEC's transport unbundling obligation to existing facilities, and did not require incumbent LECs to construct facilities to meet a requesting carrier's requirements where the incumbent LEC has not deployed transport facilities for its own use. Although we conclude that an incumbent LEC's unbundling obligation extends throughout its ubiquitous transport network, including ring transport architectures, we do not require incumbent LECs to construct new transport facilities to meet specific competitive LEC point-to-point demand requirements for facilities that the incumbent LEC has not deployed for its own use. 338. Dark Fiber. In addition, we modify the definition of dedicated transport to include dark fiber. Dark fiber is deployed, unlit fiber optic cable that connects two points within the incumbent LEC's network. As discussed above, dark or "unlit" fiber, unlike "lit" fiber, does not have electronics on either end of the dark fiber segment to energize it to transmit a telecommunications service. Thus, dark fiber is fiber which has not been activated through connection to the electronics that "light" it and render it capable of carrying telecommunications services. To provide additional capacity, new electronics are attached to previously "lit" fiber or to previously "dark" fiber. Because dark fiber is already installed and easily called into service, we find that it is similar to the unused capacity of other network elements, such as switches or "dead count" or "vacant" copper wire that is dormant until carriers put it in service. 339. We agree with state commissions and competitive LECs that dark fiber meets the statutory definition of a network element, and therefore is included within the definition of the dedicated interoffice transport network element. Section 153(29) of the Act defines the term "network element" as a "facility or equipment used in the provision of a telecommunications service, including "features, functions, and capabilities that are provided by means of such facility or equipment." The Supreme Court upheld this broad definition of a network element and acknowledged that it includes not only physical elements but non-physical elements as well. Because dark fiber is unused transport capacity, we find that it is "a feature, function, and capability of facilities used to provide telecommunications services." In addition, we note that since the Commission released its Local Competition First Report and Order, several states, acting through arbitration proceedings, have required incumbent LECs to unbundle dark fiber interoffice transport facilities, and several federal district courts, in affirming state commission decisions, have held that dark fiber meets the statutory definition of an unbundled network element. 340. We reject incumbent LECs' arguments that because dark fiber is transport that is not currently "used" in the provision of a telecommunications service, within the meaning of section 153(29), it does not meet the statutory definition of a network element or the definition of interoffice transport. Rather, we agree with the Illinois Commission that the term "used in the provision of telecommunications service" in section 153(29) refers to network facilities or equipment that is "customarily employed for the purpose" of providing a telecommunications service. Although particular dark fiber facilities may not be "lit" they constitute network facilities dedicated for use in the provision of telecommunications service, as contemplated by the Act. Indeed, most other network elements have surplus capacity or can be upgraded to provide additional capacity and therefore are not always "currently used" as the term is interpreted by incumbent LECs. For example, switches, loops, and other network elements each may have spare, unused capacity, yet each meets the definition of a network element. 341. We acknowledge that it would be problematic if some facilities that the incumbent LEC customarily uses to provide service were deemed to constitute network elements (e.g., unused copper wire stored in a spool in a warehouse). Defining such facilities as network elements would read the "used in the provision" language of section 153(29) too broadly. Dark fiber, however, is distinguishable from this situation in that it is physically connected to the incumbent's network and is easily called into service. Thus, as indicated above, we conclude that dark fiber falls within the statutory definition of a network element. 342. We also note that our reading of the term "used" comports with the Commission's interpretation of the term "provide" in the context of section 271. Specifically, in the order denying Ameritech's application to provide long distance service pursuant to section 271 of the Act, the Commission rejected competitors' arguments that the term "provide" requires the BOC to "actually furnish" a checklist item. Rather, the Commission concluded that the term "provide" requires incumbent LECs to "make available" to requesting carriers the checklist item in question upon reasonable demand. Similarly, we interpret the term "used" in the definition of a network element to mean "capable of being used" in the provision of a telecommunications service. 343. We do not agree with GTE that, unlike vacant copper, dark fiber does not qualify as interoffice transport. According to GTE, dark fiber differs from extra copper pairs in a cable because dark fiber is "unused inventory," whereas copper cable is installed to provide maximum flexibility. We find this to be a distinction without a difference. Whether located in the loop plant or in the transport network of an incumbent LEC, both copper and fiber represent unused capacity. Accordingly, we conclude that dark fiber falls within the dedicated transport network element's "facilities, functions, and capabilities." (i) Proprietary Concerns Associated with Dedicated Transport 344. In the Local Competition First Report and Order, the Commission did not identify any proprietary concerns associated with dedicated transport. No party has identified any proprietary concerns associated with unbundled dedicated transport in this phase of the proceeding, and we find none. We therefore apply the "impair" standard of section 251(d)(2) to determine whether dedicated transport is subject to the unbundling obligations of the Act. (i) Unbundling Analysis 345. We conclude that lack of access to unbundled interoffice transport impairs a carrier's ability to provide the services it seeks to offer. Requiring carriers to self- provision, or acquire from third-party providers, extensive interoffice transmission facilities materially increases the costs of market entry or of expanding service, delays broad-based entry, and limits the scope and quality of the competitor's service offerings. Neither self-provisioning interoffice transport facilities nor obtaining these facilities from third-party sources is an adequate alternative to the ubiquitous transmission facilities that a competitor can obtain from the incumbent LEC under section 251's unbundling obligations. Accordingly, we require incumbent LECs to provide unbundled access to their interoffice transmission facilities. 346. Although the record indicates that competitive LECs have deployed interoffice transport facilities along selected point-to-point routes, primarily in dense market areas, we find that the these facilities are not available, as a practical, economic, and operational matter, such that a requesting carrier's ability to provide the services it seeks to offer would not be impaired without access to the incumbent's ubiquitous interoffice transmission facilities. Specifically, the competitive transport facilities that currently exist do not interconnect all of an incumbent LEC's central offices and all interexchange carrier's points of presence within an MSA, or a substantial portion thereof. 347. Availability of Alternatives Outside the Incumbent's Network. Local competitors began deploying fiber networks in urban markets approximately 15 years ago. Incumbent LECs have provided a significant amount of data indicating the location of transport facilities deployed by competitive LECs. For example, the incumbents submitted, through the USTA UNE Report, data that indicates that, by the end of 1998, competitive LECs had deployed interoffice transport in approximately 300 cities. According to the USTA UNE Report, competitors have deployed nearly 30,000 route miles of fiber within the top 50 MSAs. 348. In addition, the USTA UNE Report states that of the top 50 MSAs, forty-seven are served by at least three competitors; 29 are served by five or more competitors; and 16 are served by seven or more competitors. The USTA Report also asserts that requesting carriers have deployed fiber in all but 15 of the MSAs ranked between 50 and 150 and that competitors have centered their deployment of competitive fiber around "dense" wire centers, which USTA defines as wire centers with 40,000 or more access lines. The USTA UNE Report also maintains that as of March 1999, incumbent LECs have the following number of wire centers that are served by at least one competitive fiber provider: Ameritech 161; Bell Atlantic 274; BellSouth 136; GTE 70; SBC 284; US WEST 118. 349. The incumbents also provide evidence of the number of collocation arrangements in many of their wire centers. Relying on this data, the incumbents argue that there are significant alternatives to interoffice transport services available. According to USTA, the fact that competitors have operational collocation arrangements in approximately 874 dense wire centers implies the presence of competitive fiber "nearby." In particular, according to the USTA UNE Report, of the wire centers with 20,000 or more lines, 90 percent in the SBC region, 72 percent in the Bell Atlantic region, and 74 percent in the US West region have collocation, which the incumbents assert signifies competitive transport is available. 350. Bell Atlantic also argues that its Competitive Alternative Transport Terminal (CATT) service, currently offered on a trial basis with Metromedia Fiber Network Services (MFN), offers high capacity interoffice dedicated transport services to any collocated carrier. Bell Atlantic claims that MFN has entered into this CATT arrangement in a large number of end offices and that CATT will be generally available to other carriers pursuant to tariff. 351. Other evidence in the record, however, undermines the incumbents' suggestion that competitive fiber is sufficiently available that transport should not be unbundled. MCI WorldCom, for example, provides information about the number of transport providers in the six major cities included in the USTA survey. According to MCI WorldCom, only eight of the 138 wire centers in Los Angeles have three or more collocators that provide transport. Similarly, MCI WorldCom states that only four of 64 wire centers in Seattle have three or more collocators providing transport and only one of 25 wire centers in San Jose has three or more collocators providing transport. In addition, MCI WorldCom reports that, in Minneapolis, Richmond and Washington DC with 135, 51, and 158 wire centers respectively, no end office has three collocators providing transport. 352. In addition, NorthPoint reports that the incumbent LEC is the only source of transport for at least 70% of central offices in which NorthPoint is collocated, even in dense wire centers in large metropolitan areas. Similarly, Sprint asserts that in New York City, which is considered the most mature market in the country, Sprint continues to use the incumbent LEC extensively for transport because competitive fiber is not available in sufficient numbers of incumbent LEC central offices for it to offer a ubiquitous service in this area. 353. Ubiquity. We conclude that, despite the evidence of some competitively deployed interoffice transmission facilities, lack of access to the incumbent's dedicated transmission facilities impairs a requesting carrier's ability to provide the services it seeks to offer. The alternatives cited in the evidence submitted by the incumbents are not ubiquitously available, and therefore competitive transport if not available as a practical, economic and operational matter. 354. As an initial matter, we are not persuaded that the incumbents' data accurately reflects the extent to which alternatives are actually available to competitors. In particular, we find that only at a granular, wire center-by-wire center level does the record show the presence of competitive alternatives to the incumbent's interoffice transport, albeit on a non-ubiquitous basis. Thus, without access to unbundled dedicated transport, requesting carriers would be forced to create a patchwork of alternative network facilities, where they have been deployed and are being offered to other carriers, or alternatively to construct their own transport facilities. The USTA UNE Report based its analysis on the markets that have attracted the most competitive transport entry. For example, the USTA UNE Report states that "[I]n the Los Angeles MSA, 72 wire centers serve 40,000 + lines. Of these, 20 have at least one collocated competitive LEC. An analysis of fiber route maps shows that CLEC fiber passes through at least 15 of 20 wire center areas with collocation." Thus, according to USTA's data, 15 of 72 dense wire centers or approximately 21% of dense wire centers in the Los Angeles MSA include competitive fiber "nearby." 355. We note that the incumbents do not explain what is meant by fiber that is "nearby." Nor do incumbents explain how having fiber "nearby" reflects the availability of ubiquitous transport alternatives. In addition, however, because the incumbents' data focuses only on the most dense wire centers, the data provides little to no information about the availability of transport in less dense wire centers in the same cities. If the analysis were expanded to include less dense wire centers, or wire centers serving less than 40,000 lines, the analysis would presumably show a lower percentage of competitive alternatives for the entire MSA than is reflected by the data provided by the incumbents. 356. Incumbents rely on the evidence of competitively deployed transport submitted in the USTA UNE Report to argue that competitive LECs are not impaired without access to unbundled transport facilities in locations where competitive LECs have already deployed transport. Specifically, the incumbents argue that the Commission should exclude dedicated transport from an incumbent LEC's unbundling obligations in any area where at least one requesting carrier has deployed transport facilities and has collocated its own transmission equipment in an incumbent LEC central office. We reject this argument. Although the incumbents' evidence shows that nearly 30,000 route miles of fiber have been deployed in the top 50 MSAs, there are few, if any alternative transport facilities outside the incumbent LECs' networks that connect all or most of an incumbent LEC's central offices and interexchange carriers' points of presence within an MSA. Even where competitive alternatives exist, the alternatives generally do not travel the same routes as the incumbent's facilities. Thus, even if competitors were able to purchase indirect routing from alternative providers, to the extent alternatives exist, competitors more than likely have to route their traffic along indirect, inefficient routing patterns, thereby increasing their costs of transport. Thus, contrary to arguments made by incumbent LECs, we find that the evidence demonstrates that a significant number of central offices in a given MSA are not effectively served by competitive fiber facilities. 357. We reject any bright-line test that triggers elimination of an incumbent LEC's unbundling obligation based on the presence of a single competitor that has self- provisioned transport in a particular market. As discussed above, in order to determine whether or not a requesting carrier's ability to provide the services it seeks to offer is "impaired" within the meaning of section 251(d)(2), we must determine whether alternatives outside the incumbent's network are available as a practical, economic, and operational matter, and determine whether unbundling a particular element is consistent with the goals of the Act. 358. In particular, we find that basing our unbundling rules on the bright-line proposed by the incumbents does not address whether lack of unbundled access to the incumbent's ubiquitous transport facilities would impair other requesting carriers' ability to provide the services they seek to offer. Indeed, under the test proposed by the incumbents, the first new entrant to deploy transport facilities in any particular market would determine the degree and pace of competition in that market as well as the scope of an incumbent LEC's unbundling obligation, and would potentially result in the presence of only two competitors in the market (e.g. a duopoly). Limiting the development of competition in such a manner is contrary to the goals of the Act and is inconsistent with the purpose of our unbundling rules. 359. In order to provide service, competitive LECs require dedicated transport facilities that are more extensive than those that are currently deployed along the point-to- point routes. The competitive alternatives that are available along limited point-to- point routes do not necessarily allow competitive LECs to connect their collocation arrangements or switching nodes according to the needs of their individual network designs. These carriers also require dedicated transport to deliver traffic from their own traffic aggregation points to the incumbent LEC's network for purposes of interconnection. Without access to the incumbent's ubiquitous transport facilities, competitive LECs are faced with the delays and costs of deploying their own transport facilities to meet the demand. Alternatively, competitive LEC's must utilize a patchwork of competitive alternatives, where available, to collect and route traffic to the required destination. 360. Entrance Facilities. Bell Atlantic and BellSouth specifically argue that extensive deployment by competitive LECs of the transport link between the interexchange carrier point of presence and an incumbent's serving wire center (the "entrance facility"), requires us to find that requesting carriers are not impaired in their ability to serve these point-to-point markets. According to Bell Atlantic, for example, there are competitors that serve approximately 90 percent of Bell Atlantic's special access transport customers. According to BellSouth, 19 percent of their wire centers have at least one actual or pending collocation arrangement and one actual or pending entrance facility. 361. We acknowledge that, based on the record before us, the entrance facility market appears to be the most mature segment of the interoffice transport market, and thus may, in some situations, provide requesting carriers with effective alternatives to unbundled transport for certain point-to-point routes. The record does not indicate, however, the extent to which these facilities are available to other requesting carriers or whether the location of these facilities serve the transport needs of requesting carriers seeking to provide service to particular locations. In particular, the incumbents' data does not indicate the locations at which competitive entrance facilities terminate, or whether the facilities connect incumbent LEC serving wire centers to all or substantially all of the interexchange carrier points of presence. Accordingly, we cannot conclude, based on the record before us, that the competitive entrance facility market is providing requesting carriers with effective alternatives to unbundled transport for all, or substantially all of the routes requesting carriers would need in order to provide the services they seek to offer. 362. Dark Fiber. Incumbent LECs argue that some competitive LECs have deployed significant amounts of fiber to meet the growing demand for transport services, and that competitive LECs are not impaired without access to the incumbent's unbundled dark fiber. Incumbent LECs further argue that the presence of competitive fiber in dense wire centers is evidence of a wholesale market for dedicated transport, and support this claim by providing anecdotal evidence that competitors are swapping fiber capacity with each other. We disagree. Rather, we agree with those commenters that argue that a competitive wholesale market for alternative network elements has not developed for dedicated transport, in part because of the lack of ubiquitous transport alternatives. 363. Although there is evidence of transport deployment by non-incumbent providers along some point-to-point routes, the record does not support a general finding that requesting carriers can, on a ubiquitous basis, practically and effectively substitute transport services provided by other competitive carriers for unbundled transport. Indeed, the record indicates that the "fiber frenzy" and "bandwidth markets" cited by incumbent LECs are largely limited to portions of inter-city, long-haul networks that do not ubiquitously reach the interoffice segments of the incumbent LEC's network. Lack of access to ubiquitous transport alternatives, which allow competitive LECs to interconnect their networks with all the central offices serving their customers, will impair these carriers' ability to provide the services they seek to offer. Accordingly, we reject the incumbent LECs' argument that the presence of a competitive transport alternatives along certain routes is evidence that requesting carriers generally are not impaired without access to the incumbents' unbundled dark fiber. 364. In addition, to the extent that there may be excess capacity along these fiber routes, non-incumbent providers of competitive transport facilities are under no legal obligation to offer their excess capacity to their competitors. Moreover, interexchange carriers (IXCs) operate both as access customers of the incumbent LEC, as well as the incumbent's competitor in the local exchange market. These inter-carrier relationships complicate the functioning of an effective wholesale transport market because the alternative provider of transport is also a significant competitor. In these circumstances, it is possible that local affiliates of IXCs could potentially discriminate against unaffiliated requesting carriers seeking access to competitive transport facilities by denying access altogether. 365. Although we include dark fiber in the unbundling obligations of section 251(c)(3), we note that GTE argues that it must maintain control of its dark fiber reserves because, as a carrier of last resort, it is obliged to provide service to any and all customers as the need arises. GTE also argues that requiring incumbent LECs to make their reserve capacity available to new entrants discourages long term business planning and deprives the incumbents of the fruits of their investment. We note that with the addition of electronics such as Dense Wave Division Multiplexing (DWDM) equipment, incumbent and competitive carriers alike can expand the bandwidth of existing capacity without installing new dark fiber. We find that technological solutions such as these largely address GTE's concern that unbundled access to dark fiber may adversely affect its ability to provide service. In addition, however, if incumbent LECs are able to demonstrate to a state commission that unbundling dark fiber threatens their ability to provide service as a "carrier of last resort," states have the flexibility to establish reasonable limitations and technical parameters for dark fiber unbundling. We conclude, however, that for a limitation on dark fiber to be reasonable, it must relate to a likely and foreseeable threat to an incumbent LEC's ability to provide service as a carrier of last resort. In establishing reasonable limitations and technical parameters for dark fiber, states should acknowledge that requesting carriers require regulatory certainty in order to implement their business plans. 366. Other Technologies. We reject Bell Atlantic's proposal that the Commission consider the availability of wireless transport in our unbundling analysis. The record does not demonstrate that wireless transport options are available across any particular MSA. Nor does the record address the question of whether integrating wireless transport offerings into a wireline transport network allows providers to offer service of the same quality and functionality as they would be able to offer using wireline alternatives. Notably, NEXTLINK, the largest Local Multipoint Distribution Service (LMDS) licensee and a potential source of competitive wireless transport, supports the continued availability of unbundled dedicated transport network elements. 367. Tariffed Offerings. We also reject GTE and US West's argument that competitive LECs have access to ubiquitous transport through the use of the incumbents' special access tariff arrangements. As discussed above, we give little weight to the incumbent LEC's special access tariffs. Moreover, the Commission previously rejected this argument in the Local Competition First Report and Order. For reasons the Commission articulated in that order, we reject the incumbents' argument here. If we were to adopt the incumbents' approach, the incumbents could effectively avoid all of the 1996 Act's unbundling and pricing requirements by offering tariffed services that, according to the incumbents, would qualify as alternatives to unbundled network elements. This would effectively eliminate the unbundled network element option for requesting carriers, which would be inconsistent with Congress' intent to make available to requesting carriers three different competitive strategies, including access to unbundled network elements. 368. Cost. We conclude that the costs of self-provisioning dedicated transport facilities materially diminish a requesting carrier's ability to provide the services it seeks to offer. We agree with commenters that argue that replicating the incumbent's vast and ubiquitous transport network would be prohibitively expensive, and delay competitive entry. In the Local Competition First Report and Order, the Commission concluded that a requesting carrier would incur "much higher costs" if it "had to construct all of its own facilities" to match the scope of an incumbent LEC's interoffice transport network. Nothing has changed in the intervening three years to cause us to alter this conclusion. Indeed, based on the record before us, we conclude that the material costs and delays associated with self-provisioning duplicate, ubiquitous transport facilities would impair a competitive LEC's ability to offer services to a broad base of consumers. Accordingly, we require incumbent LECs to offer unbundled access to their dedicated transport facilities. 369. Self-provisioning dedicated transport requires competitive LECs to incur significant direct and other costs, including the cost of fiber, the cost of deploying fiber in public rights of way, trenching and the cost of purchasing and collocating the necessary transmission equipment. For example, the record indicates that the direct equipment costs of purchasing interoffice transport equipment exceeds $300 per line, and that the cost of constructing alternative transport facilities (e.g., digging and backfilling trench) are between $200,000 - $300,000 per mile in densely populated areas. According to GTE, the direct cost of constructing a one hundred mile dedicated transport facility is close to $3 million. 370. In addition, in order to use alternative transport facilities, either through self- provisioning or through third-party providers, a competitive LEC must collocate at the incumbent's central office. Collocating in each end office imposes materially greater costs on requesting carriers than would the purchase of the incumbent's interoffice transport facilities. Based on the record, it appears that the current range for non-recurring charges for provisioning physical collocation arrangements is between $15,000 and $508,000 for each central office where a competitor serves customers with unbundled loops. This results in an increase of between 15 and 20 percent to the costs of the equipment installed in the cage. In addition to the substantial costs of constructing and collocating self-provisioned transport facilities, competitive LECs must incur additional of negotiating and obtaining municipal rights-of-way permissions. 371. If a competitive LEC were required to obtain transport from multiple, non- ubiquitous alternative providers of transport, to the extent it is available, they would incur additional costs associated with coordinating back office billing and collection arrangements, as well as the costs associated with coordinating operational issues arising out of use of multiple vendors. While we acknowledge that the precise level of costs will vary according to the business plans of each competitive LEC, we conclude that contracting with third-parties to coordinate among multiple carriers in order to serve ubiquitously would materially diminish the ability of a requesting carrier to provide the services it seeks to offer. Moreover, because purchasing transport capacity is generally less expensive at higher levels of capacity, competitive LECs using multiple providers would lose efficiencies they would otherwise achieve if they were able to aggregate their traffic over the facilities of one ubiquitous provider. 372. We reject the incumbent LECs' cost models that purport to demonstrate that the fact that competitors have deployed a significant amount of fiber in downtown business districts is evidence that the cost of self-provisioning transport facilities does not impair a competitive LEC's ability to provide the service it seeks to offer. We find that cost models estimating the costs of self-provisioning transport are highly sensitive to assumptions that are not necessarily representative of the actual market place. For example, BellSouth provides a cost model that analyzes the transport networks of several competitive LECs located in Atlanta, and projects that the costs to the competitive LECs of extending the scope of their network to reach all central offices within that city is between $35 and $38 per DS1. BellSouth does not explain the difference between its model's cost estimate of $35-$38 per month, per dedicated DS1 and the cost estimate of $84 per month, per dedicated DS1 generated by a model the Commission developed in its universal service proceeding. Nor does BellSouth explain why the costs generated by its model are significantly lower than the costs generated by the model developed by Hatfield Associates, Inc., which shows the cost of a DS-1 to be $110 per month. Moreover, it is not clear whether BellSouth's cost estimates assume full utilization of the transport facilities. For competitive LECs entering the market that have little usage, the relevant comparison between the costs of self-provisioning and purchasing unbundled transport from the incumbent should be based on the number of DS1s actually carried, not on the number of DS1s that could potentially be used by the requesting carrier. 373. Ameritech proposes the use of a model that, it asserts, shows that in two second tier cities in Ameritech's territory, it is economical for competitive LECs to build ubiquitous transport networks of less than 100 miles to wire centers with a total of 100,000 access lines. Even assuming, arguendo, that Ameritech's model accurately projects the theoretical viability or profitability of extending a competitive LEC's transport network, as noted by the Supreme Court, the ability to "amass earnings" alone is not dispositive of whether or not a requesting carrier is impaired without access to the incumbent's unbundled transport. We therefore find that cost models proposed by BellSouth, Ameritech, and others do not accurately indicate the extent to which the costs associated with self-provisioning transport materially diminish a requesting carrier's ability to provide the services it seeks to offer. Finally, as discussed above, we do not base our unbundling analysis on individual business case analyses. 374. Timeliness. We conclude that lack of access to the incumbent's interoffice transport network would materially delay a requesting carrier entry into the local market or alternatively delay expansion of an existing carrier's service offerings. Whether requesting carriers self-provision interoffice transport, or purchase it from third- party providers, they must collocate their own equipment at the incumbent's central office. Thus, collocation is an essential prerequisite to self-provisioned and third- party provisioned transport, and the time required to collocate affects a requesting carrier's ability to provide service using dedicated transport. 375. Incumbents and requesting carriers provide different estimates about the time required to implement a single collocation arrangements in an incumbent LEC's central office. In general, competitive LECs argue that each collocation arrangement requires between six months and a year to provision. In addition, these carriers argue that the delay associated with implementing collocation arrangements is compounded as competitive LECs expand their networks and seek to establish more collocation arrangements. Incumbent LECs respond that they have provisioned collocation to requesting carriers in a timely fashion and on a broad scale. 376. We acknowledge that collocation arrangements necessarily require some time to implement, and that the amount of time required to order and provision a collocation arrangement will vary from incumbent LEC to incumbent LEC and by requesting carrier. Accordingly, we do not attempt to specifically quantify what constitutes a reasonable provisioning interval for a single collocation arrangement. We agree, however, with commenters that provisioning the multiple collocation arrangements needed to provide a ubiquitous transport network within an MSA would compound significantly the inherent delays associated with provisioning a single collocation arrangement. NorthPoint contends that most incumbent LECs have imposed "governors" on the number of collocation applications they will accept. Specifically, BellSouth has limited the number of collocation applications a requesting carrier may file to five per month, thereby delaying ubiquitous rollout of services. Requiring requesting carriers to collocate in numerous end offices in order to obtain ubiquitous transport facilities would materially delay the ability of requesting carriers to enter a market or to expand its service offerings to the greatest number of consumers. 377. Several carriers argue that the process of securing necessary access to rights-of-way, pole attachments, and conduit space significantly delays their ability to compete. For example, NEXTLINK notes that it took two years to negotiate and obtain a telecommunications franchise from the City of New York before it could deploy competitive facilities, and that it must negotiate separate agreements with each municipality traversed by its fiber ring. We find that the delays of this magnitude associated with obtaining authority to access public rights-of-way materially delay the ability of a requesting carrier to self-provision transport. 378. Functionality and Quality. We conclude that requiring carriers to utilize alternative sources of transport imposes functional and quality disadvantages that materially diminish a requesting carrier's opportunity to provide the services it seeks to offer. If the Commission were to adopt the incumbent LEC proposals to eliminate unbundled access to interoffice transport in areas where there are one or more alternative suppliers in the market, carriers would have to use multiple alternative suppliers, where available, for their transport requirements. Using a patchwork of transport offerings consisting of facilities acquired from competitive LEC/competitive access providers and the incumbent LEC, in lieu of ubiquitous incumbent LEC transport facilities, would introduce additional complexity into a ubiquitous end-to-end transport network. For example, Sprint notes that when facilities of more than one carrier are involved, repair times are roughly three times longer than if the entire transport network were controlled by one carrier or provisioned exclusively through unbundled transport. In addition, Sprint argues that an end-to-end transport offering provisioned by multiple providers may require several digital-to-analog and analog-to-digital conversions or protocol conversions, which could lower total connection speeds otherwise achievable with a single provider transport offering. Although we do not conclude that digital-to-analog or analog-to-digital protocol conversions result in a material quality degradation, we find that, as a general matter, requiring requesting carriers to utilize a patchwork of competitive alternatives, to the extent they are available, to collect and route traffic to the required destination can result in a material degradation of quality in the service the requesting carrier to seeks to provide. 379. Goals of the Act. We recognize that requiring incumbent LECs to unbundle dedicated transport may be marginally overinclusive because of the presence of some alternative fiber along selected point-to-point routes in dense markets. We believe, however, that the benefits of uniform transport unbundling outweigh the costs of creating a patchwork regime in which incumbent LECs would likely seek to litigate its transport unbundling obligation on particular point-to-point routes where transport alternatives are arguably available. As we stated above, unbundling requirements that provide uniformity and certainty to the market will allow new entrants and fledgling competitors to implement national and regional business plans and attract capital investment. Litigation over the incumbents' unbundling obligations requires the parties to these agreements and the state commissions that approve them to expend vast amounts of time and resources and would impede the development of competition. 380. Creating a patchwork of transport unbundling obligations would be inconsistent with the goal of the 1996 Act to facilitate rapid entry into the local exchange market. We reiterate the Commission's conclusion in the Local Competition First Report and Order that "[w]e recognize that there are alternative suppliers of interoffice facilities in certain areas. We are convinced, however, that entry will be facilitated if competitors have greater, not fewer, options for procuring interoffice facilities as part of their local networks, and that Congress intended for competitors to have these options available from competitors." Furthermore, we believe that our decision to unbundle interoffice transport is consistent with Congress' recognition, in section 271, that providing unbundled access to interoffice transport would encourage rapid entry into the local exchange market. 381. We further find that the allegations of the competitive harms resulting from a uniform transport unbundling obligation are overstated. We believe that there are significant operational and technical incentives for a requesting carrier to eliminate its reliance upon transport provided by incumbent LECs over the long term. Where alternative providers build transport facilities to areas exclusively served by the incumbent LEC's facilities, requesting carriers may substitute those alternative sources of transport as they become available. We therefore expect the need for unbundled transport will decrease as competitive transport networks become more ubiquitous. We will closely monitor the developments in the transport market to determine whether the transport market, or a particular segment of this market, is supplying requesting carriers with effective alternatives to the incumbent LEC's unbundled network elements when we reexamine these rules in three years. a. Shared Transport 382. We find that lack of unbundled access to incumbent's shared transport would impair the requesting carrier's ability to use unbundled switching. In particular, without access to unbundled shared transport, a requesting carrier would have to self- provision or purchase dedicated transport from the incumbent, which would materially increase the costs and decrease the quality of services the requesting carrier could provide, and would materially limit the carrier's ability to serve a broad base of customers. Accordingly, where an incumbent LEC provides requesting carriers with access to unbundled switching, we require incumbent LECs also to provide access to unbundled shared transport services. (i) Definition 383. In the Local Competition Third Reconsideration Order, the Commission defined shared transport as transmission facilities shared by more than one carrier, including the incumbent LEC, between end office switches, between end office switches and tandem switches, and between tandem switches in the incumbent LEC's network. The Commission clarified in that proceeding that incumbent LECs are not required to provide shared transport between incumbent LEC switches and serving wire centers. No commenter in this phase of the proceeding specifically addressed the definition of shared transport and the record provides no basis for modifying our definition of shared transport. 384. Ameritech, however, argues that shared transport is not an "unbundled" network element within the meaning of section 251(c)(3). Specifically, Ameritech argues that under the Supreme Court's ruling, incumbent LECs must provide to requesting carriers pre-assembled combinations of individual unbundled network elements if the element can be purchased separately. Because shared transport is technically inseparable from unbundled switching requesting carriers do not have the option of using unbundled shared transport without also taking unbundled local switching. Thus, according to Ameritech, the shared transport element is not an "unbundled" element within the meaning of section 251(c)(3). 385. We reject Ameritech's arguments. The Supreme Court upheld the Commission's interpretation that the phrase "on an unbundled basis" in section 251(c) does not refer to physically separated elements but rather to separately priced elements. Shared transport is an "unbundled" element because it consists of separately priced switching and transport network elements. The fact it is technically infeasible for a competitor to use shared transport with self-provisioned switching is irrelevant to whether an element is "unbundled" pursuant to section 251(c)(3). In addition, the Eighth Circuit, in affirming our decision in the Local Competition Third Reconsideration Order, rejected Ameritech's argument when it held that shared transport meets the definition of an unbundled network element because it is a "feature, function, [or] capability," that is provided by facilities and equipment used in the provision of a telecommunications service. Accordingly, we conclude that shared transport meets the definition of an unbundled network element. (i) Proprietary Concerns Associated With Shared Transport 386. Ameritech asserts that its routing table used to provide shared transport is proprietary. As discussed above, we reject Ameritech's claim because we find that incumbent LECs may not withhold access to unbundled local switching on the grounds that switch routing tables are proprietary in nature under section 251(d)(2)(A). With the exception of Ameritech, no commenter identifies any proprietary concerns associated with the provision of shared transport, and we identify none. Accordingly, we analyze shared transport under the "impair" standard of section 251(d)(2)(B). (i) Unbundling Analysis 387. We conclude that a requesting carrier's ability to provide the services it seeks to offer is impaired without access to the incumbent's unbundled shared transport. Without access to unbundled shared transport, a requesting carrier would have to self-provision or purchase dedicated transport from the incumbent, which would materially increase the costs and decrease the quality of services the requesting carrier could provide, and would materially limit the carrier's ability to serve a broad base of customers. Accordingly, we conclude that incumbent LECs must provide unbundled access to shared transport. 388. Costs and Quality. We find that lack of unbundled access to the incumbent's shared transport facilities materially increases a requesting carrier's costs of providing service. As described above, we find that there is a lack of ubiquitous transport alternatives available to requesting carriers. Thus, without access to the incumbent's shared transport facilities, a requesting carriers must either deploy its own dedicated facilities or purchase dedicated transport from the incumbent. Because requesting carriers, in the early stages of entering the local market, may not yet have sufficient market information to forecast accurately their traffic volumes, they may miscalculate the amount of dedicated transport capacity they will need. Specifically, an inability to reasonably forecast traffic volumes would likely cause a requesting carrier to purchase an insufficient amount, or conversely, too much dedicated transport capacity. In shared transport arrangements, the switch routes the competitor's traffic through the most efficient trunking group available. The trunking group is shared among many users, including the incumbent LEC's end users, thereby reducing requesting carrier costs and utilizing capacity only when necessary to route and complete a call. 389. In addition, as traffic demands increase, a requesting carrier will incur a non- recurring charge each time it purchases additional transport capacity. In contrast, where a requesting carrier purchases unbundled shared transport to meet increased customer demand, it effectively purchases the entire capacity of the incumbent LEC's network and will not incur non-recurring charges for additional increments of dedicated transport capacity. Purchasing only those increments of capacity that the requesting carrier requires to meet demand eliminates inefficient use of dedicated transport facilities. In addition, at low volumes requesting carriers will incur significantly higher recurring, per-minute costs to substitute dedicated transport for shared transport arrangements at low volumes. We reiterate the Commission's conclusion in the Third Order on Reconsideration that "the relative costs of dedicated transport, including the associated NRCs [non-recurring charges], is an unnecessary barrier to entry for competing carriers." 390. According to Ameritech, competitive LECs have the option of using its end office integration (EOI) service, a tariffed, retail service that Ameritech claims will carry, on a minute-of-use basis, whatever interoffice transport traffic the competitive LEC delivers to its point of interconnection. Under this plan, Ameritech would not require requesting carriers to order dedicated transport facilities until their actual volume levels justified provisioning a dedicated trunk. Consistent with the little weight we afford the incumbents' tariffed offerings for consideration as an alternative to dedicated transport, we reject the argument that Ameritech's tariffed EOI service eliminates the obligation to unbundle shared transport. 391. We agree with commenters that argue that the ability to obtain access to shared transport enables them to handle traffic at peak loads and maintain call blockage levels that are at parity with those of the incumbent LECs. As the Commission stated in the Local Competition Third Reconsideration Order, a new entrant entering the local market with smaller traffic volumes would have to maintain greater excess transport capacity relative to the incumbent LEC in order to provide the same level of service quality (i.e. same level of successful call completion) as the incumbent LEC. We conclude a requesting carrier would be impaired without access to unbundled shared transport because it would have to choose between purchasing excess capacity or incurring increased call blockage rates. 392. Goals of the Act. We find that requiring incumbent LECs to provide unbundled access to shared transport is consistent with the Act's goal of encouraging requesting carriers to rapidly enter the local market and serve the greatest number of customers. Requiring unbundled access to shared transport is particularly important because it addresses the transport needs of requesting carriers in the early stages of competitive entry by allowing competitors to efficiently purchase transport facilities as they ramp up toward higher-capacity dedicated transport requirements. Furthermore, when used in conjunction with unbundled switching, requesting carriers may find it economical to serve the small business and residential markets using shared transport because these market segments may not always support traffic volumes that justify using dedicated transport services. Accordingly, we find that requiring unbundled access to shared transport promotes the prompt development of competition to serve the greatest number of customers, as intended by the Act. A. Signaling Networks and Call-Related Databases 1. Signaling Networks a. Background 393. In the Local Competition First Report and Order, the Commission concluded that incumbent LECs, upon request, must provide nondiscriminatory access to their signaling networks on an unbundled basis. The Commission stated that it was technically feasible for incumbent LECs to provide such access, and that such access was critical to entry in the local exchange market. The Commission concluded that incumbent LECs must provide unbundled access to signaling networks as part of the unbundled switch network element as well as on a standalone basis. 394. In the Notice, we sought comment on the application of the "necessary" and "impair" standards to previously identified unbundled network elements, including signaling networks. The Notice also requested that parties include specific costs and an analysis of the availability of alternative signaling facilities. 395. The majority of state commissions and competitive LECs commenting in this phase of the proceeding argue that the incumbent LECs' signaling networks should be unbundled because alternatives to the incumbents' signaling networks are more costly, have lower quality, and do not provide the ubiquity of the incumbents' networks. The incumbent LECs argue that based on the availability of alternative signaling providers, requesting carriers are not impaired in their ability to provide services. a. Discussion 396. We conclude that without unbundled access to the incumbent LECs' signaling networks, a requesting carrier's ability to provide the services it seeks to offer is impaired. Requiring a requesting carrier to obtain signaling from alternative sources would materially diminish its ability to provide the services it seeks to offer, due to the quality differences between the signaling networks available from the incumbent LEC and those available from alternative providers of signaling. As described below, we conclude that neither self-provisioning signaling networks, nor obtaining this element from third-party sources, is a sufficient substitute that would justify excluding signaling networks from the incumbent LECs' unbundling obligation under section 251(c)(3). We therefore require incumbent LECs to provide requesting carriers with unbundled access to their signaling networks. (i) Definition 397. In the Local Competition First Report and Order, the Commission defined the signaling network element as including, but not limited to, signaling links and signaling transfer points (STPs). No party commenting in this phase of the proceeding has asked us to modify our definition, and we find no marketplace developments that would cause us to re-evaluate our definition of the signaling network element. Accordingly, we reaffirm the definition of signaling networks that was adopted in the Local Competition First Report and Order. (i) Proprietary Analysis 398. We agree with commenters that signaling links and STPs are not proprietary. Moreover, we do not discern any copyright, patent, or trade secret implications to unbundling signaling links and STPs, and carriers do not generally rely upon their signaling links and STPs to differentiate themselves from their competitors. In addition, SS7 signaling networks generally adhere to Bellcore standards rather than LEC-specific protocols, and provide seamless connectivity between networks. We therefore conclude that signaling links and STPs are not proprietary elements, and we analyze signaling networks under the "impair" standard of section 251(d)(2)(B). (i) Unbundling Analysis 399. Current switch technology requires each local switch to connect to a single STP. All parties, including incumbent LECs, agree that because the incumbent LECs' switching networks are already connected to a STP, a carrier that purchases unbundled switching from an incumbent LEC must also purchase signaling from that incumbent LEC. In such cases, the incumbent LEC must provide access to its signaling network from that switch in the same manner in which it obtains such access itself. 400. A requesting carrier that has deployed its own switch, or has purchased switching from an alternative source, however, may purchase signaling from an incumbent LEC and link its switch to the incumbent LEC's signaling network. Alternatively, the requesting carrier may self-provision signaling or purchase signaling from an alternative provider. Thus, the only issue left to be resolved is whether competitive LECs that do not purchase switching from an incumbent LEC are impaired without unbundled access to the incumbent's signaling network element. 401. We conclude that regardless of whether a requesting carrier self-provisions its own switching, or purchases switching from an alternative source, the incumbent LEC must provide the requesting carrier with unbundled access to the incumbent's signaling network, pursuant to section 251(c)(3). Consistent with our framework for unbundling as set forth above, we find that in such situations, lack of access to unbundled signaling systems materially diminishes the ability of a requesting carrier to provide the services it seeks to offer. In particular, requiring a competitor to self- provision or use alternative sources of signaling materially degrades the quality of its service to end users and materially restricts its ability to provide service on a ubiquitous basis. We therefore require incumbent LECs to provide requesting carriers that have deployed their own switching facilities access to the incumbent LEC's unbundled signaling network for each of the requesting carrier's switches. This connection shall be made in the same manner as an incumbent LEC connects one of its own switches to a signaling transfer point. 402. Alternatives in the Market. The record indicates that there are several alternatives to the incumbent LECs' signaling networks available in the market. In particular, there are six major facilities-based SS7 network providers (AT&T, MCI WorldCom, Illuminet, TNX, GTE-INS, and SBC/SNET) and four mid-sized facilities-based SS7 network providers (GST, ICG, Intermedia and US LEC), that operate regional SS7 networks. In GTE's service area, twelve competitive LECs have opted to build their own signaling networks. In addition, there are several suppliers of the equipment used to operate a signaling network: Lucent, Tekelec, Nortel, Alcatel, IEX Corporation, SummaFour, and Siemens. We also note that the equipment produced by these companies is based on standard interfaces and protocols. 403. As discussed above, however, the mere existence of alternatives outside the network does not mean that requesting carriers are not impaired without unbundled access to the incumbent LEC's network. Based on our analysis of the factors identified above, we find that a requesting carrier is materially diminished in its ability to offer service if it is not able to purchase signaling as an unbundled network element. 404. Cost. In light of the significant evidence of multiple third-party providers of signaling, we disagree with parties that assert that self-provisioning signaling, or obtaining signaling from alternative providers, would involve substantial and material cost and would delay competition in the local market. Although several states and competitive LECs argue that replication of the incumbent LEC's signaling network would be "extremely costly," they have not submitted cost data in the record to support their claims. 405. Unlike self-provisioning a switch or network elements that are dedicated to individual subscribers (i.e. the NID), deploying a signaling network does not require a requesting carrier to incur substantial sunk and fixed costs, because a carrier does not need multiple signaling facilities in order to establish a signaling network that is capable of providing service to a broad base of customers. Rather, existing technology permits the carrier that is using its own switch to route signals over a single set of A-links (links between a specific end office and the SS7 network for any given point code). The carrier's single STP can serve its entire network. Alternatively, a competitive LEC can purchase signaling from the non-incumbent sources mentioned above. We agree with MGC that cost-effective SS7 signaling networks are generally available on a national basis. 406. Several parties argue that because alternative signaling providers have not established the ubiquitous presence to match the incumbent LECs' signaling footprint, the cost of transport to an alternative provider's signaling network materially increases the requesting carrier's cost. Replicating the ubiquitous signaling networks of the incumbent LECs may be prohibitively expensive for some competitive LECs. In addition, new entrants in the local market most likely do not have the scale necessary to justify the investment needed to replicate the incumbent LECs' signaling networks. We do not find sufficient evidence in the record, however, to support a conclusion that the cost of self-provisioning or purchasing signaling from alternative sources, in and of itself, would require us to unbundle the incumbent's signaling network. Accordingly, we find that the cost of non- incumbent LEC alternative signaling networks is not dispositive of whether or not a competitive LEC's ability to provide the services it seeks to offer is materially diminished. For reasons we discuss below, however, we do find that competitive LECs need to have access to a ubiquitous signaling network in order to ensure the same quality of service as the incumbents. 407. Ubiquity and Quality. Although we do not conclude that the cost of self- provisioning alternative signaling impairs a requesting carrier's ability to provide the services it seeks to offer, we find that lack of access to the incumbent LECs' ubiquitous networks materially diminishes this ability. We agree with commenters who argue that because alternative vendors of signaling networks only have a few geographically dispersed STPs, they cannot provide requesting carriers with signaling that is of comparable quality to that of the incumbent LECs' signaling networks. 408. The ubiquitous nature of an incumbent LEC's signaling network provides it with advantages that competitive LECs cannot achieve through use of alternative signaling networks. For example, the Bell Operating Companies have deployed at least one STP in every LATA. Each of the incumbent LEC's STPs is connected to one or more incumbent switches serving customers limited to a particular geographic area within the incumbent LEC's region, while alternative signaling systems typically rely on a very few or even a single STP pair as a gateway to its signaling system. Consequently, if a competitive LEC uses an alternative provider of signaling, the competitive LEC's entire customer base may be connected to a single STP pair. If an outage occurs within the incumbent LEC's signaling network, only those customers served by switches connected to that particular STP will be adversely affected. In contrast, where a competitive LEC relies on one or a small number of STPs to serve its entire network, a greater portion, if not all, of the competitive LECs' customers will be negatively affected by a network outage. The lack of access to a ubiquitous signaling network could adversely impact the competitive LEC's customer satisfaction, thereby placing the competitive LEC at a disadvantage vis-…-vis the incumbent. 409. We note that Time Warner claims that alternative signaling networks lack diversity in signaling links that provide redundant signaling paths, and this lack of diversity reduces the reliability of the signaling network. Time Warner argues that this lack of diversity also leads to more frequent outages that directly impact customer satisfaction. Although we agree that lack of diversity in signaling networks could very likely result in greater numbers of customers affected by network outages, Time Warner has not provided evidence in this proceeding to support its specific claim. 410. Other commenters identify similar quality issues associated with the use of alternative signaling networks. We agree that a lack of redundant signaling paths may increase the likelihood that more customers may be affected by signaling outages. We are, however, unable to conclude based on the record before us, if the outages attributed to a lack of diversity are isolated incidents, or if they are the result of an increased risk of failure. Thus, we do not base our decision to unbundle the incumbent LECs signaling networks on the lack of diversity in alternative providers' signaling networks. As discussed above, however, we find that a competitive LEC's ability to provide the service it seeks to offer is materially diminished, because alternative providers' signaling networks lack the ubiquity of the incumbent LECs' networks, and that larger portions of a requesting carrier's network would likely be affected by a single point of failure on the signaling network. 411. Other quality problems identified by ALTS' members include poor customer service associated with utilizing alternative signaling providers. While we agree that these quality concerns may materially decrease the ability of requesting carriers to provide the services it seeks to offer, we do not find them dispositive of whether requesting carriers are impaired in general. In particular, it is not clear from the record whether these quality concerns are isolated instances, or alternatively, are prevalent throughout the industry. 412. Goals of the 1996 Act. We conclude that unbundling the incumbent LECs' signaling networks will promote the development of facilities-based competition and thereby encourage investment and innovation in new technologies and telecommunications services. Competitive LECs deploying their own switches will not have the incentive to do so if they are faced with having to rely on less ubiquitous and less reliable alternatives for signaling. Unbundling the incumbent LECs' signaling networks will give competitive LECs incentive to deploy their own switches, because they can be connected to the ubiquitous incumbent LECs' signaling networks. The 1996 Act was designed to spur competition in the local market. Our decision to unbundle incumbent LECs' signaling networks facilitates this goal, and creates options for consumers in their local telecommunications service. 1. Call-Related Databases a. Background 413. In the Local Competition Order, the Commission determined that access to call- related databases was technically feasible, and concluded that incumbent LECs must provide nondiscriminatory access to their call-related databases on an unbundled basis, for the purpose of switch query and database response through the SS7 network. The Commission also required incumbent LECs to provide unbundled access to the Advanced Intelligence Network (AIN) platform. The Commission found that such access was technically feasible, and that competitors would be impaired without access to the AIN platform. 414. In the Notice, we asked parties to comment on unbundling the seven network elements we previously identified, including signaling and call-related databases. Most requesting carriers argue that the Commission should require incumbent LECs to provide access to call-related databases on an unbundled basis. Incumbent LECs argue that access to call-related databases on an unbundled basis is not required under section 251(d)(2). a. Discussion 415. We find that, as a general matter, requesting carriers' ability to provide the services they seek to offer is impaired without unbundled access to the incumbent LECs' call-related databases. Thus, we require incumbent LECs, upon request, to provide nondiscriminatory access to their call-related databases on an unbundled basis, for the purpose of switch query and database response through the SS7 network. We conclude that requesting carriers' ability to provide the services they seek to offer is impaired without unbundled access to the incumbent LECs' AIN platform and architecture. Thus, we find that incumbent LECs, upon request, must provide nondiscriminatory access to their AIN platform and architecture. We also conclude, however, that service software created in the AIN platform and architecture is proprietary and thus analyzed under the "necessary" standard of section 251(d)(2)(A). Based on our "necessary' standard, we conclude that incumbent LECs are not required to unbundle the services created in the AIN platform and architecture that qualify for proprietary treatment. (i) Definition 416. In the Local Competition First Report and Order, the Commission defined call- related databases as "databases, other than operations support systems, that are used in signaling networks for billing and collection or the transmission, routing, or other provision of telecommunications service." The Commission further required incumbent LECs to provide unbundled access to their call-related databases, including but not limited to: the Line Information database (LIDB), the Toll Free Calling database, the Local Number Portability database, and Advanced Intelligent Network databases. No commenter in this phase of the proceeding challenges the definitions of call-related databases or AIN that were adopted in the Local Competition First Report and Order, and we find no reason for modifying those definitions. As discussed below, however, we clarify that the definition of call- related databases includes, but is not limited to, the calling name (CNAM) database, as well as the 911 and E911 databases. 417. The Advanced Intelligent Network (AIN) uses distributed intelligence in centralized databases to control call processing and manage network information, eliminating the need for those functions to be performed at every switch. The AIN database enables some call processing functions to be performed outside the switch. There are two separate components of the AIN. The first component is the AIN platform and architecture. The AIN platform and architecture basically consists of an off-line computer known as the Service Creation Environment (SCE), Service Management System (SMS), and AIN software. AIN services are designed and tested in the SCE. Once a service is successfully tested, the software is transferred to a SMS that administers and supports service control point (SCP) databases in the network. The SMS then regularly downloads software and information to a SCP where interaction with the voice network takes place via signaling links and STPs. 418. When a software "trigger" is activated, an AIN capable switch uses the SS7 network to access databases, SCPs, that contain service software and subscriber information, for instruction on how to route, monitor, or terminate the call. The second component of the AIN is the AIN service software that is developed in the AIN platform, and is used to provide telecommunications service. Examples of AIN services include: deployment of number portability, wireless roaming, and advanced services such as same-number service (i.e. 500 number service) and voice recognition dialing. 419. As a general matter, no commenter challenges the definitions of call-related databases or AIN that were adopted in the Local Competition First Report and Order. Several commenters, however, request that the Calling Name (CNAM) database be classified as a call-related database. The CNAM database contains the name of the customer associated with a particular telephone number and is used to provide Caller ID and related services. We take this opportunity to clarify that the definition of call-related databases includes, but is not limited to, the calling name (CNAM) database, as well as the 911 and E911 databases. Call-related databases are databases that supply information or instructions used for "billing and collection or used in the transmission, routing, or other provision of telecommunications service." The CNAM, 911, and E911 databases are call- related databases, because they are used for "billing and collection, or used in the transmission, routing or other provision of a telecommunications service." CNAM databases are used to provide Caller ID and related telecommunications services, and the 911 and E911 databases are telecommunications services used to provide emergency assistance. We specifically identify the CNAM, 911 and E911 databases as being illustrative of call-related databases, and not as a comprehensive list of all call-related databases. 420. We note that Low Tech Designs requests that the Commission require AIN triggers and AIN trigger upgrades be made available to competitors on an unbundled basis. We find that there is not enough evidence in the record to make a determination about the technical feasibility of unbundling AIN triggers. We therefore decline to expand our definition of call-related databases to include AIN triggers, and reaffirm the definition of call-related databases in the Local Competition First Report and Order. Low Tech Designs also requests that the Commission mandate the interconnection of "CLEC-provided and other third-party AIN/SS7 Service Control Points and Intelligent Peripherals." We decline this request because we find that there is not enough evidence in the record to make a determination as to the technical feasibility of interconnecting third-party SCPs and Intelligent Peripherals to incumbent LECs' signaling networks. Our refusal to grant Low Tech Design's request in this proceeding does not affect the ability of any state commission to address this issue. (i) Proprietary Concerns Associated with Call-Related Databases 421. With the exception of AIN service software, commenters do not identify proprietary concerns associated with the provision of call-related databases. Moreover, with the exception of AIN service software, we do not discern any copyright, patent, or trade secret implications to unbundling call-related databases. Thus, with the exception of AIN service software, we analyze call-related databases under the "impair" standard of section 251(d)(2)(B). 422. Because certain services created in the AIN platform and architecture are proprietary, we agree with Ameritech and BellSouth that if competitive LECs receive unbundled access to incumbent LECs' AIN platforms, access to AIN service software should not be unbundled. Ameritech cites a new proprietary service, "Privacy Manager," to illustrate why its AIN service software qualifies as a proprietary network element. Privacy Manager is derived from the SCE, and allows consumers to screen telemarketing calls. Ameritech asserts that Privacy Manager "includes several new and useful improvements" that are subject to patent protection, and are the subject of several pending patent applications. Ameritech adds that Privacy Manager is currently a trade secret because it has independent economic value, is not generally known by or readily discernable to Ameritech's competitors, and has been the subject of reasonable security measures. We agree with Ameritech that services such as Privacy Manager qualify as "proprietary" treatment. We also agree that software services such as Privacy Manager are new and innovative products used to differentiate the incumbent LECs' service offering. As such, they should be evaluated under the "necessary" standard of section 251(d)(2)(A). (i) Unbundling Analysis for Call-Related Databases (a) The "Impair" Standard 423. Consistent with our framework for unbundling set forth above, we find that lack of access to call-related databases on an unbundled basis would materially impair the ability of a requesting carrier to provide the services it seeks to offer in the local telecommunications market. In particular, we are persuaded that there are no alternatives of comparable quality and ubiquity available to requesting carriers, as a practical, economic, and operational matter, for the incumbent LECs' call-related databases. Thus, we require incumbent LECs to provide non discriminatory access to their call-related databases, including, but not limited to, the CNAM Database, the 911 Database, the LIDB, Toll Free Calling Database, AIN databases, and downstream number portability databases, by means of physical access at the signaling transfer point linked to the unbundled databases. Incumbent LECs must allow requesting carriers that have purchased an incumbent LEC's local switching capability to use the incumbent LEC's service control point element in the same manner, and via the same signaling links, as the incumbent LEC itself. An incumbent LEC must allow a requesting carrier that has deployed its own switch and has linked that switch to an incumbent LEC's signaling system to gain access to the incumbent LEC's service control point in a manner that allows the requesting carrier to provide any call-related database-supported services to customers served by the requesting carrier's switch. 424. We note that our analysis of call-related databases is intertwined with our analysis of signaling, because signaling is necessary to obtain access to certain call-related databases. Thus, our decision to unbundle the signaling network leads us to unbundle call-related databases as well. We believe that access to call-related databases, such as the LIDB, Toll Free calling, CNAM and Number Portability databases, encourages efficient network architecture deployment and promotes the ability of new entrants and established competitors to provide service in the local exchange market. We also agree with commenters that access to the incumbent LECs' call-related databases is critical to permitting the seamless routing and completion of traffic both among competitors and between competitors and the incumbent LEC. 425. With respect to AIN specifically, the Commission found in the Local Competition First Report and Order that requesting carriers need equivalent access to the incumbent LECs' SMSs to populate their own information in call-related databases. The Commission explained that information bound for many call-related databases is entered into an SMS that then downloads the information to the databases for real-time use on the network. To ensure efficient access to the incumbent LECs' databases, we affirm that incumbent LECs must provide a requesting telecommunications carrier with the information necessary to enter correctly, or format for entry, the information relevant for input into the incumbent LECs' SMS. The Commission also found in the Local Competition First Report and Order, that it is technically feasible to access the SMS, through the SCE, to deploy AIN services. There is no evidence in this record to suggest otherwise, and we therefore affirm the finding in the Local Competition First Report and Order that incumbent LECs must provide a requesting carrier the same access to design, create, test, and deploy AIN-based services at the SMS, through a SCE, that the incumbent LEC provides to itself. Incumbent LECs must also provide requesting carriers with access to call-related databases and access to the SMS in a manner that complies with the privacy requirements in section 222 of the Act. 426. As we further found in the Local Competition First Report and Order, access to the incumbent LECs' SCPs, SMS and SCE for the creation and deployment of AIN services may require incumbent LECs and requesting carriers to develop measures to protect the incumbent LECs' facilities and data. We continue to believe that there may be mediation issues that need to be addressed before a competing carrier obtains access to these databases. Accordingly, if parties are unable to agree to appropriate mediation mechanisms through negotiations, we conclude that during arbitration of such issues, state commissions (or the Commission acting pursuant to section 252(e)(5)) must consider whether such mediation is necessary, and if so, whether it will protect adequately against intentional or unintentional misuse of the incumbent LEC's AIN facilities. 427. SBC argues that requesting carriers have access to alternative call-related databases to store their data in any LIDB in the nation. Similarly, BellSouth and GTE claim that requesting carriers can obtain call-related database capabilities from alternative sources. Despite these assertions, we find that as with signaling networks, requesting carriers are impaired without unbundled access to incumbent LECs call- related databases pursuant to section 251(d)(2). 428. Cost and Quality. Several commenters argue that it would be costly for requesting carriers to replicate the incumbent LECs' call-related databases, or obtain call- related database services from alternative vendors. MediaOne submitted data to support these claims for the LIDB and CNAM databases. Other commenters, however, generally have not submitted sufficient cost data in the record to support their claims that it would be extremely costly to replicate the incumbent LECs' call- related databases, or obtain call-related database services from alternative providers. Based on the record before us, we find that the cost incurred by a requesting carrier to self-provision or use alternative databases does not appear to materially diminish the carrier's ability to provide the services it seeks to offer. 429. We conclude that unbundled access to incumbent LECs' call-related databases is required in some instances for requesting carriers to offer the telecommunications services they seek to provide. For example, in some cases, access to incumbent LEC databases is the only practical way to ensure proper call flow. Specifically, incumbent LECs are the only providers of CNAM database information. Incumbent LECs' CNAM databases provide information about customers of both requesting carriers and incumbent LECs. Therefore, in order for a switch-based competitor to provide caller ID to its customers, it must have access to the incumbent LEC's CNAM database. Such access is critical, especially because a majority of calls to a competitor's customers originate from the incumbent. 430. Goals of the Act. Requiring incumbent LECs to provide access to call-related databases, including access to the AIN databases, will foster investment and innovation in the local telecommunications marketplace. Requesting carriers require access to call-related databases and AIN databases to provide the services they seek to offer in the local telecommunications market. Requesting carriers also require access to the AIN platform and architecture, so that they may have the opportunity to devise innovative AIN services that will spur competition and benefit consumers through greater choices of telecommunications services. (a) The "Necessary" Standard 431. As discussed above, we find that AIN service software qualifies as a proprietary network element, and therefore, should be analyzed under the "necessary" standard. Our interpretation of the "necessary" standard requires the Commission to determine whether, after taking into consideration alternatives outside the incumbent's network, lack of access to that element would, as a practical, economic, and operational matter, preclude the requesting carrier from providing the services it seeks to offer. 432. We agree with Ameritech that unbundling AIN service software such as "Privacy Manager" is not "necessary" within the meaning of the standard in section 251(d)(2)(A). In particular, a requesting carrier does not need to use an incumbent LEC's AIN service software to design, test, and implement a similar service of its own. Because we are unbundling the incumbent LECs' AIN databases, SCE, SMS, and STPs, requesting carriers that provision their own switches or purchase unbundled switching from the incumbent will be able to use these databases to create their own AIN software solutions to provide services similar to Ameritech's "Privacy Manager." They therefore would not be precluded from providing service without access to it. Thus, we agree with Ameritech and BellSouth that AIN service software should not be unbundled. 433. We believe that excluding AIN service software, such as "Privacy Manager," from the unbundling requirements of section 251(d)(2), will protect incentives for the incumbent LEC to invest and deploy new and innovate services. We also believe that such protection, in conjunction with our decision to unbundle the AIN platform and architecture, will promote innovation and deployment of new services by requesting carriers. A. Operations Support Systems 1. Background 434. In the Local Competition First Report and Order, the Commission concluded that incumbent LECs must provide access to operations support systems (OSS) functions on an unbundled basis to requesting carriers. The Commission also required incumbent LECs to make modifications to their OSS as necessary in order to offer nondiscriminatory access to these functions, including access to interface design systems. Specifically, the Commission determined in the Local Competition First Report and Order that the provision of access to OSS functions and the information they contain is integral to the ability of competing carriers to enter the local exchange market. The Commission further concluded that a requesting carrier that lacks access to the incumbent's OSS "will be severely disadvantaged, if not precluded altogether, from fairly competing." In Iowa Utils. Bd., the Supreme Court expressly affirmed the Eighth Circuit's holding that the Commission's designation of operations support systems as a network element was an "eminently reasonable" interpretation of the 1996 Act. 435. In the Notice, we sought comment on the application of the "necessary" and "impair" standards to previously identified unbundled network elements, including OSS. The Notice also requested comment on whether the Commission should modify the definition of OSS. 436. All commenters to this proceeding agree that OSS qualifies as an unbundled network element. Incumbent LECs, however, argue that the Commission should limit access to OSS functions to those instances when a requesting carrier purchases another network element, an interconnection offering, or resold services from the incumbent LEC. Competitors, on the other hand, argue that OSS qualifies as an independent unbundled network element and therefore is not subject to any such limitations. Additionally, they argue that an incumbent LEC's duty to provide unbundled access to OSS includes an obligation to provide loop qualification information. ALTS requests that the Commission require incumbent LECs to provide access to OSS functions for carrier-to-carrier transactions. 1. Discussion 437. We find that requesting carriers are impaired without access to the incumbent LEC's OSS as an unbundled network element. The record demonstrates that, in general, lack of access to OSS as an unbundled network element materially diminishes a requesting carrier's ability to provide the services it seeks to offer. We also clarify that the definition of OSS includes access to loop qualification information. Finally, we reject the incumbent LECs' proposal to limit access to OSS to situations where the requesting carrier is ordering other unbundled network elements or resold services. a. Definition of OSS 438. In the Local Competition First Report and Order, the Commission defined OSS as consisting of pre-ordering, ordering, provisioning, maintenance and repair, and billing functions supported by an incumbent LEC's databases and information. OSS includes the manual, computerized, and automated systems, together with associated business processes and the up-to-date data maintained in those systems. Because of the varied, and largely non-standardized, development of incumbent LECs' OSS, the Commission identified certain functions needed by competitive carriers to deliver local exchange and exchange access services at the level expected by customers and state commissions. Specifically, the Commission identified the five functions of OSS that incumbent LECs must make available to competitors on an unbundled basis: pre-ordering, ordering, provisioning, repair and maintenance, and billing. 439. We find no reason to modify our definition of OSS. The majority of commenters support the existing definition of OSS. A few parties request that we broaden the definition of OSS to include access to the incumbent LEC's electronic interface and gateways to enable the processing of orders without manual intervention. Because these requests focus on the method by which competitors access incumbent LEC OSS, we believe that interface and gateway issues are already captured in the nondiscriminatory access requirements of the Local Competition First Report and Order. Accordingly, we find it unnecessary to modify our definition of OSS in this manner. We agree with ALTS, however, that the Commission should clarify that the pre-ordering function includes access to loop qualification information. Loop qualification information identifies the physical attributes of the loop plant (such as loop length, the presence of analog load coils and bridge taps, and the presence and type of Digital Loop Carrier) that enable carriers to determine whether the loop is capable of supporting xDSL and other advanced technologies. This information is needed by carriers seeking to provide advanced services over those loops through the use of packet switches and DSLAMs. 440. We clarify that pursuant to our existing rules, an incumbent LEC must provide the requesting carrier with nondiscriminatory access to the same detailed information about the loop that is available to the incumbent, so that the requesting carrier can make an independent judgment about whether the loop is capable of supporting the advanced services equipment the requesting carrier intends to install. Based on these existing obligations, we conclude that, at a minimum, incumbent LECs must provide requesting carriers the same underlying information that the incumbent LEC has in any of its own databases or other internal records. For example, the incumbent LEC must provide to requesting carriers the following: (1) the composition of the loop material, including, but not limited to, fiber optics, copper; (2) the existence, location and type of any electronic or other equipment on the loop, including but not limited to, digital loop carrier or other remote concentration devices, feeder/distribution interfaces, bridge taps, load coils, pair-gain devices, disturbers in the same or adjacent binder groups; (3) the loop length, including the length and location of each type of transmission media; (4) the wire gauge(s) of the loop; and (5) the electrical parameters of the loop, which may determine the suitability of the loop for various technologies. Consistent with our nondiscriminatory access obligations, the incumbent LEC must provide loop qualification information based, for example, on an individual address or zip code of the end users in a particular wire center, NXX code, or on any other basis that the incumbent provides such information to itself. 441. In addition, we agree with Covad that an incumbent LEC should not be permitted to deny a requesting carrier access to loop qualification information for particular customers simply because the incumbent is not providing xDSL or other services from a particular end office. We also agree with commenters that an incumbent must provide access to the underlying loop information and may not filter or digest such information to provide only that information that is useful in the provision of a particular type of xDSL that the incumbent chooses to offer. For example, SBC provides ADSL service to its customers, which has a general limitation of use for loops less than 18,000 feet. In order to determine whether a particular loop is less than 18,000 feet, SBC has developed a database used by its retail representatives that indicates only whether the loop falls into a "green, yellow, or red" category. Under our nondiscrimination requirement, an incumbent LEC can not limit access to loop qualification information to such a "green, yellow, or red" indicator. Instead, the incumbent LEC must provide access to the underlying loop qualification information contained in its engineering records, plant records, and other back office systems so that requesting carriers can make their own judgments about whether those loops are suitable for the services the requesting carriers seek to offer. Otherwise, incumbent LECs would be able to discriminate against other xDSL technologies in favor of their own xDSL technology. 442. We disagree, however, with Covad's unqualified request that the Commission require incumbent LECs to catalogue, inventory, and make available to competitors loop qualification information through automated OSS even when it has no such information available to itself. If an incumbent LEC has not compiled such information for itself, we do not require the incumbent to conduct a plant inventory and construct a database on behalf of requesting carriers. We find, however, that an incumbent LEC that has manual access to this sort of information for itself, or any affiliate, must also provide access to it to a requesting competitor on a non- discriminatory basis. In addition, we expect that incumbent LECs will be updating their electronic database for their own xDSL deployment and, to the extent their employees have access to the information in an electronic format, that same format should be made available to new entrants via an electronic interface. 443. We also clarify that under our existing rules, the relevant inquiry is not whether the retail arm of the incumbent has access to the underlying loop qualification information, but rather whether such information exists anywhere within the incumbent's back office and can be accessed by any of the incumbent LEC's personnel. Denying competitors access to such information, where the incumbent (or an affiliate, if one exists) is able to obtain the relevant information for itself, will impede the efficient deployment of advanced services. To permit an incumbent LEC to preclude requesting carriers from obtaining information about the underlying capabilities of the loop plant in the same manner as the incumbent LEC's personnel would be contrary to the goals of the Act to promote innovation and deployment of new technologies by multiple parties. 444. Consistent with the framework we adopted in the Local Competition First Report and Order, we conclude that access to loop qualification information must be provided to competitors within the same time intervals it is provided to the incumbent LEC's retail operations. To the extent such information is not normally provided to the incumbent LEC's retail personnel, but can be obtained by contacting incumbent back office personnel, it must be provided to requesting carriers within the same time frame that any incumbent personnel are able to obtain such information. It would be unreasonable, for instance, if the requesting carrier had to wait several days to receive such information from the incumbent in the incumbent's personnel have the ability to obtain such information in several hours. In order to provide local exchange and exchange access service, a competitor needs such information quickly to be able to determine whether a particular loop will support xDSL service. a. Proprietary Concerns Associated with OSS 445. The record does not indicate, nor do commenters argue, that OSS is proprietary. Moreover, we do not discern any copyright, patent, or trademark or trade secrecy implications to unbundling OSS. We therefore conclude that OSS should be evaluated under the "impair" standard. a. Unbundling Analysis for OSS 446. We conclude that lack of access to the incumbent LEC's OSS impairs the ability of requesting carriers to provide the services they seek to offer. The incumbents' OSS provides access to key information that is unavailable outside the incumbents' networks and is critical to the ability of other carriers to provide local exchange and exchange access service. We therefore require incumbent LECs to offer unbundled access to their OSS nationwide. 447. Commenters overwhelmingly agree that the unbundling of OSS satisfies the "impair" standard of section 251(d)(2). OSS is a precondition to accessing other unbundled network elements and resold services because competitors must utilize the incumbent LEC's OSS to order all network elements and resold services. Thus, the success of local competition depends on the availability of access to the incumbent LEC's OSS. Without unbundled access to the incumbent LEC's OSS, competitors would not be able to provide their customers comparable, competitive service, and hence would have to operate at a material disadvantage. While we acknowledge that a competitive market is developing for OSS systems, these alternative providers do not provide substitutable alternatives to the incumbent LEC's OSS functionality. Alternative OSS vendors provide requesting carriers with an electronic interface that allow competitive LECs to access the incumbent LEC's OSS and internal customer care systems. These vendors cannot provide a sufficient substitute for the incumbent LEC's underlying OSS because incumbent LECs have access to exclusive information and functionalities needed to provide service (e.g., customer service record information, provisioning of orders for unbundled network element and resold services, ability to initiate repairs for unbundled network elements and resold services, etc.). 448. We reject the incumbent LECs' arguments to limit the scope of a requesting carrier's access to the incumbent's OSS functions to situations where the competitor is ordering other unbundled network elements or resold services from the incumbent LEC. We find such limitations to be discriminatory because access to the same information and support functions as the incumbent LEC is needed by requesting carriers to provide quality service over their own facilities. For example, the incumbent LEC has access to unique information about the customer's service, and a competitor's ability to provide service is materially diminished without access to that information. This is true regardless of whether the competitor is providing service entirely through its own facilities, entirely over the incumbent's, or using some combination of the two. In particular, the full facilities-based competitor runs the risk of offering a lower quality service from the perspective of the end user if it does not know all the details of the customer's current service offering. As another example, carriers may also need to access the repair and maintenance function in the incumbent's OSS to submit trouble tickets for interconnection trunks. 449. We do not decide ALTS' request at this time that incumbent LECs provide access to OSS functions even when the incumbent is no longer the retail provider of local service to an end user. The record has not been sufficiently developed to establish how, absent access to incumbent LEC OSS, requesting LECs ability to provide the services they seek to offer would be materially diminished when the incumbent LEC is not involved in providing service to a retail customer. The most apparent example of this situation would be customer changeovers where competitive LECs are serving customers through resale of the incumbent's services or use of the incumbent's unbundled network elements. This appears to us to be an industry- wide issue. Thus, as a first step, we encourage the industry to develop guidelines and standards to facilitate the orderly transition of customers from one carrier to another. We note that any solution to this problem must adhere to the requirements of the Act, including the nondiscriminatory access requirements of section 251(c)(3) and the CPNI obligations of section 222. 450. We reject commenters' proposal that the Commission establish and ensure that incumbent LECs meet OSS performance standards, both quantitative and qualitative, to demonstrate parity under the rules. Failure to satisfy these performance standards, according to MCI WorldCom, should automatically trigger a process to identify and correct the root cause of the OSS problem. We decline to adopt performance standards in this proceeding. The issue before us in this proceeding is whether OSS is subject to the unbundling obligations of section 251, not whether the Commission should establish performance standards and penalties to determine if an incumbent is providing nondiscriminatory access to its OSS functions. We note that the states have primary authority under section 252 for setting schedules and resolving disputes concerning access to OSS functions as unbundled network elements. In addition, in the Second Order on Reconsideration , the Commission did not preclude requesting carriers from bringing enforcement actions against incumbent LECs to the Commission for consideration. Thus, more appropriate forums exist for the resolution of specific allegations of noncompliance with our unbundling rules. Accordingly, we find it unnecessary, at this time, to modify our rules in the manner suggested. A. Operator Services and Directory Assistance 1. Background 451. In the Local Competition First Report and Order, the Commission required incumbent LECs to provide unbundled access to their operator services and directory assistance (OS/DA). The Commission found that access to the systems supporting both operator call completion services and directory assistance was necessary, under section 251(d)(2)(A) for new entrants to provide competing local exchange service. The Commission also concluded that a competitor's ability to provide service would be significantly impaired, under section 251(d)(2)(B), if it did not have access to the incumbent LECs' operator call completion services and directory assistance. The Commission therefore required incumbent LECs to provide unbundled access to the databases used in the provision of both call completion services and directory assistance. In Iowa Utils. Bd., the Supreme Court expressly affirmed the Eighth Circuit's holding that the Commission's designation of operator services and directory assistance as a network element was an "eminently reasonable" interpretation of the 1996 Act. 452. In the Notice, we sought comment on whether the definition of OS/DA should be modified and whether there are any proprietary concerns associated with OS/DA systems. We also sought comment on whether OS/DA should remain an unbundled network element. We sought further comment on the implications of an incumbent LEC's obligations to provide OS/DA services under the nondiscriminatory access provisions of section 251(b)(3) if those services are not provided by incumbent LECs as unbundled network elements under section 251(c)(3). 453. Commenters generally support the existing definition of operator services and directory assistance and do not identify proprietary concerns associated with OS/DA systems. The majority of commenters, including competitive LECs, interexchange carriers, alternative OS/DA providers, and most state commissions, argue that incumbents should provide unbundled access to their OS/DA services. The incumbent LECs, MGC (a facilities-based competitive LEC), and the Ohio PUC note the general availability of third-party OS/DA alternatives as evidence of a wholesale market and argue that the Commission should not unbundle the incumbent LECs' OS/DA services. 1. Discussion 454. We find that where incumbent LECs provide customized routing, lack of access to the incumbents' OS/DA service on an unbundled basis does not materially diminish a requesting carrier's ability to offer telecommunications service. The record provides significant evidence of a wholesale market in the provision of OS/DA services and opportunities for self-provisioning OS/DA services. Moreover, we do not find that the evidence regarding the differences in cost, timeliness, quality, interoperability and ubiquity between the incumbent's OS/DA service and alternative OS/DA services, provided either through self-provisioning or third-party alternatives, is sufficient to conclude that lack of unbundled access to the incumbent's OS/DA service would materially diminish a requesting carrier's ability to offer the services it seeks to provide. We note that nondiscriminatory access to the incumbent's underlying databases used in the provision of OS/DA is required under section 251(b)(3) of the 1996 Act. The additional nondiscrimination requirements of section 251(b)(3), coupled with evidence of multiple alternative providers of OS/DA service in the marketplace, provide strong evidence that competitors are not impaired without access to the incumbent's OS/DA service as an unbundled network element. 455. Accordingly, incumbent LECs need not provide access to its OS/DA as an unbundled network element. All LECs, however, must continue to provide their competitors with nondiscriminatory access to their OS/DA, pursuant to section 251(b), as implemented by the Commission. We believe that this outcome best comports with the realities of a growing OS/DA marketplace, embraces a deregulatory approach where justified, and does not unduly confine the entry strategies of competitive carriers. a. Definition of Operator Services and Directory Assistance 456. The Commission has defined operator services as "any automatic or live assistance to a consumer to arrange for billing or completion, or both, of a telephone call," and has stated that directory assistance is a service that allows "subscribers to retrieve telephone numbers of other subscribers." In the Local Competition Second Report and Order, the Commission clarified that the nondiscriminatory requirements of section 251(b)(3) included the obligation of LECs to comply with the reasonable request of a competing provider to rebrand or unbrand its OS/DA services. We recently reaffirmed this holding in the Directory Listing Information Order, where we stated that to the extent technically feasible, a LEC must identify and rebrand the traffic it provides to its competitors. 457. We decline to expand the definition of OS/DA, as proposed by some commenters, to include an affirmative obligation to rebrand OS/DA and to provide directory assistance listing updates in daily electronic batch files. We find such modifications unnecessary because, as mentioned above, these obligations already exist under section 251(b)(3), and the relevant rules promulgated thereunder. a. Proprietary Concerns Associated with OS/DA 458. With the exception of one commenter, no parties identify proprietary concerns associated with OS/DA, and we find none. Moreover, we do not discern any copyright, patent, or trademark or trade secrecy implications associated with OS/DA. Accordingly, we analyze incumbent LECs' obligations to provide unbundled access to its OS/DA under the "impair" standard. a. Unbundling Analysis 459. Consistent with the unbundling analysis set forth above, we conclude that where an incumbent LEC provides customized routing to the requesting carrier as part of the unbundled switching element, lack of access to the incumbent's OS/DA on an unbundled basis does not materially diminish a requesting carrier's ability to provide the services it seeks to offer. The record demonstrates that a variety of alternative providers of OS/DA offer services at comparable cost and quality to those of the incumbents. We agree with the incumbent LECs, MGC, and the Ohio PUC that the incumbents enjoy no material advantage obtaining the key inputs for OS/DA services. Certain commenters point to differences in cost and the amount of time required to implement services provided by these alternative sources to support their arguments that competing carriers are impaired without access to the incumbents' OS/DA services. The majority of these commenters, however, focus on the differences in the quality and accessibility of the information in the incumbent LECs' OS/DA databases relative to that available from third-party sources. As discussed more fully below, we find that these quality differences are addressed adequately by other sections of the Act. 460. Alternatives in the Marketplace. Competition in the provision of operator services and directory assistance has existed since divestiture. Such competition has accelerated in the directory assistance market as a result of the Supreme Court's decision to allow copying of carriers' white pages listings in their entirety. For example, according to SBC, more than 30 competitive LECs presently provide their own OS/DA services or resell the services of non-incumbent LECs. In Bell Atlantic's region, only 70 out of 400 interconnection agreements require Bell Atlantic to provide OS/DA as an unbundled network element. Thus, in more than 80% of Bell Atlantic's interconnection arrangements, competitive LECs have chosen to provide OS/DA for themselves or to obtain such service from wholesale providers. According to the Rural Telephone Coalition, rural incumbent LECs have obtained OS/DA services from outside sources for many years because they find third-party sources to be cost-effective. In addition, Bell Atlantic reports that its wireless affiliate, Bell Atlantic Mobile, relies on a third-party OS/DA provider. MGC advocates that OS/DA not be unbundled because, in its view, competitive LECs can purchase OS/DA from a number of vendors offering cost-effective nationwide alternatives to those of the incumbent LECs. 461. Even requesting carriers advocating the unbundling of operator and directory assistance services acknowledge that there exists a substantial number of alternative providers of operator and directory assistance services. For example, AT&T, MCI WorldCom, and Sprint have already established national operator services via toll- free numbers. McLeod USA self-provisions nationwide directory assistance service. Metro One provides OS/DA services to ALLTEL and GST Telecom. Cox and Omnipoint obtain OS/DA service from Teltrust, and WinStar obtains these services from Frontier. Requesting carriers may also obtain OS/DA services and directory listings from numerous wholesale providers, including CenturyTel Telecommunications, Clifton Forge, Consolidated Communications, Excell, Experian's TEC Group, Frontier, HebCom, InfoNXX, Metro One, Quest411 and Teltrust. 462. It appears that this increasing availability of competitive OS/DA providers coincides with a decrease in incumbent LEC OS/DA call volumes. Evidence in the record indicates that call volumes to incumbent OS/DA services have declined steadily over the past few years. For example, SBC claims directory assistance call volumes have decreased almost 30 percent since 1995, and SBC operator-assisted calls have dropped by over 50 percent during the same period. Similarly, BellSouth's operator-assisted call volumes have declined over 60 percent in the past eight years. According to Bell Atlantic, it lost greater than 67 percent of its wholesale directory assistance calls between 1994 and 1998. In fact, Bell Atlantic claims that interexchange carriers accounted for over 68% of the operator services market in 1998 and represented 72% of the wholesale operator services market by 1997. This trend, combined with the number of alternative operator services and directory assistance providers outside the incumbent LECs' networks, strongly suggests that requesting carriers are not impaired without access to the incumbent LECs' OS/DA service. Significantly, we find that the existence of multiple alternative providers of OS/DA service in the marketplace, coupled with evidence of competitors' decreasing reliance on incumbent OS/DA services, demonstrates that requesting carriers' ability to provide the services it seeks to offer is not materially diminished without access to the incumbent's OS/DA service on an unbundled basis. 463. Cost. In light of the significant evidence of multiple third-party providers of OS/DA, we find unpersuasive assertions that replication of OS/DA service facilities and functionalities would involve substantial and material cost and would delay competitive entry into the local market. The costs associated with self- provisioning OS/DA include: (1) the cost of the facility, including employees, real estate, computers; (2) the cost of transporting traffic to the facilities; and (3) the cost of obtaining the underlying subscriber information contained in OS/DA databases. We acknowledge that, in some situations, depending on the type of OS/DA service a requesting carrier seeks to provide, OS/DA service may be more expensive if it is purchased from third-party providers than it would be if purchased from the incumbent. We find, however, that such differences will not materially diminish a requesting carrier's ability to provide local exchange or exchange access service. 464. We are unpersuaded by Cox's argument that OS/DA service should be unbundled because incumbents enjoy economies of scale and scope that greatly reduce the cost of providing these services to their own customers. In light of the number of alternative providers currently providing OS/DA service and the competitive market that is developing for long distance transport, we find this argument unconvincing. We also find that incumbents do not have any particular advantage in obtaining the facilities needed to create a call center, including employees, real estate and computers. In addition, unlike many other network elements, such as switching or transport, the ability to provide a nationwide OS/DA service does not require large amounts of sunk and fixed costs in facilities that must be deployed ubiquitously in order to serve a broad customer base. Rather, a requesting carrier can establish one call center or a few regional centers to which it can transport all of the calls on its network and provide OS/DA service nationwide. Moreover, we believe that a competitive LEC or a group of competitive LECs can achieve economies of scale by aggregating demand for OS/DA services over various regions by processing them through a single call center. Unlike the self-provisioning of switches, or other such network elements, self-provisioning a single OS/DA platform would not require the competitive carrier to deploy equipment throughout the network to ubiquitously serve its customers. 465. Certain competitive LECs assert that purchasing long-haul DS1 facilities to alternative OS/DA call centers is more expensive than purchasing local loops to access OS/DA services provided by incumbent LECs. In particular, Time Warner claims that special access rates to trunk its OS/DA calls to a vendor's national call center are approximately $500,000 a year. MediaOne estimates that remote long- haul facilities cost $1500-$2000 per month for a DS1 compared to local loops provisioned by the incumbent LEC for about $500 per month. 466. While, on its face, the disparity between transport costs to carry OS/DA traffic between the competitor's switch and a self-provisioned call center appears significant, it does not persuade us that transport costs associated with self- provisioning or purchasing OS/DA from third-party vendors materially diminishes the ability of requesting carriers to provide local exchange service. The record reveals a number of alternative OS/DA providers with multiple call centers located throughout the country. For example, HebCom operates five regional call centers, Excell operates six regional call centers and InfoNXX operates four. Teltrust operates a national OS/DA service with several call centers. The availability of multiple locations of alternative providers, both regional and national, allows competitors to choose a service that will be most cost-efficient, depending on the area in which it provides service. It is not clear from the record whether Time Warner considered the availability of these regional solutions to its OS/DA needs when making its initial decision to transport calls. Additionally, the incumbent LEC itself often maintains regional call centers that are outside the local calling area of a particular call center. Bell Atlantic, for example, offers directory assistance for most of New England (Rhode Island, Vermont, New Hampshire, Maine and western Massachusetts) out of its Providence, RI, Burlington, VT and Portland, ME offices, with all calls routed through a switch in Manchester, NH. In such cases, the incumbent may also incur long-haul transport costs to trunk its OS/DA traffic to the call center. 467. Regardless of the OS/DA provider, the cost of transporting traffic to the call center is factored into the overall price of OS/DA services. Where a competitive LEC obtains OS/DA services from an incumbent LEC, even at cost-based rates, the incumbent charges the competitive LEC for transport, either separately or as part of the total cost for OS/DA service. Similarly, where a competitive LEC obtains OS/DA from an alternative OS/DA provider, the carrier or OS/DA provider must pay for transport to the call center. It is notable that rural incumbent LECs, which arguably have to haul traffic the furthest, find third-party OS/DA sources cost- effective. The fact that rural LECs and a significant number of competitive LECs and interexchange carriers presently either self-provision these services or rely on wholesale providers for their OS/DA services constitutes substantial evidence that the cost of transport does not materially diminish the ability to provide service. 468. Because OS/DA databases are available on a value added and nondiscriminatory basis under section 251(b)(3) of the Act, a competing carrier need only provide transport to an incumbent's LEC's database. We acknowledge that self- provisioning OS/DA service may require competing carriers to incur substantial start-up costs that may represent a high percentage of overall expenses until call volumes and customer penetration levels rise. We find, however that the costs of self-provisioning OS/DA do not impair a requesting carrier's ability to provide service because in addition to self-provisioning, there are multiple alternatives available in the market. In addition, regional or nationwide OS/DA call centers enable competitive carriers to aggregate call volume to reach sufficient economies of scale. We note too that carriers are not limited to self-provisioning. Carriers may choose instead to use alternative OS/DA providers, reducing the fixed costs of provisioning OS/DA services. Moreover, competitive carriers who wish to obtain OS/DA from the incumbent may do so consistent with the incumbent LEC's nondiscriminatory access obligations under section 251(b)(3). 469. Quality. We find that the functionality of third-party supplied OS/DA is sufficiently equivalent to that of the incumbent's services such that a requesting carrier's ability to provide the services it seeks to offer is not impaired without access to the incumbent's OS/DA service. Although we acknowledge that differences in quality may exist, we find that, in light of the full scope of OS/DA options available to requesting carriers, the differences identified in this proceeding do not materially diminish a requesting carrier's ability to offer local exchange or exchange access service. 470. Specifically, we find that lack of unbundled OS/DA service from the incumbent LEC does not materially diminish the ability of requesting carriers to provide the service they seek to offer; several carriers have successfully self-provisioned OS/DA, while other carriers rely upon alternative providers of OS/DA services. Requesting carriers, however, complain that the alternative sources for operator services and directory assistance are inferior because the information provided to customers is not as complete, and is not updated as frequently, as incumbent LEC databases. According to several commenters, incumbent LECs update their directory listing databases daily, and often on a real-time basis, as they complete service order processes. In contrast, alternative providers may obtain their data from sources such as yellow pages databases, scanned white page listings, postal service change of address forms, motor vehicle registration records, and voter registration records, which are not updated as often. Requesting carriers, however, have the ability, under section 251(b)(3), to obtain nondiscriminatory access to the incumbent LEC's, or any other competing LEC's, databases used in the provision of OS/DA. Where competitive LECs may obtain OS/DA information and services, directly or indirectly, from incumbent LEC sources, we do not find cognizable differences in the quality of that information or services. The record indicates that carriers that are entitled to access to incumbent LEC database information and updates, such as competitive LECs and interexchange carriers like MCI WorldCom, Sprint and AT&T, offer directory assistance on a wholesale basis to other competitive LECs. Additionally, we note that third-party OS/DA providers are often able to purchase incumbent LEC OS/DA database information and updates. We are therefore not persuaded that lack of unbundled access to incumbent LEC databases used in the provision of OS/DA necessarily results in quality differences that would materially diminish a requesting carrier's ability to offer service. 471. MediaOne claims that operators of alternative OS/DA providers may be unfamiliar with the names of the local communities because their call centers are often distantly located. We do not believe that this constitutes a material difference in quality. First, we note that MediaOne does not explain how an operator's proximity to the customer results in a difference in OS/DA service quality. Search strategies used by OS/DA operators can be based on fuzzy logic queries and phonetic spellings that enable operators to retrieve information without the exact spelling of, or familiarity with, a place or proper name. For local directory assistance, alternative providers also train their call center operators to be familiar with the localities and any necessary variations on word pronunciations. In addition, incumbents often maintain remote or regional call centers that are located outside the local calling area of a large percentage of the incumbent LEC's own customers Thus, the incumbent's operators may have no more familiarity with the names of particular locales in a geographic area than do the operators of a competitor. Thus, if a competitor wants to ensure that the operators it is utilizing are trained for a particular area, it can best achieve this result by self-provisioning OS/DA service and training its own operators. Alternatively, a competitive carrier may also select an alternative OS/DA provider with a call center closer to the carrier's customer base than the incumbent's call center or contract with the provider for special operator training to cover the names of locales within the specific geographic markets the competitive carrier serves. We are satisfied that operator-training disparities between vendor-provided operators and those of the incumbent LEC do not materially diminish a requesting carrier's ability to offer service. 472. We reject arguments that we should unbundle access to the incumbent's OS/DA service because national operator services have limited ability to connect to local public safety answering points (PSAPs) in emergency situations. Specifically, certain commenters argue that in such situations, national operator services usually advise the caller to hang-up and dial 911. While issues of public safety are of paramount concern, the standard by which we decide to unbundle a non-proprietary network element focuses on whether a carrier's ability to provide the services that it seeks to offer is impaired by lack of access to that element. Accordingly, we look to whether the ability or inability to connect OS/DA calls to a PSAP impairs the ability of a carrier to offer local exchange services. We conclude that it does not. 473. Although subscribers may mistakenly dial OS/DA to reach emergency assistance, the ability to connect a misdirected call to a PSAP is unlikely to result in a competitive advantage in the provision of local exchange service. At least one third-party provider of OS/DA service, Teltrust, states that it requires its customers to provide the emergency number of the PSAP for the originating caller so that it knows which agency to call. In cases where it receives an incoming call from an 800 number and does not have an emergency number associated with the calling party's location, the operator can call emergency services if the calling party can provide the name of the location. Should a competitive carrier decide to obtain OS/DA services for its customers from the incumbent on a nondiscriminatory basis, under section 251(b)(3), it will be able to connect its customers to the PSAP in the same manner as the incumbent. Moreover, it is not clear whether all incumbent LEC OS/DA call centers, especially those with remote OS/DA call centers, have the ability to connect their own customers to every PSAP. Thus, even if a requesting carrier had unbundled access to the incumbent's OS/DA service, its subscribers may receive instructions from the incumbent's operator that do not measurably differ from the instructions it would receive from an alternative provider's operator. Indeed the only way in which a competitor can retain control over the quality of OS/DA service is to self-provide its own OS/DA call center and train its own operators. By self-providing its own call centers it can require its customers to provide it with detailed emergency information and populate its database accordingly. 474. We find insufficient evidence in the record to suggest that, based on performance measurements, there is a material difference in the timelines with which an incumbent's operator, compared to third-party operators, can respond to an inquiry. MediaOne asserts that the average speed to answer OS/DA calls for competitors is 15-18 seconds, while the incumbent commits to answering calls to its OS/DA platform in less than six seconds. The data MediaOne provides, while helpful, is inconclusive. Specifically, the data, which consists of the performance of one incumbent and a few competitors, provides too small a sample size for us to extrapolate these results over the entire OS/DA industry and conclude that competitive carriers' ability to provide service is impaired. While we acknowledge that there are likely to be some measurable differences among OS/DA providers for particular OS/DA components, we do not find sufficient record evidence to conclude that a requesting carrier is impaired without access to the incumbent's OS/DA service. Moreover, applying the unbundling standard we set forth above, the question of whether lack of access to the incumbent's network element materially diminishes a requesting carrier's ability to provide the services it seeks to offer is determined based on the totality of the circumstances. Thus, while relevant, we cannot say that the proffered average speed to answer calls, or other OS/DA quality issues, contribute significantly to a competitor's overall ability to provide local exchange and exchange access service. 475. Timeliness. We do not find any impediments associated with self-provisioning OS/DA services that would delay a requesting carrier's entry into the local exchange or exchange access market. Although AT&T identifies delays associated with implementing the customized routing necessary to use alternative OS/DA providers, the record indicates that AT&T's customized routing issues have been resolved. We are unaware of any ongoing problems that create material delays when competing carriers purchase OS/DA service from alternative providers. We agree that customized routing is necessary to access alternative sources of OS/DA for competitors not deploying their own switches. Commenters state that a key component of providing carriers with a choice of competitive OS/DA suppliers is the availability of line class codes in the unbundled switching element. Lack of a customized routing solution that enables competitors to route traffic to alternative OS/DA providers would therefore effectively preclude competitive LECs from using such alternative providers. Thus, if an incumbent LEC does not provide customized routing to requesting carriers that use the incumbent's unbundled switching element, it must provide unbundled access to its OS/DA service. 476. Impact on Network Operations. We conclude that the interoperability issues identified in the record do not materially diminish a requesting carrier's ability to provide local exchange or exchange access service. In particular, MCI WorldCom complains that incumbent LECs should implement Feature Group D signaling, instead of outdated legacy signaling protocol. According to MCI WorldCom, to use the incumbent LECs' signaling protocol instead of Feature Group D, most competitive LECs would have to either deploy new customized operator platforms or modify their existing platforms, both of which impose substantial costs. SBC responds that the customized routing of Feature Group D is not technically feasible in all end-office switches. BellSouth, however, offers a technical solution to MCI WorldCom's concern in some of its offices and states its willingness to deploy these solutions throughout its network. In instances where the requesting carrier obtains the unbundled switching element from the incumbent, the lack of customized routing effectively precludes requesting carriers from using alternative OS/DA providers and, consequently, would materially diminish the requesting carrier's ability to provide the services it seeks to offer. Thus, we require incumbent LECs, to the extent they have not accommodated technologies used for customized routing, to offer OS/DA as an unbundled network element. 477. Finally, we find that the ability to obtain nondiscriminatory access to operator services and directory assistance under section 251(b)(3) significantly mitigates any potential impairment a requesting carrier may experience if denied access to the incumbent's OS/DA services as an unbundled network element. There are a substantial number of regional and national alternative providers of OS/DA service that are serving a variety of customers, including some incumbent LECs and IXCs. We do not find differences in cost, quality, timeliness, and ubiquity that would lead to the conclusion that requesting carriers' ability to provide local exchange and exchange access services would be materially diminished without access to the incumbent's OS/DA service as an unbundled network element. Rather, we find that these alternative sources of OS/DA service are available as a practical, economic, and operational matter. Moreover, we believe that not requiring that incumbent LECs to unbundle OS/DA service is consistent with the goals of the Act, because it will reduce competitors' reliance on the incumbent's network and create new opportunities for competitors of OS/DA service to differentiate their services through increased quality and decreased prices. CDLXXVIII. MISCELLANEOUS ISSUES A. Section 271-Related Issues 1. Background 479. Section 271(c)(2)(B) enumerates a competitive checklist that BOCs must comply with to obtain interLATA authority. In particular, prior to obtaining authority to provide long distance service, section 271(c)(2)(B) requires BOCs to demonstrate, among other things, that they are providing or "generally offering" to requesting carriers the following network elements: local loops, transport, switching, databases and signaling. 480. In the Notice, we sought comment on the interplay between the unbundling obligations of section 251(c), and the competitive checklist network elements of section 271. Among other things, we sought comment on what pricing standards would apply if a checklist network element were no longer required to be unbundled pursuant to section 251(c)(3), after considering the "necessary" and "impair" standards of section 251(d)(2). 481. Ameritech argues that if a network element on the checklist no longer needs to be unbundled, the item need not be provided to requesting carriers at prices predicated on our forward looking costs. Other commenters counter that the inclusion of network elements on the checklist is presumptive evidence that these elements must be unbundled, and thus, provided to requesting carriers at prices predicated on our forward looking costs. 1. Discussion 482. In this Order, we conclude that circuit switching and shared transport need not be unbundled in certain circumstances. Nonetheless, providing access and interconnection to these elements remains an obligation for BOCs seeking long distance approval. We therefore must decide what prices, terms, and conditions apply to these elements that no longer need to be unbundled. 483. We conclude that the prices, terms, and conditions set forth under sections 251 and 252 do not presumptively apply to the network elements on the competitive checklist of section 271. 484. The Commission must consider unbundling network elements in accordance with section 251(c)(3), while according due deference to the "necessary" and "impair" standards articulated in section 251(d)(2), and by the Supreme Court. The Commission must evaluate the network elements on the competitive checklist under the auspices of section 271. If a checklist network element is unbundled, the applicable prices, terms and conditions are determined in accordance with sections 251 and 252. If a checklist network element does not satisfy the unbundling standards in section 251(d)(2), the applicable prices, terms and conditions for that element are determined in accordance with sections 201(b) and 202(a). 485. Although section 271 does not specify that the checklist network elements must be provided in accordance with section 251(c)(3), the Commission nonetheless has independent authority to ensure that items (iv)-(vi) of the checklist are provided on a reasonable, nondiscriminatory basis. In Iowa Utils. Bd., the Supreme Court affirmed the Commission's regulatory authority over the pricing of section 251 unbundled network elements, rejecting the claim that this matter is reserved to the states. In reaching this conclusion, the Court held that the Commission's pricing authority resides broadly in section 201(b), which grants the agency authority to prescribe rules and regulations "as may be necessary in the public interest to carry out the provisions of this Act." 486. Section 201(b) provides a basis for the Commission to scrutinize the prices, terms, and conditions under which the checklist network elements are offered. Section 201(b) states that "[a]ll charges, practices, classifications, and regulations for and in connection with such communication services, shall be just and reasonable, and any such charge, practice, classification, or regulation that is unjust or unreasonable is hereby declared unlawful." Section 202(a) mandates that "[i]t shall be unlawful for any common carrier to make any unjust or unreasonable discrimination in charges, practices, classifications, regulations, facilities, or services for or in connection with like communication service." In addition, checklist items (vii) and (x) explicitly require "nondiscriminatory access" to OS/DA, databases, and signaling. 487. In circumstances where a checklist network element is no longer unbundled, we have determined that a competitor is not impaired in its ability to offer services without access to that element. Such a finding in the case of switching for large volume customers is predicated in large part upon the fact that competitors can acquire switching in the marketplace at a price set by the marketplace. Under these circumstances, it would be counterproductive to mandate that the incumbent offers the element at forward-looking prices. Rather, the market price should prevail, as opposed to a regulated rate which, at best, is designed to reflect the pricing of a competitive market. A. Combinations of Unbundled Loops and Transport Network Elements 488. A number of parties identify issues surrounding combinations of loop and transport network elements. In particular, several competitive LECs argue that the Commission should identify the "enhanced extended link" (EEL) as a separate network element or require incumbent LECs to provide requesting carriers access to loop and transport elements in combination, even if those elements are not currently combined. Incumbent LECs argue that, for loop transport elements that are currently combined requesting carriers should not be allowed to substitute such combinations of elements for existing, regulated special access services. According to incumbent LECs, allowing this substitution would either force them to increase local rates or undermine universal service. 1. Enhanced Extended Link a. Background 489. In the Local Competition Order, the Commission identified loops and transport as network elements subject to the unbundling obligation of section 251(c)(3). In rule 51.315(b), the Commission prohibited incumbents from separating network elements that are currently combined. In addition, the Commission adopted rules 51.315(c) (f) requiring incumbent LECs to combine unbundled network elements in any manner, even if those elements are not currently combined. The Eighth Circuit overturned a number of the Commission's rules, including rules 51.315(b) (f). Rule 51.315(b), however, was reinstated by the Supreme Court. In light of the reasoning set forth in the Court's opinion, the Commission asked the Eighth Circuit to reinstate rules 51.315(c) (f). 490. In the Notice, we sought comment on whether we should identify additional network elements beyond the seven listed in the Local Competition First Report and Order. We also sought comment on whether, in light of the Supreme Court's decision, we could require incumbent LECs to combine network elements that are not currently combined, such as an unbundled loop with unbundled transport. 491. In response to the Notice, a number of parties, including competitive LECs and state commissions, argue that we should either identify a new network element comprised of unbundled loop, multiplexing/concentrating equipment, and dedicated transport (the enhanced extended link or "EEL") or, alternatively, reinstate rules 51.315(c) (f) which require incumbent LECs to provide unbundled loop and transport elements on a combined basis. Incumbent LECs argue that we should not identify the EEL as a separate network element because it would constitute an unlawful combination of two or more elements not currently combined. The incumbent LECs also argue that we cannot reinstate rules 51.315(c) (f) because they are currently pending before the Eighth Circuit. a. Discussion 492. We decline to define the EEL as a separate network element in this Order. As discussed above, the Eighth Circuit is currently reviewing whether rules 51.315(c) (f) should be reinstated. We see no reason to decide now whether the EEL should be a separate network element, in light of the Eighth Circuit's review of those rules. 493. A number of commenters argue that we should reaffirm the Commission's decision in the Local Competition First Report and Order. In that order the Commission concluded that the proper reading of "currently combines" in rule 51.315(b) means "ordinarily combined within their network, in the manner which they are typically combined." Incumbent LECs, on the other hand, argue that rule 51.315(b) only applies to unbundled network elements that are currently combined and not to elements that are "normally" combined. Again, because this matter is currently pending before the Eighth Circuit, we decline to address these arguments at this time. 494. We note that in the Local Competition First Report and Order, and again in this proceeding, we identify the loop and dedicated transport as separate unbundled network elements. In particular, as discussed above, we define the loop as the functionality that extends from the customer demarcation point to the main distribution frame associated with the incumbent LEC's central office switch. We define dedicated transport as the transmission facilities dedicated to a particular customer between wire centers owned by the incumbent LECs or requesting telecommunications carriers, or between switches owned by incumbent LECs or requesting carriers. To the extent an unbundled loop is in fact connected to unbundled dedicated transport, the statute and our rule 51.315(b) require the incumbent to provide such elements to requesting carriers in combined form. Thus, although in this Order, we neither define the EEL as a separate unbundled network element nor interpret rule 51.315(b) as requiring incumbents to combine unbundled network elements that are "ordinarily combined," we note that in specific circumstances, the incumbent is presently obligated to provide access to the EEL. In particular, the incumbent LECs may not separate loop and transport elements that are currently combined and purchased through the special access tariffs. Moreover, requesting carriers are entitled to obtain such existing loop-transport combinations at unbundled network element prices. 495. We also decline at this time to reinstate rules 51.315(c) (f). As discussed above, this issue is currently pending before the Eighth Circuit. As a general matter, however, we believe that the reasoning of the Supreme Court's decision to reinstate rule 51.315(b) based on the nondiscrimination language of section 251(c)(3) applies equally to rules 51.315(c) (f). Specifically, the Court held that section 251(c)(3)'s nondiscrimination requirement means that access provided by the incumbent LEC must be at least equal in quality to that which the incumbent LEC provides to itself. We note that incumbent LECs routinely combine loop and transport elements for themselves. For example, incumbent LECs routinely provide combinations of loop and transport elements for themselves in order to: (1) deliver data traffic to their own packet switches; (2) provide private line services; and (3) provide foreign exchange service. In addition, we note that incumbent LECs routinely provide the functional equivalent of the EEL through their special access offerings. 496. We believe that the basis upon which the Eighth Circuit invalidated rules 51.315(c) (f) has been called into question by the Supreme Court's decision. In particular, the Eighth Circuit determined that "unbundled" meant physical separation of network elements. The Supreme Court clarified that "unbundled" means "separate prices." The Supreme Court also stated that section 251(c) "does not say, or even remotely imply, that elements must be provided [in discrete pieces, and never in combined form.]" We also note that an additional basis for the Eighth Circuit's decision to invalidate rules 51.315(b) (f) was its understanding that incumbents "would rather grant their competitors access to their facilities" than combine elements on behalf of requesting carriers. Experience over the last year demonstrates that incumbent LECs have refused to provide access to network elements so that competitors could combine them, except in situations where competitive LECs have collocated in the incumbent's central offices. Accordingly, we believe that section 251(c)(3) provides a sound basis for reinstating rules 51.315(c) (f). 1. Use of unbundled network elements to provide exchange access services a. Background 497. As discussed above, in some situations in the incumbent's network, loops and dedicated transport network elements are already combined to provide special access services for interexchange carriers. In ex parte filings, incumbent LECs, including BellSouth and SBC, argue that the Commission should restrict a requesting carrier from obtaining such combined facilities as unbundled network elements in order to prevent requesting carriers from by-passing existing special access services. BellSouth and SBC both argue that such a restriction is necessary to prevent interexchange carriers from benefiting from the difference between special access rates and unbundled network element prices and thus, protect the incumbent LECs' current exchange access revenue streams. Competitive LECs respond that the plain language of section 251(c)(3) precludes the Commission from imposing any restrictions on the use of unbundled network elements. a. Discussion 498. Section 251(c)(3) of the Act requires incumbent LECs to provide to requesting carriers access to unbundled network elements "for the provision of a telecommunications service . . . ." In the Local Competition First Report and Order, the Commission found that section 251(c)(3) "permits interexchange carriers and all other requesting carriers, to purchase unbundled elements for the purpose of offering exchange access services, or for the purpose of providing exchange access services to themselves in order to provide interexchange services to consumers." In particular, the Commission found that its conclusion not to impose restrictions on the use of unbundled network elements was "compelled by the plain language of the 1996 Act" because exchange access and interexchange services are "telecommunications services." Moreover, in the Local Competition First Report and Order, the Commission found that "the language of section 251(c)(3), which provides that telecommunications carriers may purchase unbundled elements in order to provide a telecommunications service, is not ambiguous." This conclusion that the Act does not permit usage restrictions was codified in Rule 51.309(a), which provides that "[a]n incumbent LEC shall not impose limitations, restrictions, or requirements on request for, or the use of, unbundled network elements that would impair the ability of a requesting telecommunications carrier to offer a telecommunications service in the manner the requesting telecommunications carrier intends." That rule was not challenged in court by any party. 499. Parties have raised again arguments that allowing requesting carriers to use unbundled network elements to provide exchange access would have significant policy ramifications. As BellSouth explains, existing combinations of unbundled loops and transport network elements are a "direct (and often physically identical) substitute for the incumbent LEC's regulated access services . . . ," but priced significantly lower than tariffed special access services. The special access service that BellSouth and SBC refer to consists of entrance facilities from the interexchange carrier's point of presence (POP) to an incumbent LEC's switch or serving wire center (SWC), a dedicated transport link from the SWC to an end office, and a channel termination facility from the end office to the end user. 500. As an initial matter, under existing law, a requesting carrier is entitled to obtain existing combinations of loop and transport between the end user and the incumbent LEC's serving wire center on an unrestricted basis at unbundled network element prices. In particular, any requesting carrier that is collocated in a serving wire center is free to order loops and transport to that serving wire center as unbundled network elements because those elements meet the unbundling standard, as discussed above. Moreover, to the extent those unbundled network elements are already combined as a special access circuit, the incumbent may not separate them under rule 51.315(b), which was reinstated by the Supreme Court. In such situations, it would be impermissible for an incumbent LEC to require that a requesting carrier provide a certain amount of local service over such facilities. 501. Moreover, we wish to make clear that in situations where the requesting carrier is collocated and has self-provisioned transport or obtained transport from an alternative provider, but is purchasing unbundled loops, that carrier may provide only exchange access over those facilities. Thus, for instance, a requesting carrier is entitled to purchase unbundled loops in order to provide advanced services (e.g., interstate special access xDSL service). 502. Finally, we clarify that interexchange carriers are entitled to use unbundled dedicated transport from their POP to a serving wire center in order to provide local telephone exchange service. Such carriers are entitled to obtain such dedicated transport links pursuant to the unbundling standard discussed above. The fact that such carriers may also provide exchange access over those facilities does not alter our conclusion. 503. We conclude that the record in this phase of the proceeding is insufficient for us to determine whether or how our rules should apply in the discrete situation involving the use of dedicated transport links between the incumbent LEC's serving wire center and an interexchange carrier's switch or point of presence (or "entrance facilities"). Only a handful of parties commented on the special access arbitrage issue that was first raised by BellSouth's August 9, 1999, ex parte filing. We believe that we should fully explore the policy ramifications of applying our rules in a way that potentially could cause a significant reduction of the incumbent LECs' special access revenues prior to full implementation of access charge and universal service reform. Therefore, we set certain discrete issues for further comment below. A. Nondiscrimination Obligations of Incumbent LECs 504. We reaffirm the conclusion the Commission adopted in the Local Competition First Report and Order that national rules defining "nondiscriminatory access" to unbundled network elements will reduce the costs of entry and speed the development of competition in local telecommunications markets. We find that the phrase "nondiscriminatory access" in section 251(c)(3) means at least two things: first, the quality of an unbundled network element that an incumbent LEC provides, as well as the access provided to that element, must be equal between all carriers requesting access to that element; second, where technically feasible, the access and unbundled network element provided by an incumbent LEC must be provided in "substantially the same time and manner" to that which the incumbent provides to itself. 505. In those situations where an incumbent LEC does not provide access to network elements to itself, we reaffirm our requirement that incumbent LECs must provide access in a manner that provides a requesting carrier with a meaningful opportunity to compete. Because we believe that the technical infeasibility problem will arise rarely, we expect incumbent LECs to fulfill the non-discrimination requirement in nearly all instances where they provision unbundled network elements. In the rare instances where technical feasibility issues arise, incumbent LECs must prove to a state commission that it is technically infeasible to provide access to unbundled elements at the same level of quality that the incumbent LEC provides to itself. DVI. FOURTH FURTHER NOTICE OF PROPOSED RULEMAKING A. Background 507. As noted above, in the Local Competition First Report and Order, the Commission held that for all unbundled network elements, including combinations of network elements, incumbent LECs may not impose any usage restriction on the use of such elements, or combinations thereof. In that order, however, the Commission imposed a temporary access charge on the purchase of unbundled switching. In particular, the Commission required requesting carriers to pay, for a limited time period, the carrier common line charge (CCL) and 75 percent of the Tandem Interconnection Charge (TIC). The Commission found that it had discretion under the Act "to adopt a limited, transitional plan to address public policy concerns raised by the bypass of access charges via unbundled elements." This decision was upheld by the Eighth Circuit, which found that the Commission decision was reasonable. 508. In the Third Order on Reconsideration, the Commission required incumbent LECs to provide access to shared transport as an unbundled network element in conjunction with local and tandem switching. In that order, the Commission limited the obligation of incumbent LECs to provision shared transport to end users to whom the requesting carrier was providing local exchange service. The Commission sought comment on whether requesting carriers may use unbundled dedicated or shared transport facilities in conjunction with unbundled switching, to originate or terminate interstate toll traffic to customers to whom the requesting carrier does not provide local exchange service. Specifically, the Further Notice of Proposed Rulemaking requested comment on the "intensely interrelated" question of whether such use would conflict with the Commission's implementation of access charge reform and universal service. A. Discussion 509. Parties have argued in this proceeding that allowing requesting carriers to obtain combinations of loop and transport unbundled network elements based on forward- looking cost would provide opportunities for arbitrage of special access services. We are cognizant that special access pre-dates passage of the 1996 Telecommunications Act and has historically been provided by incumbent LECs at prices that are higher than the unbundled network element pricing scheme of section 252(d)(1). Accordingly, in this Further Notice we consider whether there is any basis in the statute or our rules under which incumbent LECs could decline to provide entrance facilities at unbundled network element prices. 510. We seek comment on the argument that the "just and reasonable" terms of section 251(c) or section 251(g) permit the Commission to establish a usage restriction on entrance facilities. Parties should also address whether there is any other statutory basis for limiting an incumbent LEC's obligation to provide entrance facilities as an unbundled network element. 511. We acknowledge that resolution of this issue potentially could have large financial impact on incumbent local exchange carriers. We seek comment on this issue, and on the extent to which any such impact should be considered in reaching a decision on this issue. We seek comment on the policy implications, if any, of a significant reduction in special access revenues for our universal service program. Finally, because the record developed in the Further Notice of Proposed Rulemaking in the Shared Transport Order is two years old, we invite parties to refresh the record on whether requesting carriers may use unbundled dedicated or shared transport facilities in conjunction with unbundled switching to originate or terminate interstate toll traffic to customers to whom the requesting carrier does not provide local exchange service. DXII. PROCEDURAL ISSUES A. Final Regulatory Flexibility Analysis 513. As required by the Regulatory Flexibility Act (RFA), an Initial Regulatory Flexibility Analysis (IRFA) was incorporated in the Notice in CC Docket No. 96- 98. The Commission sought written public comments on the proposals in the Notice, including comments on the IRFA. This present Final Regulatory Flexibility Analysis (FRFA) conforms to the RFA. 1. Need for, and Objectives of, the Third Report and Order 514. This Order responds to the Supreme Court's January, 1999, decision that directs the Commission to revise the standards used to determine which network elements incumbent LECs must unbundle pursuant to section 251 of the Act. More specifically, this Order gives substance to the "necessary" and "impair" standards in section 251(d)(2) of the Act. Applying these standards, and considering the availability of elements outside of the incumbent's network, this Order adopts a list of network elements that must be unbundled on a national basis, subject to certain discrete geographic and product market exceptions. This Order also announces that the Commission will reexamine the national list of unbundled network elements in three years. It reaffirms a state commission's authority to require incumbent LECs to unbundle additional elements, as long as the unbundling obligations: (1) are consistent with the requirements of section 251; (2) do not substantially prevent implementation of the requirements of that section and the purposes of the Act; and (3) are consistent with the national policy framework established in this Order. Finally the Order reaffirms that incumbent LECs are obligated to offer combinations of network elements that are already combined, including combinations of loop, multiplexing/concentrating equipment, and dedicated transport if they are currently combined. 1. Summary of Significant Issues Raised by the Public Comments in Response to the IRFA 515. We received no comments in response to the IRFA in the Notice. We did, however, receive some general small-business-related comments which are discussed throughout the Order and are summarized in subsection 5 of the FRFA, infra. 1. Description and Estimate of the Number of Small Entities to which Rules will Apply 516. In the FRFA to the Commission's Local Competition First Report and Order, we adopted the analysis and definitions set forth in determining the small entities affected by this Order for purposes of this FRFA. The RFA directs agencies to provide a description of and, where feasible, an estimate of the number of small entities that will be affected by rules. The RFA generally defines the term "small entity" as having the same meaning as the terms "small business," "small organization," and "small governmental jurisdiction." The RFA defines a "small business" to be the same as a "small business concern" under the Small Business Act, unless the Commission has developed one or more definitions that are appropriate to its activities. Under the Small Business Act, a "small business concern" is one that: (1) is independently owned and operated; (2) is not dominant in its field of operation; and (3) meets any additional criteria established by the Small Business Administration (SBA). Below we further describe and estimate the number of small entities that may be affected by the rules adopted in this Order. 517. We have included small incumbent LECs in this RFA analysis. As noted above, a "small business" under the RFA is one that, inter alia, meets the pertinent small business size standard (e.g., a telephone communications business having 1,500 or fewer employees), and "is not dominant in its field of operation." The SBA's Office of Advocacy contends that, for RFA purposes, small incumbent LECs are not dominant in their field of operation because any such dominance is not national in scope. We have therefore included small incumbent LECs in this RFA analysis, although we emphasize that this RFA action has no effect on the Commission's analyses and determinations in other non-RFA contexts. 518. The United States Bureau of the Census (the Census Bureau) reports that at the end of 1992, there were 3,497 firms engaged in providing telephone services, as defined therein, for at least one year. These firms include a variety of different categories of carriers, including LECs, interexchange carriers, competitive access providers, wireless providers, operator service providers, pay telephone operators, wireless providers, and resellers. At least some of these 3,497 telephone service firms may not qualify as small entities because they are not "independently owned and operated." For example, a wireless provider that is affiliated with a LEC having more than 1,500 employees would not meet the definition of a small business. It seems reasonable to conclude, therefore, that fewer than 3,497 of these telephone service firms are small entities that may be affected by this Order. Since 1992, however, many new carriers have entered the telephone services marketplace. At least some of these new entrants may be small entities that are affected by this Order. 519. The SBA has developed a definition of small entities for telephone communications companies other than radiotelephone (wireless) companies. The Census Bureau reports that there were 2,321 such telephone companies that had been operating for at least one year at the end of 1992. According to the SBA's definition, a wireline telephone company is a small business if it employs no more than 1,500 persons. All but 26 of the 2,321 wireline companies listed by the Census Bureau were reported to have fewer than 1,000 employees. Thus, even if all 26 of those companies had more than 1,500 employees, there would still be 2,295 wireline companies that might qualify as small entities. Although it seems certain that some of these carriers are not independently owned and operated, we are unable at this time to estimate with greater precision the number of wireline carriers and service providers that would qualify as small business concerns under the SBA's definition. Consequently, we estimate that fewer than 2,295 of these wireline companies are small entities that this Order may affect. Since 1992, however, many wireline carriers have entered the telephone services marketplace. Many of these new entrants may be small entities that are affected by this Order. 520. Incumbent Local Exchange Carriers. Neither the Commission nor the SBA has developed a definition specifically directed toward small incumbent LECs. The closest applicable definition under SBA rules is for telephone communications companies other than radiotelephone (wireless) companies. The most reliable source of information regarding the number of LECs nationwide of which we are aware appears to be the data that we collect annually in connection with the Telecommunications Relay Service (TRS). According to our most recent data, 1,410 companies reported that they were engaged in the provision of local exchange services. Although it seems certain that some of these carriers are not independently owned and operated or have more than 1,500 employees, we are unable at this time to estimate with greater precision the number of small incumbent LECs that would qualify as small business concerns under SBA's definition. Consequently, we estimate that there are fewer than 1,410 small incumbent LECs that may be affected by the decisions and rules adopted in this Order. 1. Description of Projected Reporting, Recordkeeping, and Other Compliance Requirements 521. Pursuant to sections 251(c) and (d) of the 1996 Act, incumbent LECs, including those that qualify as small entities, are required to provide nondiscriminatory access to unbundled network elements. The only exception to this rule is those carriers that qualify and have gone through the process of obtaining an exemption, suspension or modification pursuant to section 251(f) of the Act. This Order interprets the necessary and impair standards of section 251(d)(2) in such a way that it fulfills the Supreme Court's requirement that we apply some limiting standard to an incumbent LEC's 251(c) obligations. In this Order, we identify a minimum set of network elements that incumbent LECs are obligated to offer to requesting carriers on an unbundled basis nationwide: (1) local loops, including dark fiber and high-capacity loops; (2) subloops; (3) network interface devices; (4) local switching, except under certain conditions; (5) interoffice transport; (6) signaling and call-related databases; (7) operations support systems; and (8) in very limited situations, packet switching. State commissions may require incumbent LECs to provide additional network elements on an unbundled basis. The Order also clarifies that incumbent LECs are obligated to provide access to combinations of loop, multiplexing/concentrating equipment and dedicated transport if they are currently combined. Compliance with the rules and decisions adopted in this Order may require the use of engineering, technical, operational, accounting, billing, and legal skills. 1. Steps Taken to Minimize the Economic Impact of this Order on Small Entities, and Alternatives Considered 522. As we concluded in the original FRFA, and as discussed more thoroughly above, we believe that our actions establishing a minimum national list of unbundled network elements in this Order facilitates the development of competition in the local exchange and exchange access markets. This decision decreases entry barriers and provides reasonable opportunities for all carriers, including small entities, to provide local exchange and exchange access services. 523. National requirements for unbundling allows requesting carriers, including small entities, to take advantage of economies of scale in network. Requesting carriers, which may include small entities, should have access to the same technologies and economies of scale and scope available to incumbent LECs. Having such access will facilitate competition and help lower prices for all consumers, including individuals and small entities. A minimum national list of unbundled network elements also should facilitate the development of consistent standards and help resolve issues without imposing additional litigation costs on parties, including small entities. 524. Establishing a minimum national list of unbundled network elements facilitates negotiations and reduces regulatory burdens for all parties, including small entities. Adopting a national list lowers requesting carrier's cost by enabling them to implement regional and/or national business plans. In reaching this conclusion we considered one proposal to adopt national standards that would be applied by state commissions on a market-by-market basis. We concluded that this approach would lead to greater uncertainty in the market and would hinder the development of competition. We also found that it would complicate the negotiation of interconnection agreements and lead to increased litigation. Furthermore, this approach would increase the administrative burden on state commissions and parties arbitrating interconnection agreements before these state commissions. All of these factors would slow the development of competition. Therefore we reaffirmed our decision in the Local Competition First Report and Order to adopt a national list. 1. Report to Congress 525. The Commission will send a copy of the Third Report and Order, including this FRFA, in a report to be sent to Congress pursuant to the Small Business Regulatory Enforcement Fairness Act of 1996. In addition, the Commission will send a copy of the Third Report and Order, including the FRFA, to the Chief Counsel for Advocacy of the Small Business Administration. The Third Report and Order and FRFA, or summaries thereof, will also be published in the Federal Register. A. Initial Regulatory Flexibility Analysis (IRFA) 526. As required by the RFA, the Commission has prepared this present Initial Regulatory Flexibility Analysis (IRFA) of the possible significant economic impact on small entities by the policies and rules proposed in this Fourth Further Notice of Proposed Rulemaking. Written public comments are requested on this IRFA. Comments must be identified as responses to the IRFA and must be filed by the deadlines for comments on the Fourth Further Notice of Proposed Rulemaking provided above in section VII. The Commission will send a copy of the Fourth Further Notice of Proposed Rulemaking, including this IRFA, to the Chief Counsel for Advocacy of the Small Business Administration. In addition, the Fourth Further Notice of Proposed Rulemaking and IRFA, or summaries thereof, will be published in the Federal Register. 1. Need for, and Objectives of, the Proposed Rules 527. In this proceeding commenters have argued that allowing requesting carriers to obtain combinations of loop and transport unbundled network elements based on forward-looking cost would provide opportunities for arbitrage of special access services. We recognize that special access has historically been provided by incumbent LECs at prices that are higher than the unbundled network element pricing scheme of section 252(d)(1). Accordingly, in this Fourth Further Notice, the Commission seeks comment on the legal and policy bases for precluding requesting carriers from substituting dedicated transport for special access entrance facilities. We ask whether there is any basis in the statute or our rules under which incumbent LECs could decline to provide entrance facilities at unbundled network element prices. 528. Finally, because the record developed in the Further Notice of Proposed Rulemaking in the Shared Transport Order is two years old, we invite parties to refresh the record on whether requesting carriers may use unbundled dedicated or shared transport facilities in conjunction with unbundled switching to originate or terminate interstate toll traffic to customers to whom the requesting carrier does not provide local exchange service. 1. Legal Basis 529. Sections 1-4, 10, 201, 202, 251-254, 271, and 303(r) of the Communications Act, as amended, 47 U.S.C.  151-54, 160, 201, 202, 251-54, 271, and 303(r). 1. Description and Estimate of the Number of Small Entities to Which the Proposed Rules Will Apply 530. In the FRFA in the Third Report and Order, supra, we have described the entities possibly affected by that decision. We anticipate that the same entities, as well as those described below, could be affected by any action taken in response to the Fourth Further Notice. We therefore incorporate the description and estimates used in the FRFA in the Third Report and Order, and add the following descriptions. 531. Competitive Local Exchange Carriers. Neither the Commission nor SBA has developed a definition of small entities specifically directed toward providers of competitive local exchange services. The most reliable source of information regarding the number of competitive LECs nationwide of which we are aware appears to be the data we collected in the August, 1999 Local Competition Report. According to our most recent data, 158 companies reported that they were local service competitors holding numbering codes. Although it seems certain that some of these carriers are not independently owned and operated, or have more than 1,500 employees, we are unable at this time to estimate with greater precision the number of competitive LECs that would qualify as small business concerns under SBA's definition. Consequently, we estimate that there are fewer than 158 small entity competitive LECs that may be affected by the decisions and rules adopted in response to the Fourth Further Notice of Proposed Rulemaking. 532. Competitive Access Providers. Neither the Commission nor SBA has developed a definition of small entities specifically directed toward providers of competitive access services (CAPs). The closest applicable definition under SBA rules is for telephone communications companies other than radiotelephone (wireless) companies. The most reliable source of information regarding the number of CAPs nationwide of which we are aware appears to be the data that we collect annually in connection with the TRS Worksheet. According to our most recent data, 129 companies reported that they were engaged in the provision of competitive access services. Although it seems certain that some of these carriers are not independently owned and operated, or have more than 1,500 employees, we are unable at this time to estimate with greater precision the number of competitive LECs that would qualify as small business concerns under SBA's definition. Consequently, we estimate that there are fewer than 129 small entity competitive LECs that may be affected by the decisions and rules adopted in response to the Fourth Further Notice of Proposed Rulemaking. 1. Description of Projected Reporting, Recordkeeping, and Other Compliance Requirements 533. If the Commission does not establish any restrictions on the use of unbundled network elements or combinations of network elements, no additional compliance requirements are anticipated from further consideration of this issue. If, however, restrictions on access to network elements are imposed, and depending on how the restrictions are imposed, competitive LECs, CAPs and other purchasers of unbundled network elements, including small entities, may be subject to additional reporting, recordkeeping and other compliance requirements. Incumbent LECs, including small incumbent LECs, would also be impacted because they would have to keep track of competitive LEC filings and whether the use of the unbundled network element changed in such a way that a restriction would attach. If restrictions are placed on the use of unbundled network elements or combinations of such elements, compliance with these requests may require the use of engineering, technical, operational, accounting, billing, and legal skills. 1. Steps Taken to Minimize Significant Economic Impact on Small Entities, and Significant Alternatives Considered 534. If requesting carriers can substitute unbundled network elements, such as transport, for entrance facilities, incumbent LECs, including small entities, may be significantly economically impacted. On the other hand, substituting unbundled network elements for entrance facilities could benefit competitive LECs, CAPs, and other purchasers of unbundled network elements. The Commission will evaluate in this proceeding whether there are legal grounds for restricting such access. If no such grounds exist, and instead if the statute requires unrestricted access to these unbundled network elements or combinations, then the Commission will have no alternative other than implementation of the statutory requirements for unrestricted access. 1. Federal Rules that May Duplicate, Overlap, or Conflict with the Proposed Rules 535. None. DXXXVI. ADMINISTRATIVE MATTERS 537. Pursuant to Sections 1.415 and 1.419 of the Commission's rules, 47 C.F.R.  1.415, 1.419, interested parties may file comments on or before January 12, 2000 and reply comments on or before February 11, 2000. Comments may be filed using the Commission's Electronic Comment Filing System (ECFS) or by filing paper copies. See Electronic Filing of Documents in Rulemaking Proceedings, 63 Fed. Reg. 24,121 (1998). 538. Comments filed through the ECFS can be sent as an electronic file via the Internet to . Generally, only one copy of an electronic submission must be filed. If multiple docket or rulemaking numbers appear in the caption of this proceeding, however, commenters must transmit one electronic copy of the comments to each docket or rulemaking number referenced in the caption. In completing the transmittal screen, commenters should include their full name, Postal Service mailing address, and the applicable docket or rulemaking number. Parties may also submit an electronic comment by Internet e-mail. To get filing instructions for e-mail comments, commenters should send an e-mail to ecfs@fcc.gov, and should include the following words in the body of the message, "get form