******************************************************** NOTICE ******************************************************** This document was converted from WordPerfect 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 ) ) Telephone Number Portability ) CC Docket No. 95-116 ) RM-8535 ) FIRST MEMORANDUM OPINION AND ORDER ON RECONSIDERATION Adopted: March 6, 1997 Released: March 11, 1997 By the Commission: Paragraph Table of Contents Number I. INTRODUCTION. . . . . . . . . . . . . . . . . . . . . . 1 II. BACKGROUND . . . . . . . . . . . . . . . . . . . . . . 2 A. First Report & Order . . . . . . . . . . . . . . . . 2 B. Number Portability Methods . . . . . . . . . . . . . 5 C. Current State Efforts. . . . . . . . . . . . . . . . 8 III. DISCUSSION. . . . . . . . . . . . . . . . . . . . . 11 A. Issues Relating to Long-Term Number Portability Methods 11 1. Performance Criteria . . . . . . . . . . . . 11 a. Background . . . . . . . . . . . . . . . 11 b. Pleadings. . . . . . . . . . . . . . . . 14 c. Discussion . . . . . . . . . . . . . . . . 19 (1) Service Degradation . . . . . . . . . 21 (2) Network Reliability . . . . . . . . . 25 (3) Intranetwork Use of QOR . . . . . . . 30 2. Public Interest Considerations. . . . . . . . . 31 a. Overview . . . . . . . . . . . . . . . . . 31 b. Purported Cost Savings Associated with QOR 33 c. Impact of QOR on the Implementation Schedule 44 d. Impact on the States . . . . . . . . . . . 46 e. Conclusion . . . . . . . . . . . . . . . . 47 B. Implementation Schedule for Wireline Carriers. . . . 48 1. Background. . . . . . . . . . . . . . . . . . . 48 2. Deployment Only in Requested Switches . . . . . 50 3. Extension of Implementation Schedule. . . . . . 72 4. Acceleration of Implementation Schedule . . . 100 5. Exemptions for Rural and/or Smaller LECs. . . 108 6. Implementation Requirements for Intermediate (N-1) Carriers 124 C. Implementation Schedule for Wireless Carriers. . . 127 D. Deferral of Implementation Until Resolution of Cost Recovery Issues 143 IV. ORDERING CLAUSES . . . . . . . . . . . . . . . . . . 149 APPENDIX A List of Parties APPENDIX B Final Rules APPENDIX C Description of Number Portability Methods APPENDIX D Supplemental Final Regulatory Flexibility Analysis APPENDIX E Implementation Schedule I. INTRODUCTION 1. On June 27, 1996, the Commission adopted the First Report and Order and Further Notice of Proposed Rulemaking (First Report & Order) in this docket implementing the requirement under Section 251(b) of the Communications Act of 1934, as amended (the Act), that all local exchange carriers (LECs) offer, "to the extent technically feasible, number portability in accordance with requirements prescribed by the Commission." By this action, we resolve certain petitions for reconsideration or clarification of our number portability rules adopted in the First Report & Order. Twenty-two parties filed petitions for reconsideration or clarification, nineteen parties filed oppositions or comments on the petitions, and sixteen parties filed reply comments. While the petitions raise a broad range of issues, we address three primary issues in this First Memorandum Opinion and Order on Reconsideration (First Reconsideration Order). We will address the remaining issues in one or more subsequent reconsideration orders in this docket. First, we conclude that Query on Release (QOR) is not an acceptable long-term number portability method. Second, we extend the completion deadlines in the implementation schedule for wireline carriers by three months for Phase I and by 45 days for Phase II, clarify the requirements imposed thereunder, and address issues raised by rural LECs and certain other parties. Finally, we affirm and clarify our implementation schedule for wireless carriers. II. BACKGROUND A. First Report & Order 2. Pursuant to the statutory requirement of Section 251(b), the First Report & Orderrequires all LECs to implement a long-term number portability method in the 100 largest Metropolitan Statistical Areas (MSAs) according to a phased deployment schedule that commences October 1, 1997, and concludes December 31, 1998. Thereafter, in areas outside the 100 largest MSAs, each LEC must make long-term number portability available within six months after a specific request by another telecommunications carrier. The First Report & Orderalso requires all cellular, broadband personal communications services (PCS), and covered Specialized Mobile Radio (SMR) providers to be able to deliver calls from their networks to ported numbers by December 31, 1998, and requires cellular, broadband PCS, and covered SMR providers to offer number portability throughout their networks and have the capability to support roaming nationwide by June 30, 1999. 3. Rather than choosing a particular technology for the provision of number portability, the Commission established performance criteria that any long-term number portability method selected by a LEC must meet. The Commission noted, however, that one of the criteria it adopted effectively precludes carriers from implementing QOR. The First Report & Order further concludes that long-term number portability should be provided through a system of regional databases that will be managed by one or more independent administrators selected by the North American Numbering Council (NANC). 4. The First Report & Order also requires wireline LECs, pending their deployment of a long-term number portability method, to provide currently available number portability measures upon request by another telecommunications carrier. Consistent with Section 251(e)(2) of the Communications Act, the First Report & Order sets forth principles that ensure that the costs of currently available measures are borne by all telecommunications carriers on a competitively neutral basis, and permits states to utilize various cost recovery mechanisms, so long as they are consistent with these statutory requirements and our principles. The Commission also concurrently adopted a Further Notice of Proposed Rulemaking (Further Notice) seeking comment on cost recovery for long-term number portability. B. Number Portability Methods 5. Because most telephone numbers within the North American Numbering Plan (NANP) are associated with a particular switch operated by a particular service provider, they currently cannot be transferred outside the service area of a particular switch or between switches operated by different service providers without technical changes to the switch or network. Industry participants have developed several methods for providing service provider portability that would be suitable for long-term use by carriers. These methods for providing long-term number portability employ databases containing the customer routing information necessary to route telephone calls to the proper terminating locations. All of these methods depend on Intelligent Network (IN) or Advanced Intelligent Network (AIN) capabilities. 6. While various methods for providing long-term number portability have been developed, two methods have emerged as the primary ones advocated by parties in this proceeding: Location Routing Number (LRN) and Query on Release (QOR). Under LRN, a unique 10-digit number, or location routing number, is assigned to each central office switch. Carriers routing telephone calls to customers that have transferred their telephone numbers from one carrier to another perform a database query to obtain the location routing number that corresponds to the dialed telephone number. The database query is performed for all calls to switches from which at least one number has been ported. The carrier then routes the call to the new carrier based on the location routing number. 7. QOR, also known as Look Ahead, is a triggering mechanism that operates in conjunction with the LRN addressing scheme. Under QOR, the signalling used to set up a telephone call is routed to the end office switch to which the dialed telephone number was originally assigned (known as the donor switch or the release switch) according to the NPA- NXX of the dialed number. If the dialed number has been transferred to another carrier's switch, the release switch sends a release message back, and the previous switch in the call path queries the database to obtain the routing information. The call is then completed to the new carrier's switch. C. Current State Efforts 8. Prior to the adoption of our First Report & Order, a number of state commissions had selected LRN as the method for implementing number portability in areas within their states' boundaries. These states include Colorado, Georgia, Illinois, Maryland, New York, and Ohio. On August 2, 1996, the California Public Utilities Commission (CA PUC) issued an order mandating the use of LRN as the long-term number portability method to be implemented in California. No states have selected QOR as the preferred method for long-term number portability. 9. Since adoption of the First Report & Order, planning and implementation of long- term number portability has progressed significantly. A number of state commissions have spent the past eight months developing state-specific plans for implementing LRN and resolving technical issues associated with the deployment of LRN. For example, the Illinois Commerce Commission Number Portability Workshop (ICC Workshop) remains in the forefront with respect to implementation of number portability. It has developed a detailed LRN test plan and has resolved numerous operational issues relating to switching, signalling, and SCP requirements. In addition, the Maryland Public Service Commission has determined a ranking and timeline for deployment of LRN in every switch in Maryland; established a comprehensive operations plan for LRN implementation; resolved issues relating to interfaces, ordering, provisioning, repair and maintenance processes as well as operator services; and studied switch and SCP requirements. Number portability task forces in Indiana, Michigan, and Ohio have also developed switch-specific implementation plans for those states. 10. The industry, under the auspices of the NANC, has been working on the design of the number portability regional database system. The NANC, a Federal Advisory Committee established under the provisions of the Federal Advisory Committee Act to advise the Commission on numbering issues, held its first meeting on October 1, 1996. The NANC's Local Number Portability Administration (LNPA) Selection Working Group and its task forces have been meeting regularly to assist the NANC in recommending to the Commission resolution of issues related to the selection and duties of an entity or entities to serve as the local number portability administrator(s), the database architecture plan, and the technical and operational requirements for the number portability database system. The NANC has committed to making its recommendation to the Commission on LNPA issues by May 1, 1997. Under NANC oversight, carriers in Illinois, Georgia, California, Maryland, Colorado, New York, and Texas have formed a Limited Liability Corporation (LLC) and issued a Request for Proposal (RFP) for each state to construct and maintain a number portability database. Each LLC has contacted neighboring states seeking to expand these state databases into regional databases covering the RBOC service areas. The LNPA Selection Working Group projects that all seven regional databases will be ready for testing on dates ranging from April 18, 1997, to July 1, 1997, and will be ready to support number portability deployment on or before October 1, 1997, in accordance with the deployment schedule set forth in the First Report & Order. III. DISCUSSION A. Issues Relating to Long-Term Number Portability Methods 1. Performance Criteria a. Background 11. The Act requires all LECs "to provide, to the extent technically feasible, number portability in accordance with requirements prescribed by the Commission." The Act states that "[t]he term 'number portability' means the ability of users of telecommunications services to retain, at the same location, existing telecommunications numbers without impairment of quality, reliability, or convenience when switching from one telecommunications carrier to another." We interpreted these statutory provisions to mean that the Commission should develop a national number portability policy and prescribe the requirements that all local exchange carriers, both incumbents and others, must meet to satisfy their statutory obligations. 12. In the First Report & Order, we concluded that establishing performance criteria that a LEC's number portability architecture must meet would better serve the public interest than choosing a particular technology or specific architecture. We thus adopted the following nine minimum criteria, which require that any long-term number portability method must: (1) support existing network services, features, and capabilities; (2) efficiently use numbering resources; (3) not require end users to change their telecommunications numbers; (4) not require telecommunications carriers to rely on databases, other network facilities, or services provided by other telecommunications carriers in order to route calls to the proper termination point; (5) not result in unreasonable degradation in service quality or network reliability when implemented; (6) not result in any degradation of service quality or network reliability when customers switch carriers; (7) not result in a carrier having a proprietary interest in any long-term method; (8) be able to accommodate location and service portability in the future; and (9) have no significant adverse impact outside the areas where number portability is deployed. We concluded that a number of these criteria implement the statutory requirement that customers switching their carrier be able to retain their numbers "without impairment of quality, reliability, or convenience." 13. In addition, we concluded that criterion four precludes carriers from using such number portability methods as QOR. When discussing criterion four, we stated that carriers may experience several undesirable effects if they are forced to rely on the networks of their competitors in order to route calls. For example, the use of number portability methods that first route the call through the original service provider's network in order to determine whether the call is to a ported number, and then perform a query only if the call is to be ported, would treat ported numbers differently than non-ported numbers, resulting in ported calls taking longer to complete than unported calls. This differential in efficiency would disadvantage the carrier to whom the call was ported and impair that carrier's ability to compete effectively against the original service provider. b. Pleadings 14. Most petitioners focus primarily on our conclusion in the First Report & Orderthat incumbent LECs are prohibited from using QOR as a long-term number portability method. They argue that QOR does not violate the performance criteria the Commission established, and that QOR has many public interest benefits that will make it more efficient, easier, and less costly to deploy than other number portability methods. They assert that, at a minimum, we should allow a carrier to use QOR within its own network in order to route calls made by its own customers to NXXs assigned to that carrier. Furthermore, they claim that a LEC's "intranetwork" use of QOR would not impact other carriers, nor would it present network interoperability issues. Some petitioners also argue that we should allow the use of QOR between networks if the carriers mutually agree to do so. 15. With respect to the performance criteria, petitioners assert that QOR does not violate performance criterion four (i.e., it does not "require dependency on another carrier's network") any more than other number portability methods, such as LRN. For example, Pacific and USTA argue that there is no rational basis for the Commission to conclude that a number portability method that requires an SS7 message to be sent to a switch to which the NXX code of the called number has been assigned (the "essence" of QOR) causes "undue reliance" on the networks of other carriers, while a number portability method that requires an SS7 message to be sent to the incumbent's Service Control Point (SCP) (the "essence" of LRN) does not involve such undue reliance. Because every number portability method requires some dependence on another carrier's network facilities, several petitioners suggest that we eliminate criterion four as one of the performance criteria that a long-term number portability method must meet. 16. Petitioners also argue that QOR does not result in any degradation of service quality or network reliability when customers switch carriers (performance criterion six). These parties contend that claims that use of QOR will cause significant additional delays in the routing of calls to ported numbers are unfounded. They allege that any additional delay attributable to the use of QOR is insignificant and imperceptible, as compared to the delay associated with LRN. U S West urges the Commission to conduct a survey to determine whether callers would perceive differences in call set-up time between LRN and QOR. In addition, NYNEX asserts that QOR would impose post-dial delay on a much smaller set of calls than LRN, because LRN without QOR increases post-dial delay for all interswitch calls, whether ported or not. USTA, on the other hand, argues that the standard should not be whether network routing for ported and non-ported numbers is identical, but whether service quality is discriminatory. 17. With respect to network reliability, proponents of QOR assert that QOR is less likely to threaten the reliability of the network than LRN, because QOR requires far fewer database queries than LRN. These parties contend that QOR imposes less of a burden on the SS7 network and, therefore, poses a lower risk of a network outage. In a late-filed ex partepresentation, SBC submitted a network reliability study conducted by Bellcore that purportedly demonstrates that there is a 0.036 percent (0.00036) probability of a "catastrophic outage" if LRN is implemented under what they characterize as a "normal" schedule, and no probability of such a catastrophic network outage if QOR is implemented under a "normal" schedule and less than one percent of the numbers are ported. MCI and AT&T vigorously dispute the purported findings of the Bellcore study, arguing, among other things, that the figures for "catastrophic outage" assume that all number portability databases in Houston fail simultaneously, which they argue is a highly improbable scenario, given that SBC has never experienced a single dual SCP failure, much less a dual failure of all SCPs. MCI also notes that, according to the Bellcore study, probabilities for FCC reportable outages with LRN and QOR are virtually identical under the same scheduling scenarios. In response, Bellcore argues that the types of failures contemplated by the Bellcore study are not "too improbable to be of concern" as AT&T claims, because a combination of events and errors has resulted in various switch failures and outages in the past. Thus, Bellcore asserts that, even though a complete failure has not occurred, there is certainly a reasonable probability that such an event could occur in the future. In fact, Bellcore claims that such partial and complete failures have been relatively rare in the United States, because the industry has been diligent in anticipating failures, guarding against them, and adopting designs and procedures that minimize their effects when they occur. 18. Parties opposing these petitions argue that QOR violates both criterion four and criterion six. They argue that QOR requires greater dependence on an incumbent LEC's network than other number portability methods, such as LRN. For example, opponents of QOR claim that QOR requires more signalling and routing steps than LRN before the call is delivered to a customer that has ported a number. They also assert that QOR relies to a greater extent on an incumbent LEC's facilities, because QOR uses both the signalling and trunking networks to reserve a call path to the incumbent LEC's terminating switch to which the NXX code of the called party was originally assigned. Opponents of QOR further assert that the use of QOR affects service quality and network reliability. They contend that: (1) QOR results in service degradation by causing an incremental increase in the post-dial delay for calls ported to a new carrier; (2) QOR impairs network reliability because additional network routing increases the potential for dropped calls and call blocking for ported calls, and that (3) QOR is therefore not "competitively neutral," even when it is used only "within a carrier's network" or between consenting carriers. c. Discussion 19. Criterion Four. Based on our analysis of the record in this proceeding, we now conclude that criterion four should be removed from our list of minimum performance criteria required for number portability, because all interconnected carriers are likely to rely upon each other's networks to some extent to process and route calls in a market in which a long-term number portability method has been deployed. For example, under both LRN and QOR, the competitive LEC may be dependent upon facilities provided by the original service provider for the proper routing of all ported calls, because the original service provider is the entity that launches a query to the number portability database to obtain the location routing number for the dialed number. Furthermore, we find no basis in the record for drawing a principled distinction between permissible and impermissible levels of reliance on the original service provider's network. For these reasons, we find that criterion four -- which requires that any number portability method may not "require telecommunications carriers to rely on databases, other network facilities, or services provided by other telecommunications carriers in order to route calls to the proper termination point" -- is, from a practical perspective, unworkable. Moreover, many of our concerns about reliance on a competitor's network (e.g., the possibility of service degradation and call blocking) are addressed by criterion six. Thus, criterion four does not appear to be necessary in order to implement the statutory definition of number portability. In light of our decision to eliminate criterion four, we conclude that AirTouch's requested clarification of criterion four is moot. 20. Criterion Six. With respect to criterion six, we affirm our conclusion in the First Report & Order that any long-term number portability method must not result in any degradation of service quality or network reliability when customers switch carriers. We further conclude, based on the record in this proceeding, that criterion six prohibits the use of QOR as a long-term number portability method. We agree with the commenters, primarily potential new providers of local exchange services (also referred to as "competitive LECs"), that: (1) QOR results in degradation of service by imposing post-dial delay only on calls ported to new carriers; (2) if network reliability problems were to arise as a result of QOR, those problems would disproportionately affect customers who port their numbers; and (3) QOR should not be permitted on an intranetwork basis, because it is not "competitively neutral." We discuss each of these conclusions in more detail below. (1) Service Degradation 21. After considering petitioners' arguments and concerns, we affirm our conclusion in the First Report & Order that, in accordance with criterion six, a long-term number portability method may not cause customers to experience "a greater dialing delay or call set up time" as compared to when the customer was with the original carrier. Criterion six implements the statutory requirement that consumers be able to retain their numbers "without impairment of quality, reliability, or convenience when switching from one telecommunications carrier to another." 22. At the outset, we agree with AT&T and Time Warner that the time it takes to receive a call is an important factor for many subscribers, particularly businesses that receive and respond to a large number of calls on a daily basis. If the party making a call to a business experiences additional delay because that business has switched carriers, that delay may negatively impact how the business is perceived, which, in turn, could dissuade the business from switching carriers in the first place. Therefore, we clarify that performance criterion six requires that calls to customers who change carriers (not just calls from customers who change carriers) must not take longer to complete merely because the customer has switched local service providers. In order to implement the statutory requirement that consumers should be able to change carriers and retain their original phone number without impairment of quality, reliability, or convenience, we conclude that any post-dial delay imposed by a number portability method should be roughly equivalent for all consumers, whether they are calling to or from a ported or a non-ported number. 23. We further conclude that consumers that switch telecommunications carriers and retain their numbers would experience "impairment of quality" if QOR were used, because the post-dial delay imposed by QOR is not equivalent for all consumers. Under QOR, calls that are placed to ported numbers must undergo a series of signalling and routing steps that result in longer post-dial delay than occurs for calls that are placed to non-ported numbers. (The additional steps in the call flow required by QOR are illustrated in Appendix C.) No party disputes that QOR causes additional post-dial delay. There is disagreement, however, over the appropriate baseline for comparison. Proponents of QOR erroneously focus on the post-dial delay of alternative number portability technologies, comparing the incremental post-dial delay associated with a call to a ported number using LRN with that of a call to a ported number using QOR. That is not the statutory standard. We agree with AT&T and MCI that the proper comparison for incremental post-dial delay is the difference in delay between calls placed to ported numbers and calls placed to non-ported numbers, because that is the delay that occurs "when switching from one telecommunications carrier to another." According to the most conservative estimates, calls to ported numbers from a network that uses QOR would experience an additional post-dial delay of approximately 1.3 seconds as compared to calls placed to non- ported numbers. Because we find that post-dial delay of 1.3 seconds is significant, we conclude that QOR violates the statutory definition of number portability and criterion six. By contrast, under LRN, there is no differential between ported and non-ported calls; for all calls, it takes the same amount of time to query the database for appropriate routing instructions. LRN therefore does not impair service quality when a customer changes carriers. Accordingly, we conclude that LRN is consistent with the statutory definition of number portability and performance criterion six. 24. We also reject petitioners' argument that some degree of added post-dial delay should be acceptable, provided that it is not "perceptible" to the public. First, we agree with AT&T that the studies submitted by petitioners fail to demonstrate that 1.3 seconds of post-dial delay is imperceptible to the public. Second, we agree with those parties that contend that, even if the additional post-dial delay were imperceptible to the caller, QOR could adversely affect competitors, because the incumbent LEC could truthfully advertise the fact that calls to customers that remain on the incumbent LEC's network are completed more quickly than calls to customers that switch to a competitor's network. MCI points out that this could create a marketplace perception that competitive LECs are operating inferior networks, which could harm competition. In response, six incumbent LECs have voluntarily committed not to mention the call set-up time differences between LRN and QOR in their advertising materials. As AT&T and MCI point out, however, the incumbent LECs' voluntary commitment is limited to "advertising materials," and therefore does not preclude them from mentioning call set up in all other aspects of their marketing, such as direct sales and telemarketing, news releases, studies commenced to compare competitors' service performance, and editorials. Furthermore, because only six incumbent LECs signed the letter, we have no basis on which to conclude that all incumbent LECs will refrain from using the differences in call set-up time to influence marketplace perceptions and inhibit competition. Thus, we decline to designate a threshold below which added post-dial delay is permissible. Moreover, given our concerns about these marketplace perceptions, we find U S West's suggestion that the Commission survey consumers to ascertain whether they can perceive the post-dial delay associated with QOR to be unnecessary. (2) Network Reliability 25. QOR. As discussed above, criterion six requires that no long-term number portability method may result in "any degradation of service quality or network reliability when customers switch carriers." We agree with the opponents of QOR that technical concerns raised by QOR are more likely to impact ported numbers adversely than non-ported numbers. For example, QOR requires fewer SS7 links to the number portability database than LRN because of the lower number of queries to support. There is a risk, therefore, that an SS7 network engineered to accommodate a lower traffic level would not be able to handle an unexpected sharp increase in the number of calls to ported numbers. Such increases could occur in response to advertising or promotions by competitive LECs with ported numbers. Difficulties in querying the database may result in call blockage (i.e., lost or incomplete calls) and increased post-dial delay, but only on calls to ported numbers. We also note that the apparent advantage of QOR in requiring fewer queries to the database is offset by the fact that it will require at least two additional signalling messages for each call to a ported number before routing instructions are obtained. This additional load on the signalling network creates the potential for reliability problems for ported calls. We conclude that network reliability concerns posed by QOR violate criterion six and the statutory definition of number portability because, if any network problems arise as a result of QOR, they would disproportionately affect consumers who port their numbers. 26. LRN. As a related matter, proponents of QOR assert that deployment of LRN is more likely to result in network failure than if carriers are permitted to use the QOR enhancement to LRN. Although the proponents of QOR do not frame their arguments in terms of the performance criteria we adopted in the First Report & Order, the thrust of their argument appears to fall within the scope of criterion five, which requires that no number portability method should result in "unreasonable degradation in service quality or network reliability when implemented." 27. Based on the record before us, we conclude that petitioners have not demonstrated that LRN fails to meet criterion five. Although the initial deployment of any new technology may pose some risk to the network, we are not persuaded that deployment of LRN will result in unreasonable degradation of network reliability when deployed under the revised schedule adopted in this First Reconsideration Order. Indeed, petitioners' concerns about LRN's impact on network reliability are mitigated by a number of factors. First, as we noted previously, LRN has been examined extensively by a number of state commissions and industry workshops, and had been selected for deployment by at least six states prior to the adoption of the First Report & Order. Second, we provided in the First Report & Order for a field test of LRN in the Chicago MSA (Chicago trial), which should help to protect against network reliability problems. If technical problems with LRN arise with respect to the Chicago trial, we can take appropriate action at that time. Third, as discussed in more detail in Section III.B.3 below, we are extending the implementation schedule for Phase I to allow carriers additional time to test number portability in a live environment, and to take appropriate steps to safeguard network reliability. Indeed, the Bellcore study submitted by SBC supports our conclusion that additional time for testing, integration, and soaking (limited use of the software in a live environment for a length of time sufficient to find initial defects) will help to reduce the probability of network failure. Fourth, as we clarify below, the Commission's implementation schedule does not require a flashcut implementation on October 1, 1997, for those MSAs in the first phase of the deployment schedule. Rather, number portability may be implemented gradually throughout the initial phase, provided that implementation in the designated markets is completed by the end of that phase. 28. Moreover, petitioners' fears about LRN's impact on the SS7 network are not grounds for abandoning LRN. Because of the pre-deployment procedures we adopt in this order, carriers will know in advance the specific switches in each MSA that require local number portability capabilities. Furthermore, the task of forecasting signalling load requirements should be easier with LRN than QOR, because queries are required for all interoffice intraLATA calls. As a result, carriers should be able to use historic traffic flows to help predict how many of these calls are typically destined to switches where local number portability has been deployed. In contrast, for QOR, signalling loads are dependent upon the percentage of numbers actually ported, which is a figure more difficult to predict in advance. 29. In sum, we conclude that claims that LRN will threaten network reliability are speculative and are mitigated by the added time we have provided for carriers to implement number portability during Phase I and Phase II. We expect the industry to continue to anticipate failures, guard against them, and minimize their effects when they occur, which, as Bellcore points out, has helped to make such failures rare events in the United States in the past. Thus, given all of the safeguards and mitigating factors discussed above, we are persuaded that deployment of LRN will not result in "unreasonable degradation of network reliability." (3) Intranetwork Use of QOR 30. Incumbent LECs ask us to permit them to use QOR on all calls that originate on their network and are placed to numbers that originally were assigned to one of their end offices (i.e., calls "within their own network" or "intranetwork calls"). We conclude that their request is misleading insofar as it implies that only calls to and from their own customers would be affected. In fact, calls that are placed to numbers that have been ported would require a query to the number portability database after the originating switch is notified by the terminating switch in the incumbent LEC's service area that the called number has been ported. We agree with MCI that, as customers subscribe to alternative carriers, the only calls that will remain "within" the incumbent LEC's network will be calls from one of the incumbent LEC's customers to another. As discussed above, however, the call to the ported number would experience increased post-dial delay because of the additional signalling and routing preparations required by QOR. Such disparity in treatment between ported and non-ported numbers violates criterion six and the statutory definition of number portability. 2. Public Interest Considerations a. Overview 31. Petitioners further assert that, regardless of our performance criteria, incumbent LECs should not be prohibited from using QOR as a number portability method, because deployment of QOR serves the public interest. First, they claim that QOR will result in significant cost savings. Second, they claim that permitting incumbent LECs to use QOR will make it easier for them to meet the Commission's implementation schedule. 32. As an initial matter, we disagree with the petitioners' premise that LECs should be permitted to implement QOR regardless of the performance criteria, if the Commission determines that QOR serves the public interest. As stated above, we conclude that QOR violates criterion six, which is required by the statute. Thus, we are not at liberty to apply a public interest analysis that could result in an abrogation of the statutory mandate. Nevertheless, because the parties raised public interest concerns, we address them here in order to establish that our decision to prohibit QOR is not contrary to the public interest. b. Purported Cost Savings Associated with QOR 33. Background. In the First Report & Order, we concluded that there was little evidence on the record to support the claim that deployment of QOR would result in significant cost savings. We found, based on the record, that the competitive benefits of ensuring that calls are not routed through the original carrier's network outweighed any cost savings that QOR might bring in the immediate future. Although Pacific submitted summary figures purporting to indicate that it would save approximately $14.2 million per year if it implemented QOR (assuming that 20 percent of subscribers ported their numbers), we concluded that these purported savings, which represent less than two-tenths of a percent of Pacific's total annual operating revenues, appeared insignificant in relation to the potential economic and non- economic costs to competitors if QOR is used. There was also record evidence that using QOR would only be cost-effective at low levels of ported numbers, depending on the switch type. In addition, we expressed concern that, because carriers using QOR may be required to send QOR signalling to another carrier's switch to determine whether a customer has ported his number, this would require the second carrier to have the ability to recognize and respond to the QOR message, thereby increasing its costs. 34. Pleadings. Petitioners again contend they should be allowed to use QOR because they would achieve significant cost savings. These parties claim that QOR would result in a reduction in the number of database queries, which, in turn, would reduce the costs that must be incurred to complete the infrastructure upgrades necessary to implement QOR compared to those necessary to utilize LRN. Specifically, petitioners allege that QOR would require a carrier to install fewer additional SCP pairs and SS7 signalling links, and to upgrade fewer STPs, than would be the case for LRN. Petitioners also allege that QOR would place less additional load on switch processors, and would thereby delay the need to upgrade those switch processors. 35. Petitioners further argue that QOR would allow carriers to expand the capacity of their SS7 signalling networks more gradually to handle an increased number of queries, as numbers are ported to other carriers. They allege that LRN, in contrast, will require carriers to engineer their networks to accommodate queries on every call from a given NXX once one telephone number has been ported from that NXX. According to these parties, this will require their networks to be grossly "over-engineered" when number portability is initially deployed. Several petitioners note that carriers using QOR would be able to decide on a switch-by-switch basis when it is more cost effective to disable the QOR triggering mechanism and use LRN alone. 36. In response, both AT&T and MCI claim that the LECs have overestimated the costs of LRN and underestimated the costs of QOR, thereby grossly exaggerating the relative cost savings associated with QOR. Their principal objections to the carriers' cost studies are that they: (1) overstate the number of SCP pairs needed to deploy LRN; (2) exaggerate the impact of LRN on switch processor capacity and fail to account for the impact of QOR on switch processor capacity; (3) overstate the number of queries from non-participating carriers, which results in overstated cost estimates, and fail to account for offsetting revenues; (4) fail to estimate the cost of unnecessary call set-up under QOR; (5) fail to account for the additional cost of provisioning QOR in all intermediate and terminating switches, including modifications to Operator Support Systems (OSS); and (6) exaggerate the speed of number portability deployment outside of the top 100 MSAs. In addition, MCI points out that GTE, SBC, NYNEX, and Bell Atlantic failed to specify the assumptions underlying their cost studies, and Pacific submitted its cost study on a confidential basis, making it difficult to undertake a detailed analysis of those cost studies. 37. Furthermore, there is a dispute in the record over the point at which it becomes more cost effective to use LRN rather than QOR. As we noted in the First Report & Order, AT&T contends that it is more cost effective to deploy LRN in Lucent switches when 12 percent of the customers served by such a switch have ported their numbers, and to deploy LRN in Siemens switches when 23 percent of the customers served by such a switch have ported their numbers. In this phase of the proceeding, BellSouth asserts that the crossover point occurs when 68 percent of its customers have ported their numbers. NYNEX contends that the appropriate transition is a function of its SCP costs, its signalling costs, and its switch costs, and suggests that "this point may occur when 50 percent of numbers have ported." In contrast, several interexchange carriers have argued that the crossover point for Pacific occurs when 20 percent of numbers have ported. 38. Discussion. At the outset, it is important to clarify the nature of the asserted "cost savings" associated with QOR. As most carriers recognize, LRN is the more economical way to provide long term number portability once ported numbers for a given switch reach a certain level, although the point at which it becomes more cost-effective to use LRN rather than QOR remains in dispute. From an economic perspective, the question is whether the present discounted value of the cost of initially deploying LRN is less than the present discounted value of the cost of deploying QOR initially and LRN at some later date. Proponents of QOR contend that the use of the QOR enhancement to LRN would result in real cost savings, not just a short-term deferral of expenses, because the number of ported calls in some areas will never reach the level where it is more cost effective to disable QOR and complete the build-out necessary to support LRN. We conclude, however, that the statutory scheme that Congress has put in place should, over time, result in vigorous facilities-based competition in most areas, and therefore LRN will be the most economical long-term solution. Thus, deploying QOR would most likely result in short-term cost savings, not overall cost savings. In fact, at least one incumbent LEC, Ameritech, has already decided that it is beneficial to deploy LRN from the outset, rather than converting from QOR to LRN at some later date. Even if facilities-based competition does not develop in the immediate future, however, we conclude that the harm that QOR imposes on competitors (as discussed in Section III.A.1 above) outweighs the benefit of allowing incumbent LECs to defer the cost of implementing a superior long-term number portability solution. 39. Moreover, we are not convinced that the incumbent LEC's estimates of the short- term savings associated with QOR are reliable. We are particularly concerned by the fact that the cost savings estimates submitted by incumbent LECs have varied significantly over the course of this proceeding. In some cases, estimates from the same carrier have changed by 100 percent or more. Further, the changed estimates have not moved in the same direction; some carriers' estimates of the cost savings increased drastically and other carriers' estimates decreased equally drastically. While we recognize that carriers have worked over time to refine their projections, the wide variation in the estimates submitted by individual carriers at different points in this proceeding raises questions about the reliability of these estimates. Furthermore, the fact that some carriers have not explained the basis for the assumptions underlying their estimates precludes us from conducting an independent evaluation of the reasonableness and reliability of their projected cost savings and, consequently, limits the weight we can reasonably assign to those estimates. 40. In addition, MCI alleges that the cost savings that would be realized by permitting the deployment of QOR are far less than the estimated $54 million to $136.3 million in annual savings alleged by individual incumbent LECs. The following chart shows the difference between estimated savings submitted by the petitioners and estimated savings calculated by MCI: Carrier Estimated Savings as Reported by the Carrier (Millions) Estimated Savings as Reported by MCI (Millions) Bell Atlantic $68 $15 BellSouth $102 N/A GTE $136 $28 NYNEX $54 N/A Pacific $130 -$12 SBC $84 $19 U S West $50-$75 N/A 41. MCI's calculation of the asserted cost savings associated with QOR challenges a key assumption underlying the incumbent LECs' estimates. Specifically, MCI claims that the LECs substantially underestimate the number of transactions (i.e., queries) per second (tps) that an SCP pair can perform and, consequently, their estimate of the number of SCP pairs that must be deployed to provide LRN is overstated. AT&T also alleges that the incumbent LECs' savings estimates do not take into account offsetting increases in additional switching facilities costs that would be required for QOR. MCI and AT&T further contend that the incumbent LECs' estimates of the relative costs of deploying LRN and QOR must be adjusted downward to account for revenues that they will receive to perform database queries at the request of rural and other LECs that do not have the capability to perform such queries themselves. Although incumbent LECs would obtain such revenues with both the LRN and QOR methodologies, the revenue stream is likely to be significantly greater with LRN because the number of database queries is likely to be much greater. Indeed, Pacific, a proponent of QOR, acknowledges that its estimate of the cost savings associated with QOR would be reduced by as much as $18 million if such revenues were included in the estimate. In view of the significant changes in the estimates of the cost savings associated with QOR submitted by individual incumbent LECs over the past months, a lack of data explaining many of the assumptions underlying their estimates, and the questions raised by MCI and AT&T with respect to specific aspects of the estimates, we find, on balance, that the incumbent LECs have not substantiated their claim that deployment of QOR will produce significant cost savings. 42. Moreover, a recent submission by Illuminet, a provider of SS7, database, and other services to independent LECs and other entities, casts doubt on the reasonableness of one of the most basic assumptions underlying the incumbent LECs' estimates of the relative costs of QOR and LRN. Incumbent LEC estimates assume that the LEC number portability architecture will be deployed through a network of SCPs, and that a major cost driver of LRN is the number of SCPs needed to handle increased traffic volumes. On the other hand, Illuminet advocates using an STP-based architecture, in which call routing information from the regional database is transferred to a carrier's STP instead of an SCP, and the SCP is not involved in processing the number portability query. Illuminet asserts that STPs are designed specifically to do ten-digit translations such as LRN query processing and can process number portability queries at a much faster rate than SCPs. In contrast, SCPs are designed to support multiple call processing applications and process significantly fewer queries per second. Carriers using an STP-based architecture, therefore, would need to purchase and install a relatively smaller number of STPs instead of the larger number of SCPs alleged by the LECs, and would not need to purchase and install additional SS7 links between the SCPs and STPs. Thus, according to Illuminet, use of an STP-based architecture would reduce dramatically the cost of LRN. In response, Pacific acknowledges that a combined STP-SCP approach may reduce some costs, but that expenses related to upgrading switch processors, links, and existing STPs will still be substantial. Although we acknowledge that carriers deploying LRN will incur costs other than those associated with SCPs, we agree with Illuminet that an STP-based approach should reduce the relative cost differential between LRN and QOR. 43. In addition, as we discuss more fully in Section III.B.2 below, we are modifying our implementation schedule to require LECs to deploy number portability only in those switches requested by a competitive LEC within a given MSA on the implementation schedule, rather than in every switch in that MSA. As a result, fewer switches should require upgrading in each phase of the deployment schedule, with a corresponding reduction in the cost of implementation for all carriers. Moreover, if number portability capabilities are not deployed in all switches, then there will be fewer switches generating database queries, and thus fewer SCPs and signalling links will be needed than the LECs have estimated. Sprint, for instance, has estimated that it would save approximately 25 percent of its number portability budget of $60 million for 1997 if it were not required to deploy number portability in the smaller exchanges within the MSAs on its deployment schedule. While it is impossible at this time to quantify the precise magnitude of this effect nationwide because we do not know in how many switches competitive LECs will request number portability, this modification to our number portability requirements should lessen somewhat whatever actual cost differences may exist between LRN and QOR. c. Impact of QOR on the Implementation Schedule 44. Pleadings. Bell Atlantic and Pacific claim that allowing the use of QOR would make it easier for carriers to meet the Commission's implementation schedule, because they would not need to deploy as many databases and as extensive a signalling infrastructure as would be needed under LRN. MCI disputes the claim that QOR would help carriers meet the implementation schedule. MCI argues that QOR has never been fully examined and specified by the industry in any state task force. MCI further argues that the proponents of QOR have not established that it would be technically infeasible to deploy LRN fully under the existing implementation schedule. AT&T claims that, even after QOR software becomes available, additional time would be necessary to complete the installation, testing, and training necessary actually to implement QOR. 45. Discussion. We are not persuaded by Bell Atlantic and Pacific that number portability would be deployed more rapidly if incumbent LECs are permitted to use QOR. We find speculative petitioners' arguments that problems will arise in LRN implementation, and that the Commission therefore should allow the use of QOR. We agree with AT&T that no party has demonstrated that schedules for completing installation, testing, training, and other tasks necessary to implement QOR could be developed and coordinated with the schedules for completing tasks necessary to implement LRN. Furthermore, no party has alleged that a field trial of QOR could be performed earlier than or even contemporaneously with the Chicago trial for LRN. To the contrary, as discussed in the next subsection, we have reason to believe that allowing the use of QOR would delay the Chicago trial and the implementation schedule. d. Impact on the States 46. As discussed in Section II.C above, seven state commissions have specifically ordered implementation of LRN. These and a number of other states have invested considerable time, effort, and resources in developing LRN implementation plans and technical standards. Illinois is proceeding with the field trial of LRN in the Chicago MSA. Illinois, Georgia, California, Maryland, Colorado, New York, and Texas have undertaken significant efforts to form LLCs to develop and issue RFPs to construct and maintain a number portability database, to plan for expanding these state databases into regional databases, and to prepare in each state for database testing, in order to be ready to support number portability deployment in accordance with the schedule set forth in the First Report & Order. These states have been in the forefront of opening markets to local competition, and we applaud and support their ongoing commitment to take actions necessary to make local number portability a reality in their jurisdictions. If we were to reverse our earlier finding that QOR is not acceptable as a long-term number portability method, these state activities could be greatly disrupted. Much of the testing and development of technical standards already done for implementation of LRN would have to be redone in order to accommodate a scenario in which both QOR and LRN may be in use in a given state. Moreover, the states that have been leaders in number portability implementation would likely be forced to reopen their state number portability proceedings to reconsider QOR, which could delay implementation for months while those proceedings are pending. e. Conclusion 47. Congress recognized that there are costs associated with the implementation of local number portability. Although carriers may realize some short-term cost savings if permitted to use QOR instead of LRN, the exact amount of savings from utilizing QOR is unclear. Even if the cost savings figures submitted by the LECs were correct, we believe that the benefits to consumers of such savings do not outweigh the harm that QOR would impose on competitive LECs, the cost of disrupting state efforts to implement LRN, or any delay in implementation that might result from such disruption. Thus, we conclude that permitting carriers to deploy QOR as a long-term number portability method does not serve the public interest. B. Implementation Schedule for Wireline Carriers 1. Background 48. In the First Report & Order, the Commission required local exchange carriers operating in the 100 largest MSAs to offer long-term service provider portability, according to a phased deployment schedule commencing on October 1, 1997, and concluding on December 31, 1998. The Commission noted that, in establishing the deployment schedule, it relied upon representations of switch vendors regarding the dates by which the necessary switching software will be generally available for deployment. In particular, vendors estimated that they could begin to make software for at least one long-term number portability method generally available for deployment by carriers around mid-1997. In addition, a carrier may file a specific request for number portability beginning January 1, 1999, for areas outside the 100 largest MSAs, and each LEC must make long-term number portability available in that MSA within six months after the specific request. The Commission also directed the carriers that are members of the Illinois Commerce Commission Local Number Portability Workshop (ICC Workshop) to conduct in the Chicago MSA, concluding no later than August 31, 1997, a field test of LRN or another technically feasible long-term number portability method that comports with our performance criteria. The Commission noted that Section 251(f)(2) of the Act permits a LEC with fewer than two percent of the country's total installed subscriber lines to petition a state commission for suspension or modification of the interconnection requirements of Sections 251(b) and (c). 49. The Commission delegated to the Chief, Common Carrier Bureau, the authority to monitor the progress of LECs implementing number portability, and to direct carriers to take any actions necessary to ensure compliance with its deployment schedule. The Commission also delegated to the Chief, Common Carrier Bureau, the authority to waive or stay any of the dates in the implementation schedule, for a period not to exceed nine months (i.e., no later than September 30, 1999, for the MSAs in Phase V of the deployment schedule), as is necessary to ensure the efficient development of number portability. In the event a carrier is unable to meet our deadlines for implementing a long-term number portability method, it may file with the Commission, at least 60 days in advance of the implementation deadline, a petition to extend the time by which implementation of long-term number portability in its network will be completed. The Commission emphasized, however, that carriers are expected to meet the prescribed deadlines, and a carrier seeking relief must present extraordinary circumstances beyond its control in order to obtain an extension of time. The Commission required a carrier seeking such relief to demonstrate through substantial, credible evidence the basis for its contention that it is unable to comply with our deployment schedule. 2. Deployment Only in Requested Switches 50. Pleadings. Ameritech urges the Commission to limit initial deployment of number portability in an MSA to exchanges where bona fide demand exists. Ameritech argues that excluding exchanges in rural and less densely populated suburban areas of an MSA, where competition is not likely to develop immediately, will significantly reduce costs and the demand on carriers' limited technical personnel and resources, and simplify deployment and testing. 51. Ameritech suggests delegating to state commissions the task of supervising the selection of exchanges where demand exists, and cites as a model the procedure used by the ICC Workshop in the Chicago MSA, prior to the release of the First Report & Order, under which each competing LEC submitted to the ICC staff a list of the exchanges in which the LEC sought number portability as a part of the initial deployment. The sole criterion for designation of an exchange was that the carrier anticipated needing immediately the capability to port numbers from that exchange. The ICC staff then aggregated the lists and released one consolidated list to serve as the master deployment plan for the Chicago MSA. According to Ameritech, this procedure excluded from deployment 103 out of 206 exchanges in the Chicago MSA, which serve primarily rural and less densely populated suburban areas and include many areas served by small independent telephone companies and by switches with older technology. 52. According to Ameritech, the incumbent LECs then categorized the unrequested exchanges according to the type of switch serving that exchange, and planned to convert each exchange upon a bona fide request according to the following time frames: (1) remote switches supported by a host switch equipped for portability ("Equipped Remote Switches") within 30 days; (2) switches that require software but not hardware changes to provide portability ("Hardware Capable Switches") within 60 days; (3) switches that require hardware changes to provide portability ("Capable Switches Requiring Hardware") within 180 days; and (4) switches not capable of portability that must be replaced ("Non-Capable Switches") (no agreement was reached on a time frame). Ameritech explains that, because unconverted offices would be identified prior to the initial deployment in the MSA, new LECs could request additional offices at any time, and thus notify the incumbent LECs to begin planning for conversion of those offices as soon as possible after the initial deployment in the MSA. Therefore, claims Ameritech, additional conversion could, in most cases, occur within 30-60 days after the initial deployment in the MSA. 53. BellSouth also seeks clarification that portability need not be deployed in every switch within an MSA. BellSouth reports that industry participants in the Georgia number portability workshop conducted an exercise similar to that of the ICC Workshop prior to release of the First Report & Order, in which the competing carriers selected 21 offices in the Atlanta MSA for initial implementation in late 1997. Number portability task forces in Indiana, Michigan, and Ohio, following the work of the ICC Workshop, have also established procedures under which each competing LEC must submit a list of the exchanges in which it desires number portability as a part of the initial deployment. In Maryland, each carrier submitted to the Maryland commission staff a ranking of the fifty end offices in the Baltimore and Washington, DC LATAs for which it most desired portability, and the five end offices in the Salisbury and Hagerstown LATAs for which it most desired portability. The Maryland commission staff then prepared a consolidated ranking that became the implementation roll-out schedule for Maryland. There were 25 end offices in the Baltimore and Washington, DC LATAs (out of 92 total end offices), and seven end offices in the Salisbury and Hagerstown LATAs (out of 13 total end offices), that no carrier included in its list of end offices for which it requested number portability. 54. USTA proposes that competing carriers be required to specify, in a request to a LEC, those switches for which they wish the ability to port numbers. USTA argues further that, if a carrier does not receive a request for portability in an end office by April 1, 1997, then the carrier should be able to obtain from the Commission a waiver of the deployment schedule until the LEC receives a request. Upon receiving such a request, the LEC would have nine months, or a period of time specified by the Chief of the Common Carrier Bureau, to deploy portability. USTA also proposes that state commissions and/or state number portability workshops be empowered generally to alter the timing of deployment for particular switches within their state boundaries. USTA and several other rural LEC representatives argue that, without a procedure to limit deployment to switches for which a competitor has expressed interest, many rural and small LECs will have to upgrade their networks at significant expense even though no competitors plan to enter their markets and use number portability. 55. GTE urges us to establish a "limited waiver" process for exempting smaller offices in the 100 largest MSAs from the deployment deadlines where competitive entry in that area will not be immediate, and implementation would require significant network upgrades. A LEC wishing to take advantage of GTE's proposed procedure would first determine whether any prospective entrant "expresses an immediate interest in entry" in the relevant area, and whether those prospective entrants, or the state commission, have any objection to waiving the schedule for that area. If the prospective entrants and state commission do not object, then the LEC would present the Commission with a petition for waiver "with the expectation that it will be granted." Afterward, the LEC would not have to implement portability until six months after a request from a competing carrier, assuming the switch already has SS7 and AIN capabilities. According to GTE, its proposal would enable LECs to devote their resources to upgrading offices in the more densely populated and competitive areas, and would recognize that portability requires expensive upgrades in many smaller offices. 56. NEXTLINK presents a "swapping" proposal, under which an incumbent LEC seeking a waiver for a switch within the 100 largest MSAs instead would deploy switches outside the 100 largest MSAs which a competitor requests. NEXTLINK cautions, however, that we should rely on state commissions to determine the extent of competition in markets in their states, but not, as suggested by USTA, to determine whether waivers should be granted. 57. AT&T does not oppose proposals to limit deployment of number portability to those switches for which a carrier requests deployment. Sprint supports Ameritech's proposal, which does not entail LECs requesting waivers for unrequested offices. Sprint predicts that as many as 127 out of a total of 360 of its central offices will not face immediate facilities-based competition and will be relieved from initial deployment under a procedure whereby carriers identify the switches for which they desire portability, at an estimated savings of over $15 million in 1997 alone (approximately 25 percent of Sprint's total number portability budget). Sprint emphasizes that we should determine a specific time frame within which the carrier must deploy portability once a bona fide request for portability is received, absent some other extenuating (and fully documented) circumstances. Sprint asserts that the state public utilities commissions in Florida, Ohio, Indiana, and Illinois, have established procedures by which carriers request deployment in specific exchanges. Time Warner supports allowing carriers to apply for waivers of the deployment schedule for the 100 largest MSAs for end offices serving areas that competitors do not plan to enter initially. 58. MCI, in contrast, opposes relaxing the mandate of MSA-wide deployment. According to MCI, forcing competitive LECs to defend the need for MSA-wide portability and to justify deployment in each end office would create an environment of uncertainty for competitive LECs. MCI claims that, if competitive LECs must request deployment each time a new customer requesting service is located in an end office that was not deployed according to the original deployment schedule, any incentive and ability to market their services widely will be impaired. According to MCI, once portability is introduced in an area, the incremental cost and resources needed to add additional end offices are relatively minor because most costs, i.e., SCP hardware and signalling links, OSS modifications, and shared regional database costs, will have already been incurred. 59. Discussion. We agree with the majority of the parties commenting on this issue that it is reasonable to focus initial efforts in implementing number portability in areas where competing carriers plan to enter. This approach will permit LECs to target their resources where number portability is needed and avoid expenditures in areas within an MSA in which competitors are not currently interested. We further agree that such a procedure will foster efficient deployment, network planning, and testing, reduce costs, and lessen demands on software vendors. Moreover, we believe that limiting deployment to switches in which a competitor expresses interest in number portability will address the concerns of smaller and rural LECs with end offices within the 100 largest MSAs that they may have to upgrade their networks at significant expense even if no competitors desire portability. Limiting deployment to switches in which a competitor expresses interest in deployment will be consistent to a large extent with procedures suggested by Ameritech and BellSouth and already considered by several state commissions, as well as our past practice in implementing conversion to equal access for independent telephone companies. 60. We therefore conclude that LECs need only provide number portability within the 100 largest MSAs in switches for which another carrier has made a specific request for the provision of portability. We leave it to the industry and to state commissions to determine the most efficient procedure for identifying those switches in which carriers have expressed interest and which will be deployed with number portability according to the original deployment schedule for the 100 largest MSAs. We find, however, that any procedure to identify and request switches for deployment of number portability must comply with certain minimum criteria to ensure that minimal burden is imposed upon carriers requesting deployment in particular switches, and that carriers that receive requests for deployment in their switches have adequate time to fulfill the requests. As explained below, we require that: (1) any wireline carrier that is certified, or has applied for certification, to provide local exchange service in the relevant state, or any licensed CMRS provider, must be allowed to make a request for deployment; (2) requests for deployment must be submitted at least nine months before the deadline in the Commission's deployment schedule for that MSA; (3) carriers must make available lists of their switches for which deployment has and has not been requested; and (4) additional switches must be deployed upon request within the time frames described below. 61. First, any wireline carrier that is certified (or has applied for certification) to provide local exchange service in a state, or any licensed CMRS provider, must be given a reasonable opportunity to make a specific request for deployment of number portability in any particular switch located in the MSAs in that state designated in the First Report & Order. According to the Act, any carrier that desires number portability from a LEC must be able to obtain portability, in accordance with the requirements established by the Commission. A state commission, however, may review whether the requests made by a carrier are unreasonable, given the state commission's knowledge of that carrier's plans to enter the state. Based on the limited information available to us at this time, the states that are reviewing seemingly unreasonable requests appear to be acting in good faith to accommodate carriers' interests in number portability capabilities. If we receive evidence in the future that states are unreasonably limiting deployment, then we can revisit this issue at that time. 62. Second, a carrier must make its specific requests for deployment of number portability in particular switches at least nine months before the deadline for completion of implementation of number portability in that MSA. We conclude that this deadline will enable a LEC to plan ahead for the deployment of number portability in multiple switches in a given MSA. We encourage carriers to make such requests earlier than the nine-month deadline to give the LEC that operates the switch in which portability is requested more time to implement number portability capabilities. In addition, carriers may agree among themselves, or state commissions may require carriers, to comply with a deadline for submitting requests that is more than nine months prior to the implementation deadline. 63. We encourage carriers, before requests for deployment are submitted, to seek to reach a consensus on the particular switches that initially will be deployed with number portability. We note, moreover, that the state commission may decide, or carriers affected in the state may agree, that it would be preferable for the state commission to aggregate the requests to produce a master list of requested switches. In addition, we conclude that carriers may negotiate private agreements specifying that a carrier will not request that certain switches be deployed according to the Commission's schedule if the LEC from which deployment is requested agrees to deploy other number portability-capable switches, either inside or outside the 100 largest MSAs, at an earlier date than the deadlines in the Commission's schedule. 64. Third, after carriers have submitted their requests, a carrier must make readily available upon request to any interested parties a list of its switches for which number portability has been requested and a list of its switches for which number portability has not been requested. We find that simplifying the task of identifying the switches in each MSA in which number portability is initially scheduled to be deployed is consistent with our policy of facilitating the deployment of number portability in areas where new competitors plan to enter. 65. Fourth, carriers must be able to request at any time that number portability be deployed in additional switches. LECs must provide portability in these additional switches upon request, after the deployment deadline mandated by the Commission's schedule for that MSA, within the time frames that we adopt here, unless requesting carriers specify a later date. Although carriers may make specific requests for deployment in additional switches in a particular MSA at any time, the time frames set forth below will commence after the deadline for deployment in that particular MSA in our implementation schedule. We agree with Sprint and Time Warner that specific time frames within which number portability must be deployed in all switches that were not initially requested are necessary to ensure that competitive LECs can be certain that portability will be available in areas in which they plan to compete and can formulate their business plans accordingly. Absent this certainty, competing carriers would have an incentive to request more switches during the initial request process, including those serving markets which they do not plan to enter in the near future, in order to ensure deployment of portability in any switch in which they might ever want portability. We find, therefore, that establishing specific time frames for deployment in all additional switches will benefit competitive LECs by ensuring that portability will be available to them at a designated future time, and will benefit incumbent LECs by reducing their initial deployment burdens. 66. We find that the time frames developed by the carriers participating in the ICC Workshop generally successfully balance the needs of competitive LECs for certainty of deployment and the burdens faced by incumbent LECs in deploying number portability in additional switches that require different levels of upgrades. We therefore adopt, with slight modification, the time frames developed by the ICC Workshop for the conversion of additional exchanges: (1) Equipped Remote Switches within 30 days; (2) Hardware Capable Switches within 60 days; (3) Capable Switches Requiring Hardware within 180 days; and (4) Non-Capable Switches within 180 days. For example, if carriers request deployment in a certain number of switches in the Pittsburgh, PA MSA nine months before that MSA's Phase III deadline of June 30, 1998 (i.e., they make requests by September 30, 1998), and a carrier requests on April 1, 1998, deployment in an additional Equipped Remote Switch in Pittsburgh, then the additional switch must be equipped with number portability capability on or before July 30, 1998 (i.e., 30 days after June 30, 1998). We note that the ICC Workshop developed the time frames for the first three switch categories, but did not reach agreement on a time frame for converting a Non- Capable Switch. Since we find, as discussed above, that specific time frames for deployment of all additional switches are necessary, we find that it is reasonable to allow no more time for deployment of any switches within the 100 largest MSAs than is allowed for deployment of switches outside the 100 largest MSAs. Deployment in additional switches will be less burdensome for carriers with networks within the 100 largest MSAs that have already made network-wide upgrades, e.g., SCP hardware and OSS modifications, to support number portability in the initially requested switches. 67. Carriers seeking relief from these deadlines may file a petition for waiver under the procedures set forth in the First Report & Order. We note that the deadlines for switches in categories (1) and (2) are shorter than switches in categories (3) and (4) because the former require less extensive upgrades. We realize that the shorter deadlines for switches in categories (1) and (2) do not allow time for carriers to file a petition for waiver under the procedure established in the First Report & Order on the grounds of extraordinary circumstances that prevent it from complying with the Commission's deployment requirements. We therefore will suspend the deadlines for switches in categories (1) and (2) during the period that the Commission is considering a carrier's petition for waiver. 68. We agree with MCI that, after portability has been introduced in an MSA, the incremental cost and resources needed to add additional end offices are relatively minor because most costs, e.g., SCP hardware and signalling links, OSS modifications, and shared regional database costs, will have already been incurred. Number portability, consequently, can be deployed more quickly in the switches for which number portability is requested after the initial deployment of number portability. We therefore decline to adopt suggestions by USTA and GTE to allow a longer time after receipt of a request for deployment of number portability capability in switches not in the initial deployment. 69. We emphasize that a carrier operating a non-portability-capable switch must still properly route calls originated by customers served by that switch to ported numbers. When the switch operated by the carrier designated to perform the number portability database query is non-portability-capable, that carrier could either send it to a portability-capable switch operated by that carrier to do the database query, or enter into an arrangement with another carrier to do the query. 70. We conclude that permitting carriers to specify those switches within the 100 largest MSAs in which they desire portability is more workable than the procedures proposed by some petitioners that would require incumbent LECs to file waiver requests for specific switches for which the incumbent LECs believe that no competitor is interested. A waiver procedure would create a period of uncertainty for both the incumbent LEC and the competitive LEC as to whether portability would actually be deployed in that switch. Moreover, a waiver procedure would burden the incumbent LEC with preparing and filing the petition for waiver, require that we review the petition, and potentially burden the state commission with determining whether there is actual competitive interest in the switch. In addition, these proposals by petitioners appear to assume generally that no competitive LEC would oppose the waiver petition; if this is not the case, then a waiver procedure would burden competing carriers with challenging the waiver. A waiver procedure would also burden both competing carriers and consumers by hampering competitive entry into the market while waiting for a determination by the Commission or a state commission. 71. We believe that the criteria set forth above adequately address MCI's concern that requesting carriers would bear an unnecessary burden of justifying deployment in each end office and endure uncertainty as to deployment. The only burden on requesting carriers is to identify and request their preferred switches. In addition, carriers have a time frame for deployment of the initially unrequested switches within the 100 largest MSAs. Competitive LECs can thus market their services as widely as they desire with assurance that number portability will be available in the areas where, and at the times when, they desire to compete. As an additional safeguard against anticompetitive abuses of the procedures to identify and request those switches for which a carrier desires deployment of number portability, we delegate authority to the Chief, Common Carrier Bureau, to take action to address any problems that arise over any specific procedures. 3. Extension of Implementation Schedule 72. Pleadings. Several BOCs and GTE argue that the current schedule for implementation by wireline carriers allows too little time for implementing a technology that requires such extensive network-wide modifications. These petitioners argue that the present schedule could jeopardize network reliability because it does not allow sufficient time to complete numerous tasks, many of which, they allege, are beyond their control, including: (1) review and incorporation of the results from the Chicago trial, and resolution of critical carrier- specific operational issues that the Chicago trial will not address; (2) development and testing of number portability-specific and "generic" software upgrades; (3) development and testing of infrastructure modifications and additions to support number portability capabilities; (4) modification of operational support systems (OSS); (5) modification of vendor software if state commissions dictate inconsistent rate centers for identifying and billing calls; and (6) establishment of regional databases and associated technical standards by the NANC. On February 19, 1997, SBC submitted a study it commissioned from Bellcore that purportedly demonstrates that the deployment schedule set forth in the First Report & Order for Phase I would threaten network reliability in Houston. 73. In their petitions and comments, some of the incumbent LECs recommend specific ways to relax the deployment schedule for wireline carriers. U S West suggests extending the deadline for each phase by three months, claiming this would give carriers not participating in the Chicago trial the necessary time to study the results of the trial and conduct tests within their own networks. BellSouth, CBT, and GTE recommend that the deadlines for completing implementation of Phases I and II each be extended from 90 to 180 days. Under this plan, new Phase I would extend from October 1997 through March 1998; new Phase II would extend from January 1998 through June 1998, and the remaining phases would remain the same (e.g., Phase III would still extend from April 1998 through June 1998). BellSouth claims that, because its plan would extend only the deadlines for completing implementation of long- term number portability for Phases I and II, LECs would still start implementation of all phases, and complete deployment in Phases III through V, according to our original schedule. In addition, BellSouth seeks clarification that Phase I implementation may begin at any time during Phase I (i.e., from October 1997 through March 1998, under its proposed schedule). GTE urges us to clarify that LECs will be entitled to a waiver of the deployment deadlines if they cannot meet the deployment schedule for reasons "outside the control of the LECs." USTA proposes allowing each state commission and/or its workshop to evaluate evidence of local competition in areas within that state, and either accelerate or decelerate the deployment schedule in those areas, as long as the "overall burden" on carriers implementing number portability is not increased. 74. NYNEX urges us to expedite the Chicago trial, or, in the alternative, to select other areas to hold field trials. NYNEX also urges us to encourage states to be flexible in opting out of the regional database or choosing to construct joint databases, and to work with less active neighboring states to establish regional databases. NYNEX also suggests that, during Phase I of the schedule (fourth quarter of 1997), we allow LECs to deploy long-term number portability in smaller MSAs as test beds, instead of requiring deployment in the largest MSAs. 75. The prospective entrants generally oppose any delay in the implementation schedule for wireline carriers. AT&T responds that the Commission's schedule is justified by specific showings in the record that an industry Service Management System (SMS) could be deployed, upgrades of carrier networks could be performed, and operational issues could be addressed in time for completion of widespread deployment (i.e., in 84 MSAs) of long-term number portability by the third quarter of 1998. MCI argues that our schedule is reasonably based on the schedules that several states had already established which ordered deployment to begin in the third or fourth quarter of 1997. In ex parte filings, AT&T and MCI both argue that the late-filed Bellcore study does not provide an adequate basis for extending the implementation schedule, and that the study is "fatally flawed." 76. MCI argues that the safeguards in the First Report & Order -- monitoring of implementation by the Chief of the Common Carrier Bureau, the Chicago trial, and the waiver procedure for extending the deployment deadlines if necessary -- will be adequate to avoid alleged network reliability risks and technical problems. MCI also urges us to instruct the LECs that they will not receive a waiver of the schedule if they introduce new services or technologies that are incompatible with LRN, experience implementation problems as a result, and then claim more time is needed to modify LRN and resolve the problems caused by the introduction of incompatible services or technologies. AT&T, ICG, NEXTLINK, Sprint, and TRA also argue that a procedure for relief already exists if carriers show that they cannot meet the implementation schedule. Therefore, argues NEXTLINK, requests for delay of the implementation schedule are premature and fail to demonstrate the "extraordinary circumstances" required by the First Report & Order. ALTS argues that the incumbent LECs challenging the technical aspects of the schedule should instead first try to resolve their claims with the involved carriers and vendors, and then seek Commission intervention with respect to any remaining issues. Time Warner argues that, given the incumbent LECs' strong incentive to delay competition, we should closely scrutinize claims of infeasibility in case-by-case waiver requests, and deny a waiver if another carrier facing similar technical challenges (e.g., upgrading similar generic software on similar switches) has met the deployment deadlines. 77. ICG claims BellSouth's request to extend the number portability implementation schedule for Phases I and II is unjustified and would slow implementation in the later phases as well. ICG suggests that if we do grant BellSouth's request, however, then the implementation dates for Phases III, IV, and V should not be changed. ICG and NEXTLINK oppose U S West's request that carriers not participating in the Chicago trial receive an extension of the implementation schedule, arguing that: the four month period between the completion of the Chicago trial and the completion of implementation in Phase I is ample time for carriers to review the results of the Chicago trial, and carriers can schedule their own trials if they want more time; LECs need not wait for the outcome of the Chicago trial before testing and modifying their own networks, as the trial's results will be available as it progresses; and U S West participates in state and industry fora where implementation and inter-carrier OSS impacts of number portability have been extensively analyzed. Similarly, MCI claims that the results of the Chicago trial will be applicable to all networks, because all carriers use switches from the same few vendors and have similar network designs. Finally, MCI claims that USTA's proposal to allow states to alter the Commission's deployment schedule would let incumbent LECs influence the states to delay the schedule and thus "cripple" deployment of long-term number portability in every MSA. 78. Discussion. We grant, with some modifications, the requests by BellSouth and other parties to extend the deadlines for completion of deployment of long-term number portability for Phases I and II, as set forth in Appendix E of this First Reconsideration Order. On reconsideration, we extend the end date for Phase I by three months. Thus, deployment in Phase I will now take place from October 1, 1997, through March 31, 1998. We take this action because we are now persuaded that initial implementation of this new number portability technology is likely to require more time than subsequent deployment once the technology has been thoroughly tested and used in a live environment. For example, initial implementation of this new technology is likely to involve more extensive testing, and may require extra time to resolve any problems that may arise during the testing. It therefore is appropriate that Phase I be longer than subsequent phases in the schedule to allow carriers to take appropriate steps to safeguard network reliability. 79. We also note that the participants in the Chicago trial have recently informed us that the completion date of the Chicago trial, previously scheduled for August 31, 1997, has been postponed by approximately one month until September 26, 1997. While the Chicago trial participants have committed to providing the Commission with weekly updates on trial progress, the full report on the Chicago trial that participants had planned to file September 30, 1997, is now scheduled to be filed October 17, 1997. Consistent with this notification by the Chicago trial participants, we hereby extend our deadline for carriers that are members of the ICC Workshop to conduct a field test of any technically feasible long-term database method for number portability in the Chicago, Illinois, MSA and to report the results of that trial. While we understand that participants in the Chicago trial are prepared to commence implementation in Chicago immediately upon conclusion of the trial and still expect to meet the original December 31, 1997, deadline, we recognize that carriers operating in other MSAs may require additional time to interpret the results of the Chicago trial in light of their individual network configurations. Finally, we find some merit in CBT's argument that an extra 90 days for initial implementation may permit small and mid-size LECs to reduce their testing costs by allowing time for larger LECs to test and resolve the problems of new technology. Given all the factors listed above, we conclude that a three-month extension of the time period for initial deployment in Phase I markets appropriately safeguards network reliability, and therefore is warranted. 80. We also extend the end date for Phase II by 45 days. Thus, deployment in Phase II will now take place from January 1, 1998, through May 15, 1998. We extend Phase II to alleviate potential problems that may arise if deployment in markets in Phase I and II must be completed on the same date. Requiring that implementation be completed in a greater number of markets by a specific deadline may make that deadline more difficult to meet (e.g., by straining vendor resources to perform software upgrades in any given period of time). For the same reason, we decline to extend Phase II by 90 days as requested by BellSouth, as such an extension would establish the same deadline for completion of deployment for Phases II and III. We conclude that the modest adjustment of the deadline for Phase II adopted in this First Reconsideration Order will more effectively stagger the deadlines for deployment in different markets than BellSouth's proposal. 81. We clarify, per BellSouth's request, that implementation of number portability for a phase may begin at any time during that phase, provided that implementation in the designated markets is completed by the end of that phase. Contrary to the allegations of Pacific and other parties, number portability thus need not be introduced "on virtually the same day" in the seven of the largest MSAs, especially because it may now be phased into the first markets more gradually over six months, instead of three. 82. We strongly advise carriers to begin implementation early in each phase, however, as they will not be able to obtain a waiver of the schedule if they cannot demonstrate, through substantial, credible evidence, at least sixty days before the completion deadline, the extraordinary circumstances beyond their control that leave them unable to comply with the schedule, including "a detailed explanation of the activities that the carrier has undertaken to meet the implementation schedule prior to requesting an extension of time." This is especially applicable to Phases I and II, given that we now are granting carriers additional time during those phases specifically so that they can implement number portability more gradually. We will not look favorably upon a waiver request if the carrier has not taken significant action to implement portability, if the carrier does not place orders with switch manufacturers in a timely manner, or, for example, if the carrier requests a waiver for a Phase II market because it only began preparing for implementation for a Phase I market in the first quarter of 1998, and then claims that it has too many software upgrades to perform from January through May 15, 1998. Carriers should be able to identify any specific technical problems that may necessitate an extension of the deployment deadline for Phase I during the four months between the scheduled end of the Chicago trial and the deadline for requesting an extension for Phase I, especially because carriers will be receiving initial feedback from testing in Chicago far in advance of the Chicago trial's conclusion. As noted above, the participants in the Chicago trial have committed to providing weekly progress reports as the trial progresses. Initial tests of LRN hardware and software on a subset of switches in the Chicago MSA began in January 1997. Intra-network and database testing in Chicago is scheduled to take place for several months before the start of the Chicago trial mandated by the Commission. 83. Our decision to extend the deadlines for completing Phases I and II of our deployment schedule reflects the fact that we consider network reliability to be of paramount importance. Consistent with that commitment, in the First Report & Order we delegated authority to the Chief, Common Carrier Bureau, to monitor generally the progress of number portability implementation and take appropriate action, as well as establishing a procedure for individual LECs to obtain an extension of the deployment deadlines as necessary for their specific markets. The Chief, Common Carrier Bureau, will monitor the weekly reports from the Chicago trial and any other pertinent developments. We find that further adjustment of the deployment schedule in response to these developments is more properly a matter for the Chief, Common Carrier Bureau, to handle as number portability technology is tested and carriers discover any actual, specific difficulties. If significant problems arise during the Chicago trial, or other significant implementation problems arise during Phase I, the Chief, Common Carrier Bureau, has the authority to adjust the schedule for the Chicago trial or the deadline for Phase I implementation, as appropriate, to ensure network reliability. 84. Although the findings of the Bellcore study submitted by SBC were vigorously challenged by AT&T and MCI, it bears mention that extending the Phase I completion date by three months is responsive to the recommendation in the Bellcore study that we should allow additional "time for testing, integration, and soaking (limited use of the software in a live environment for a length of time sufficient to find initial defects) of the software." In fact, the Bellcore study specifically recommended that the Commission "[e]xtend the time interval for introduction of [number portability] by 3 months." Our extension of Phase I, in combination with our conclusion that carriers need provide portability only in requested switches, also allows carriers the flexibility to introduce portability more gradually, beginning with a subset of switches within the MSA. 85. We deny the petitions to extend the deployment deadlines for all markets or otherwise provide wireline carriers greater flexibility in the schedule to implement long-term number portability. Although we conclude that initial implementation of this new number portability technology may require additional time, we are not persuaded that implementation in subsequent phases, after the technology has already been tested and installed in the initial markets, need be delayed to the extent requested by some petitioners. We find on the basis of the record in this proceeding that the implementation schedule as revised in this First Reconsideration Order is reasonable, and that granting any further delay of the schedule at this time is premature and unnecessary, especially because there is still approximately one year before LECs must complete deployment for the earliest phase. Petitioners have only speculated that unpredictable events may, at some point in the future, generally delay implementation, and have not shown that a specific factor will render the later schedule impossible to meet for any particular reason, much less for any particular LEC. 86. For example, despite NYNEX's vague claim that switch vendors cannot meet current workloads, no party has submitted any evidence refuting the specific vendor representations cited in the First Report & Order that vendors will be able to begin providing software for at least one long-term number portability method around mid-1997. Indeed, GTE admits that it "has no reason to doubt that [the switch vendors Lucent, Northern Telecom, Siemens, and Ericsson] can meet their commitments" to "begin supplying LRN software in early- to-mid 1997." It is our understanding that the switch vendors Lucent and Nortel are still on schedule to provide LRN software by mid-1997. 87. NYNEX also claims that the time estimated for software upgrades does not account for the fact that most upgrades will take place on weekends in order to minimize system disruptions, and generally alleges that "[t]here are probably not enough weekends" to complete deployment according to the schedule. NYNEX fails, however, to specify the additional time that, according to its estimate, would be necessary to complete the necessary upgrades. Similarly, GTE claims generally that more time is necessary to install the software for long-term number portability in all switches and remove transitional number portability capabilities, but does not estimate the additional time it believes it would need. GTE also claims in general terms that the schedule does not accurately reflect the time needed to modify vendor software if state commissions dictate inconsistent rate centers, but does not explain to what extent the rate centers are inconsistent and thus need modified software, nor does it show that a specific amount of additional time will be needed. 88. Petitioners' arguments are even more speculative given that their implementation obligations are likely to be significantly lighter than they assume, because, as we discuss above, LECs are required to deploy number portability only in switches for which they receive requests for number portability capability. Moreover, even if the problems identified by petitioners do in fact develop, in our First Report & Order we established a procedure for LECs to obtain an extension of the deployment deadlines as necessary, and delegated authority to the Chief, Common Carrier Bureau, to monitor the progress of number portability implementation. 89. In addition, contrary to petitioners' claims, the schedule set forth in the First Report & Order did allow time for factors such as the need to modify OSS (e.g., ordering and billing systems) and the need to upgrade the number portability-specific and the generic switch software. We noted in the First Report & Order that we based the schedule largely on state commission deployment schedules. State commissions and workshops, organized under their auspices and composed of industry representatives, have been and still are working to resolve deployment issues and many of the associated issues that petitioners now claim warrant delay. While studying these issues in detail, those states that established deployment schedules prior to our First Report & Order nonetheless concluded that deployment could commence in certain MSAs in those states by mid-1997. Moreover, since October, 1993, the industry, under the auspices of the Number Portability Workshop of the Industry Numbering Committee (INC), has been "assess[ing] the technical feasibility and implementation requirements, impacts, and attributes of number portability," including the factors identified by petitioners. We conclude that state commissions were well aware of the concurrent need to resolve associated issues such as modification of OSS, and allowed time to resolve these associated issues when setting their long-term number portability implementation schedules. While the First Report & Order did not expressly consider how much time is necessary to modify OSS and to upgrade software, the schedule the Commission adopted relies upon the prior work and expertise of the state commissions in establishing those state implementation schedules. At any rate, the extensions of Phases I and II, as discussed above, should alleviate any potential concerns about the sufficiency of time for modifying OSS and upgrading switch software. 90. We have concluded that a modest extension of the deployment schedule for Phase I (and Phase II) markets is warranted to allow more time for testing and modifications to be made when local number portability is first implemented. We do not believe, however, that speculative and unspecified concerns about possible future technical concerns are sufficient to justify an across-the-board delay in implementing number portability in view of the adverse effects of delay on competition in local markets. The Commission found in the First Report & Order that number portability is essential to effective facilities-based competition in the provision of local exchange services. Extending the schedule now for deployment of long- term number portability, beyond the modifications adopted in this First Reconsideration Order, based on unsubstantiated concerns will thus hamper the development of that competition. Such an extension, moreover, would conflict with the 1996 Act's intent to open monopoly local telecommunications markets to competition as soon as possible. 91. Moreover, we are not persuaded by the argument that we should delay the implementation schedule to account for problems that some other LECs may experience, due to differences in LEC networks that may prevent them from deploying number portability at the same speed. We believe that Congress, in requiring the provision of number portability "to the extent technically feasible," did not intend for LECs that are capable of providing number portability according to our deployment schedule to delay deployment on the grounds that some other LECs may encounter technical obstacles in adapting their networks. We recognize, as Bell Atlantic points out, that the BOCs were permitted to develop and deploy equal access pursuant to a more relaxed schedule. The BOCs, however, did not have a statutory mandate to deploy equal access as soon as it was technically feasible to do so, and no party has shown that the schedule established by the Modification of Final Judgment (MFJ) for deployment of equal access could not have been accelerated. 92. Furthermore, we find it unnecessary to act on GTE's request that we clarify that LECs may obtain a waiver if they cannot meet the schedule for reasons beyond their control. The waiver procedure established in the First Report & Order for extending deployment deadlines as necessary provides an effective vehicle for addressing any problems in implementing number portability that LECs can document. We note that carriers may file petitions for waiver of the deployment schedule more than 60 days in advance of an implementation deadline, and thus receive relief earlier, if they are able to present substantial, credible evidence at that time establishing their inability to comply with our deadlines. 93. We reject USTA's proposal to give every state commission and/or workshop the authority to extend independently our deployment deadlines according to their assessments of the level of local competition in an area. As set forth above, we require carriers to identify the switches in which they desire number portability capability well before the deadline for deployment in a particular MSA. We find that this requirement will enable LECs to deploy number portability in areas in which local competition is likely to develop at an early stage, while relieving LECs of the obligation to install the capability in areas that competitive LECs have no initial interest in serving. This requirement, in our view, addresses USTA's concerns by striking a reasonable balance between a LEC's interest in avoiding unnecessary switch upgrades, and a competitive LEC's interest in having assurances that number portability will be available in areas where it plans to compete to serve existing LEC customers. 94. We decline to expedite the Chicago trial, as requested by NYNEX. The First Report & Order scheduled the completion date for the Chicago trial for as early as appeared reasonably possible at that time. Given the record before us now, we conclude that it would not be possible to accelerate the commencement of that trial. Moreover, we agree with the Chicago trial participants that it would be inappropriate to shorten or delete any of the planned testing. 95. We also decline to order additional field tests, as requested by NYNEX. The requirement that there be a field trial in Chicago is only intended to ensure that at least one field trial is held to identify technical problems in advance of widespread deployment, which will provide all carriers, as well as the Commission, with information on implementation. All carriers will have an opportunity to monitor testing in Chicago and evaluate the results of the testing on an ongoing basis. We find, moreover, that LECs currently have access to additional information concerning the impact of number portability on their systems, because many LECs are, and have been for some time, analyzing extensively implementation and inter-carrier OSS impact of number portability under the auspices of state and industry fora. As we stated in the First Report & Order, we do not routinely schedule field trials in rulemaking proceedings; our requiring a field trial in the Chicago MSA is an exceptional step that we adopted to safeguard against any risk to the public switched telephone network. The need for any further trials should be determined by the industry. 96. To the extent that other networks differ in design or switch use or other relevant variables, we do not preclude the testing of either software or hardware in other areas or by other carriers, either contemporaneously with the Chicago trial or even before that trial begins. Indeed, we encourage carriers to test portability within their own networks as early as possible. For example, Bell Atlantic plans to do "first office application" testing in Gaithersburg, Maryland, from July 15, 1997, to August 30, 1997. The Gaithersburg test, therefore, will have been completed seven months before Bell Atlantic's March 31, 1998, deadline to complete implementation in Philadelphia, the market in which it must deploy long-term number portability in Phase I under our revised schedule. In any event, carriers should have the opportunity to perform their own testing, including on "live traffic," well before the date by which they must request any waiver of the Phase I implementation requirements. 97. We also decline to adopt NYNEX's proposal to deploy portability in smaller MSAs instead of the largest ones during Phase I of the deployment schedule. At this time, there is only speculation that starting with the most populous MSAs will result in technical problems. Indeed, carriers are further ahead in preparing for number portability in many of the larger MSAs than in the smaller ones; for example, several state commissions that had addressed the issue of number portability before issuance of the First Report & Order had ordered that deployment begin in several major cities that are currently in Phases I or II of our schedule. Therefore, switching the deadlines of those larger MSAs with other, smaller MSAs now would, at a minimum, disrupt planning by competitive LECs and state commissions in those jurisdictions. Moreover, our three-month extension of the end date of Phase I, in combination with our conclusion that carriers need provide portability only in requested switches, will serve much the same purpose as NYNEX's request by allowing carriers the flexibility to begin deployment in a subset of switches within each of the Phase I MSAs and gradually increase coverage over the six-month period. In addition, we do not prohibit, but rather encourage, carriers to take whatever additional actions they believe are necessary to safeguard their networks, including testing deployment of portability in one of their smaller MSAs before or during Phase I of our deployment schedule. For example, Bell Atlantic is testing number portability in the smaller market of Gaithersburg, MD before Phase I. 98. We also deny NYNEX's request that we explicitly encourage states to be flexible in opting out of the regional database or choosing to construct joint databases, or to work with less active neighboring states to establish regional databases. We find that the First Report & Order allows sufficient flexibility for states to opt out of the regional databases. In addition, NYNEX's concern that the NANC would not resolve the database issues in time for carriers to meet the deployment schedule is now largely moot, given the recent activities of the NANC. The NANC has committed to making its final recommendations to the Commission on the database system by May 1, 1997. The NANC's working groups and task forces relating to number portability are already organized and holding regular meetings to resolve the database issues. The Local Number Portability Administration Selection Working Group projects that all seven regional databases will be ready for testing on dates ranging from April 18, 1997, to July 1, 1997, and will be ready to support number portability deployment on or before October 1, 1997, in accordance with the deployment schedule set forth in the First Report & Order. 99. Finally, we clarify that the first performance criterion, that any method "support existing network services, features, and capabilities," refers only to services existing at the time of the First Report & Order. We caution LECs that problems in implementing their chosen number portability method due to modifications necessitated by the introduction of a new service or technology will not justify a delay of the deployment schedule. We decline, however, specifically to prohibit the introduction of any new service that is incompatible with LRN, as the First Report & Order did not adopt LRN or mandate use of any specific long-term number portability method. 4. Acceleration of Implementation Schedule 100. Pleadings. Several competitive LECs urge us to accelerate the deployment schedule in smaller markets. ACSI contends that the present schedule incorrectly assumes larger markets will experience competition first. ACSI claims that in fact many competitors are focusing on MSAs beyond the largest 50. Consequently, accelerating the deployment of number portability in those smaller markets would promote competition in all markets and treat competitors more equally regardless of the size of market they are entering. ACSI proposes that the schedule for the 100 largest markets be accelerated so that all BOCs implement number portability "according to roughly the same schedule as a function of population served." ACSI also proposes requiring non-BOC incumbent LECs to deploy portability in their largest market in the fourth quarter of 1997, or, at the latest, the first quarter of 1998. In the alternative, ACSI urges us to allow carriers with "operational networks in the 100 largest MSAs and the authority to provide local exchange services" to request, beginning July 1, 1997, the deployment of number portability on a specified date six or more months in the future. ACSI would place upon an incumbent LEC the burden of proving that it cannot provide number portability, and proposing an alternative date for implementation no more than three months later than the date requested. ALTS agrees that incumbent LECs should be required to implement portability in a region before the scheduled implementation date for that region, if the incumbent LEC is able to do so. 101. For markets outside the 100 largest MSAs, KMC contends that we should require LECs to accept bona fide requests for deployment of portability, after January 31, 1997, and to require implementation of such requests within six months. Alternatively, KMC urges us to require all LECs immediately to accept bona fide requests for markets outside the 100 largest MSAs, and to satisfy such requests within 24 months, unless the LEC can prove technical infeasibility. Another option presented by KMC and ACSI is to permit carriers to submit requests for markets outside the 100 largest MSAs as early as July 1, 1998, so that implementation of these requests can begin immediately upon completion of deployment in the 100 largest MSAs. NEXTLINK urges us to accept requests earlier than the First Report & Order allows for the provision of number portability for markets outside the 100 largest MSAs, consult with the relevant state commission regarding the extent of competition in that requested market, and grant the request if there is "sufficient evidence" of competition. NEXTLINK contends that accelerating the schedule will not be overly burdensome because incumbent LECs may still seek a waiver. 102. Several BOCs, GTE, USTA, and ALLTEL oppose accelerating the deployment schedule for markets below the top 100 MSAs. BellSouth and GTE assert that accelerating deployment will impede the phased deployment and jeopardize carriers' ability to meet the original schedule. GTE, USTA, Sprint, and Pacific argue that resources would be diverted from deployment in larger markets where competitors are more likely to be interested in entering. BellSouth asserts that, if we add new central offices to the schedule, then we should remove a corresponding number of central offices from the original schedule. BellSouth also suggests that, if we add new central offices to the schedule, then failure to implement any of those additional MSAs according to deadlines on the original schedule should not constitute a failure to meet a Section 271 checklist requirement. In response to ACSI's claim that the Commission's schedule discriminates against smaller markets, GTE asserts that the phased schedule takes into account the differing levels of local exchange competition in different areas, the burden on carriers serving multiple regions, and the fact that more significant upgrades may be necessary for carriers operating in smaller areas. 103. USTA supports KMC's recommendation that LECs be able to submit requests for deployment in markets outside the 100 largest MSAs earlier than January 1, 1999, but proposes that such requests be fulfilled on a negotiated timetable subject to the decisions of each state commission, instead of by December 1998, or within 24 months, as suggested by KMC. ALLTEL contends that accelerating the schedule will force carriers to file waivers or seek suspensions of implementation. ALLTEL argues, moreover, that smaller providers should not be required to invest in number portability technologies until they have been proven reliable in larger markets. 104. Discussion. We deny the petitions for reconsideration that advocate: (1) accelerating deadlines for certain MSAs; (2) allowing carriers with operational networks in the 100 largest MSAs and the authority to provide local exchange service to request portability in any MSA in the 100 largest MSAs beginning July 1, 1997, and requiring LECs to fulfill such requests on a specified date six or more months in the future; (3) adding MSAs outside the largest 100 MSAs to the initial deployment schedule; or (4) combining the deadlines of consolidated MSAs. The current schedule is based on the projected availability of switch software, and recognizes the burden on carriers serving multiple regions and the fact that more significant upgrades may be necessary for carriers operating in smaller areas. Petitioners have not made a showing that the necessary software, hardware, and other resources will be available earlier in areas originally scheduled for later deployment, or will be available in quantities sufficient to support deployment in additional areas, particularly in areas outside the 100 largest MSAs. If such hardware and software is not available for deployment early enough or in sufficient quantities to support deployment in additional areas, then accelerating deployment deadlines for smaller MSAs may divert these limited resources from deployment in other, larger MSAs, and thus delay deployment of number portability where a greater population might benefit from competition. 105. For the reasons stated above, we also reject ACSI's request to require deployment in Phase I in certain additional markets in which the incumbent LECs are not BOCs. In addition, we continue to believe that non-BOC incumbent LECs, most of which have more limited resources than the BOCs, should have additional time to upgrade and test their networks. Moreover, we conclude above that LECs need deploy number portability in the 100 largest MSAs only in switches for which another carrier has made a specific request for the provision of portability. Requiring that additional MSAs be deployed in Phase I does not give sufficient notice to carriers or states to establish switch-requesting procedures in MSAs for which they had no previous notice that deployment was required in Phase I. We also decline to adopt USTA's proposal that state commissions be free to accelerate the deployment schedule. While we are sympathetic to the desires of some states to advance deployment where actual competitive interest exists, we conclude that the schedule adopted in the First Report & Order, as modified in this First Reconsideration Order, represents a reasonable balancing of competing interests, and carriers need to have certainty that these are the requirements with which they must comply. Our First Report & Order was silent on the issue of whether states could accelerate the deployment schedule. We therefore grandfather any state decisions to accelerate deployment for a particular market from one phase to an earlier phase that were adopted prior to release of this First Reconsideration Order. 106. We do not prohibit LECs from agreeing to accelerate implementation, either for specific MSAs or specific switches within MSAs. We find, however, that acceleration of our schedule is more properly determined by private agreements among carriers. Competitive LECs are free to negotiate with incumbent LECs for deployment of number portability ahead of our schedule. Moreover, to the extent that carriers agree to "swap" the implementation deadlines for specific MSAs or switches within MSAs, they can jointly file specific waiver petitions to do so. 107. We grant in part the petitions of ACSI, KMC, and NEXTLINK to allow requests for deployment of number portability in areas outside the 100 largest MSAs to be submitted earlier than January 1, 1999. We therefore modify our rules to permit carriers to submit requests for deployment of number portability in areas outside the 100 largest MSAs at any time. We decline, however, to require that deployment be completed within six months of request for requests filed prior to January 1, 1999. This modification to our rules will benefit all parties, because receiving earlier notice to upgrade switches will likely ease a LEC's compliance burden and help to ensure that competing carriers will receive portability within the time requested. Finally, we clarify that, contrary to KMC and ACSI's view, our current schedule does not leave an implementation gap between December 31, 1998, and July 1, 1999, since implementation of requests for deployment of number portability in areas outside the 100 largest MSAs filed on or before January 1, 1999, will occur during the first six months of 1999. KMC and ACSI's suggestion that we permit requests for markets outside the 100 largest MSAs beginning July 1, 1998, and require fulfillment of those requests within six months, would actually require that those smaller markets be completed at the same time as the MSAs in the last phase of our deployment schedule, thus sharply increasing the burden on carriers during that phase. 5. Exemptions for Rural and/or Smaller LECs 108. Pleadings. JSI, NECA, and NTCA/OPASTCO argue that requiring rural LECs to provide number portability where no competitor has requested it will burden rural LECs significantly without benefitting the public by increasing competition. NECA and NTCA/OPASTCO state that requiring rural LECs to provide portability absent such a request contravenes our intent to let the pace of competitive entry into local markets determine the need for number portability. NTCA/OPASTCO asserts that these small businesses do not have the resources to test portability technologies. GTE argues that the same concerns that prompted us to forego an implementation schedule for areas outside the 100 largest MSAs, i.e., lack of imminent competition and the need for significant network upgrades, apply to smaller offices within the 100 largest MSAs. JSI, NECA, and NTCA/OPASTCO contend that the need for such an exemption is apparent in Congress' Joint Explanatory Statement which states, "Duties imposed under new Section 251(b) make sense only in the context of a specific request from another telecommunications carrier . . . ." 109. JSI and NTCA/OPASTCO suggest that we exempt rural LECs operating within the 100 largest MSAs from complying with the implementation deadlines until receipt of a request for deployment. Several other parties agree with JSI and NTCA's suggestion, and would expand the exemption to include: (1) LECs with less than five percent of their subscribers in an MSA, or LECs with only 10 percent of their access lines within an MSA; (2) rural LECs with study areas that only partially overlap one of the 100 largest MSAs; or (3) any carrier with less than two percent of the nation's access lines. JSI further argues that we should not apply our deployment requirements to rural LECs until there is factual evidence that number portability is technologically feasible, and will not disproportionately burden rural LECs. GTE suggests that, if no competitors express an interest in entering the market, and the state commission does not object, smaller LECs should be allowed to present a waiver to us that, if approved, would exempt them from portability requirements until six months after a request is made. CBT suggests that we exempt from the implementation schedule carriers granted a suspension or modification of the number portability requirements under Section 251(f)(2) until the state commission removes the suspension. 110. USTA urges us to exempt from the deployment schedule rural LECs that are exempt from interconnection requirements under Section 251(f). JSI goes further and argues that Section 251(f)(1) prohibits the imposition of number portability requirements on rural LECs because rural LECs are automatically exempt from the interconnection requirements of Section 251(c). JSI states that this exemption from interconnection requirements permits us to impose number portability requirements upon rural LECs only to the extent it is technically feasible for rural LECs to provide portability without upgrading their networks to utilize databases, installing SS7 or AIN capabilities, or installing and furnishing functions requiring new switching software. JSI adds that this exemption may be terminated only by a state commission. In addition, JSI argues, the Commission recognized in the First Report & Orderthat carriers meeting the 251(f) criteria may be exempt from number portability requirements. 111. NTCA/OPASTCO claims that the First Report & Order's Final Regulatory Flexibility Analysis does not address the impact of the rules on small incumbent LECs, and is thus inconsistent with the Local Competition Order. NTCA/OPASTCO suggests that exempting rural LECs from number portability requirements absent a specific request would fulfill our responsibility under the Regulatory Flexibility Act. 112. Time Warner and MCI oppose any "blanket waiver" of number portability requirements for smaller and/or rural LECs. MCI argues that such waivers will "lessen the likelihood" that competition will ever reach areas served by smaller and rural providers. MCI claims that a blanket waiver is unnecessary, because smaller and rural LECs can receive waivers under the statutory provision or under the