UNITED STATES DEPARTMENT OF JUSTICE
July 23, 2002
UNITED STATES DEPARTMENT OF JUSTICE
Introduction and Summary
The United States Department of Justice ("the Department"), pursuant to Section 271(d)(2)(A) of the Telecommunications Act of 1996(1) ("the 1996 Act"), submits this evaluation of the joint application filed by Qwest Communications International, Inc. on June 13, 2002, to provide in-region, interLATA services in Colorado, Idaho, Iowa, Nebraska, and North Dakota. Qwest's Joint Application to the Federal Communications Commission ("FCC" or "Commission") for these states is its first for long distance authority in any of its states.
As the Department has explained, in-region, interLATA entry by a regionalBell Operating Company ("BOC") should be permitted only when the local markets in a state have been "fully and irreversibly" opened to competition.(2) This standard seeks to measure whether the barriers to competition that Congress sought to eliminate with the 1996 Act have in fact been fully eliminated and whether there are objective criteria to ensure that competitive local exchange carriers ("CLECs") will continue to have nondiscriminatory access to the facilities and services they will need from the BOC in order to enter and compete in the local exchange market. In applying its standard, the Department considers whether all three entry paths contemplated by the 1996 Act -- facilities-based entry involving the construction of new networks, the use of the unbundled elements of the BOC's network ("UNEs"), and resale of the BOC's services -- are fully and irreversibly open to competitive entry to serve both business and residential customers.
Qwest's application demonstrates that it has succeeded in opening its local markets in Colorado, Idaho, Iowa, Nebraska, and North Dakota in most respects. However, Qwest's application as filed does not demonstrate that it provides CLECs with electronically auditable wholesale bills for the UNE-platform nor does it adequately address issues relating to Qwest's manual processing of wholesale orders. Thus, the Department cannot support Qwest's application as filed. However, Qwest has since submitted substantial additional evidence which, if sufficiently meaningful and reliable for the Commission to assure itself that Qwest is providing electronically auditable wholesale bills and manually processing wholesale orders timely and accurately, would justify the granting of long distance authority in Colorado, Idaho, Iowa, Nebraska, and North Dakota.(3)
As a threshold matter, the Department addresses the public interest violations that have been alleged by various commenters. Among the most serious is the allegation that Qwest entered into "secret" interconnection agreements with various competitors, offering preferential pricing and access to operations support systems ("OSS"), as well as requiring these competitors to refrain from participating in regulatory dockets, including Section 271 proceedings, in opposition to Qwest.(4) Commenters assert that the existence of such agreements evidences discrimination in violation of the 1996 Act and undermines the integrity of both the OSS test and commercial performance measurement data on which Qwest relies to support the application, and therefore requires its dismissal.(5)
These allegations are serious and deserve the Commission's careful attention.(6) The Department does not comment on whether Qwest's earlier failure to file the agreements violated Section 251 or 252. If the Commission finds that a violation has occurred, sanctions may be appropriate and could include suspension or revocation of any Section 271 authority that the Commission may have granted in the interim.(7) However, it is not apparent that the remedy for such prior violations, if any, lies in these proceedings rather than in effective enforcement through dockets in which such matters are directly under investigation.(8) Moreover, such allegations of past discrimination do not appear to implicate the Department's inquiry into whether local exchange markets are fully and irreversibly open to competition for purposes of providing its Evaluation of a pending Section 271 application to the Commission.(9)
Nevertheless, these allegations appear to implicate evaluation of this Section 271 application to the extent they ultimately may raise questions as to the quality of the record.(10) Performance data relating to the CLECs that are alleged to have received preferential treatment are included in the aggregate data included in Qwest's filing and were relied upon by KPMG during portions of the Regional Oversight Committee("ROC") OSS test.(11) AT&T and others assert that such preferential treatment, assuming it occurred, undermines the meaningfulness of those data as gauges of Qwest's current performance and as benchmarks to enforce adequate performance in the future.(12) The Department agrees that accurate benchmarks of performance attained are critical, but arguably any enhanced performance caused by the allegedly preferential treatment will have resulted in higher benchmarks for Qwest to maintain.(13)
Since the "secret" agreements were entered into, the extraordinary efforts of the ROC OSS test and the various checklist workshops appear to have improved Qwest's documentation, systems, and processes to the point where they support all CLECs equally, if imperfectly. The three-year process might well have been more efficient and comprehensive with the full and open participation of all interested CLECs, but the fact that certain CLECs did not participate does not appear to have had a significant impact on the result.(14) For these reasons, the Department has chosen to consider the merits of Qwest's pending application as reflected in the record rather than recommending that the Commission dismiss the application until the allegations regarding the "secret" agreements have been resolved.
I. State Commission Section 271 Proceedings
The Colorado Public Utilities Commission ("Colorado PUC"), Idaho Public Utilities Commission ("Idaho PUC"), Iowa Utilities Board ("Iowa UB"), Nebraska Public Service Commission ("Nebraska PSC"), and North Dakota Public Service Commission ("North Dakota PSC"), working independently and together, have facilitated the development of competition in their respective local telecommunications markets.
A. Regional Oversight Committee Proceedings/OSS Test
In 1999, the ROC, a cooperative group of state regulatory commissions in the Qwest local service region, including all five of the applicant states,(15) initiated a collaborative process focusing on the attainment of Section 271 authority by examining the legal framework for opening the local markets, and by designing and executing a third-party OSS test to require and prove the practical availability of the necessary wholesale support systems.(16) A thorough, detailed, and open process was used to set up the collaborative governing structure, determine the overall scope of the test, select the third-party testers,(17) and design the Master Test Plan and Performance Indicator Definitions ("PIDs"). Believing that the relevant OSS were regional, but recognizing the possibility of variations based on the legacy system differences,(18) the ROC included a "Regional Differences Assessment" by KPMG as a prelude to the actual test effort.(19) KPMG did a preliminary assessment of both processes and systems, and concluded that these were generally "materially consistent" across the three Qwest operating regions, with the exception of the three different service order processors ("SOPs") and billing systems.(20) KPMG recommended, and the ROC concluded, that a regional test could be meaningfully executed as long as the test transaction volumes were set and distributed across the three regions.(21) The test was so executed.
No CLEC has alleged that the regional approach was inappropriate, or that the underlying OSS are too dissimilar to permit such an overall evaluation.(22) In fact, a separate OSS test was conducted in Arizona during the pendency of the ROC testing process,(23) and CLECs commonly took preliminary findings and concerns raised in each forum to the other, asserting the regional nature of the systems and the relevance of each test's findings to the other. Qwest, too, has relied on the Arizona test reports as additional support for its OSS Declarations.(24)
When the third-party test(s) began in 2000, Qwest's documentation, systems, and processes all appeared to have been unsatisfactory. KPMG and HP issued 256 Exceptions and 242 Observations(25) throughout the course of the test. During that time, repeated iterations of documentation, systems and processes, and substantial retesting, improved Qwest's wholesale support systems to the point where only a few questions regarding their adequacy to support competitive local entry remain.(26)
B. Multistate Collaborative Process
The Idaho PUC, Iowa UB, and North Dakota PSC worked with four other states in the Multistate Collaborative Process ("MCP"), which involved a fact-finding process on aspects of the competitive checklist, Section 272 Track A requirements, and the public interest analysis, including post-entry performance assurance.(27) The MCP included numerous collaborative workshops, during which CLECs, Qwest, and state agency staffs addressed and recommended resolution of disputed issues, with each state retaining authority to accept or reject such recommendations.(28)
C. Individual State Commission Proceedings
The Colorado PUC established several dockets to address independently those issues dealt with by the MCP, among others(29); conducted extensive pricing proceedings to establish wholesale rates for UNEs,(30) which provided the benchmarks for the recent adjustment of rates in Idaho, Iowa, Nebraska, and North Dakota as well(31); and adopted performance measurements developed through the ROCand an independently devised Colorado Performance Assurance Plan ("CPAP"), intended to ensure that an appropriate level of wholesale performance is maintained once Qwest's Section 271 application is approved.(32)
The Idaho PUC independently reviewed the MCP record(33); adopted the performance measurements and standards developed through the ROC, including the "QPAP" post-entry performance plan(34); and conducted arbitrations to establish initial UNE rates and recently accepted Qwest's adjustment of core UNE rates using the new Colorado rates as benchmarks.(35)
The Iowa UB conducted an independent review of the MCP record(36); adopted the performance measurements and standards developed through the ROC, as well as the QPAP(37); and held a series of UNE pricing proceedings to establish initial rates and recently accepted Qwest's adjustment of core UNE rates using the new Colorado rates as benchmarks.(38) The Iowa UB also investigated Qwest's compliance with Sections 251 and 252 of the 1996 Act, tentatively concluding that Qwest had violated those provisions by failing to file certain amendments to interconnection agreements it made with CLECs that subsequently dropped their opposition to Qwest's regulatory initiatives.(39)
The Nebraska PSC, although not initially involved in the MCP, reviewed and adopted the record of the MCP and held independent hearings on competitive checklist items, Track A and Section 272 requirements, and Qwest's actual commercial performance(40); adopted the performance measurements and standards developed through the ROC, as well as the QPAP(41); and conducted proceedings to establish initial UNE rates and recently accepted Qwest's adjustment of core UNE rates using the new Colorado rates as benchmarks.(42)
The North Dakota PSC conducted an independent review of the MCP record(43); adopted the performance measurements and standards developed through the ROC, as well as the QPAP(44); and conducted arbitrations to establish initial UNE rates and recently accepted Qwest's adjustment of core UNE rates using the new Colorado rates as benchmarks.(45)
II. Entry into the Local Telecommunications Markets
In assessing whether the local markets in a state are fully and irreversibly open to competition, the Department looks first to the actual entry in a market.(46) But the Department does not broadly presume that all three entry tracks -- facilities-based, unbundled network elements ("UNEs"), and resale -- are open or closed on the basis of an aggregate level of entry alone.(47) The following table reports CLEC entry in Colorado, Idaho, Iowa, Nebraska, and North Dakota in terms of shares of total residential and business lines served and shares of residential and business lines served by each mode of entry.
Given the regional nature of Qwest's OSS, the Department evaluates the state of entry regionwide, taking note that pricing or other state-specific factors may significantly affect the degree to which CLECs use a specific mode in a specific state. In the states addressed in this application, the amounts of entry, and the absence of evidence that entry has been unduly hindered by problems with obtaining inputs from Qwest, lead the Department to conclude that opportunities are available to competitive facilities-based and UNE-platform carriers serving business customers.(52)
Regarding competition for residential customers, the Department finds that the facilities-based mode of entry is open in Colorado and Nebraska, due to the amounts of entry and the increasingly widespread availability of service provided by cable companies to these customers. Although in the other states there is less competition to serve residential customers via facilities (including UNE loops) or the UNE-platform, or both -- and the Department therefore cannot directly presume openness -- the lower levels of penetration may reflect the higher UNE pricing that was in effect for most of the period preceding this application. The Colorado PUC ordered reductions in certain UNE rates within weeks of Qwest's filing of this application,(53) and Qwest reduced prices in the other four states on the basis of the reductions ordered by the Colorado PUC.(54)
III. Pre-Order/Order Integration
Qwest asserts it provides integrated pre-order/order processes such that a CLEC can electronically transfer information returned on a pre-order response to a subsequent order without having to manipulate the data.(55) The purpose is not to transfer the information per se but to use the pre-order information provided by Qwest to populate an order that will be successfully processed and provisioned.(56) AT&T and WorldCom argue that Qwest's pre-order and ordering systems are not adequately integrated, namely, that the documentation is unclear, the fields do not match, and a disproportionate percentage of orders are rejected due to this faulty integration.(57)
Qwest's regional reject rates for orders received via EDI, whether rejected automatically or manually, are high: approximately 30 percent of all local service requests ("LSRs") are rejected.(58) WorldCom reports similar results for its UNE-platform orders and contrasts this unfavorably with much lower reject rates in other regions.(59) A significant percentage of orders submitted by HP, acting as the pseudo-CLEC during the ROC OSS test, was rejected.(60) Yet these figures are consistent with those reported in prior applications,(61) as to which the FCC has found -- notwithstanding such aggregate reject rates -- that successful integration was possible based on the range of results in the commercial data for various CLECs, including relatively low rates for certain CLECs.(62)
Qwest's application does include HP's conclusion that successful integration was possible(63) as well as evidence that two service bureaus have successfully integrated the pre-order and order interfaces.(64) However, although Qwest maintains that it is providing a fully parsed customer service record ("CSR"),(65) some CLECs claim it is not.(66)
Qwest's evidence of integration is similar to that offered by BellSouth in support of its Georgia/Louisiana application,(67) except that Qwest does not offer migration by telephone number ("TN migration") and its unique definition of "migrate as specified" requires the CLEC to specify the features to remove as well as those to place on an account, which doubles the CLECs' reliance on the CSR for those orders.(68) Both distinctions will tend to increase the reject rate, whether due to the inadequate integration of Qwest's pre-order/order systems or due to wrong address or feature information in the CSR,(69) and may increase the level of manual processing.(70) The Department has therefore considered the integration issues as primarily relevant to the degree and adequacy of Qwest's manual handling of CLEC orders.
IV. Manual Handling of Orders
CLECs allege that Qwest subjects far too many orders to manual handling, and that its manual handling processes are prone to error such that CLEC customers are likely to have their orders inaccurately or untimely processed and provisioned.(71) In evaluating the adequacy of Qwest's manual processes, the Department examines not only the level of flow-through achieved by Qwest but also the timeliness and accuracy of its electronic order status notices and its order provisioning.(72)
Qwest's flow-through of eligible orders generally meets the relevant benchmarks of 70 to 90 percent and has done so from January through May of this year.(73) The flow-through of all orders electronically received is, as would be expected, much lower -- approximately 40 to 80 percent, depending on the product.(74) Clearly, a large quantity of electronically submitted orders are being manually handled by Qwest; however, in determining the adequacy of Qwest's OSS, the quality of the manual handling is more important than the quantity of orders manually handled.(75) Key indicia of such order processing include Qwest's return of firm order commitments ("FOCs") and rejects, return of error messages (including jeopardy notices) after having returned a FOC,(76) rescheduling of due dates for orders,(77) as well as Qwest's ability to timely and accurately install orders. Qwest's commercial performance reports indicate it has largely met the benchmarked intervals for FOC and reject returns.(78) Similarly, Qwest's performance on the ROC OSS test appears to have been timely for the return of both FOCs and rejects.(79)
In general, the timeliness of Qwest's wholesale provisioning appears analogous to Qwest's retail provisioning.(80) However, if Qwest manually mis-coded the order receipt date, the result would be both to extend the CLEC's due date and to mis-report the receipt date for performance reporting purposes, thus "hiding" the additional day and making the interval appear appropriately prompt.(81) Qwest asserts that its systems will not allow entry of receipt dates that occur after the date of entry, and that 98 percent of orders are handled on the day received, so that any due date delay/data integrity problem is miniscule.(82)
Qwest does not have any regularly reported commercial performance data on the accuracy of its manual order processing.(83) The Department has thus analyzed Qwest's performance in providing jeopardy notices and in changing the due dates on CLEC orders as an indicator of processing quality. Qwest has provided, on average, jeopardy notices closer to the due date for wholesale non-designed services in Colorado than for analogous retail services.(84) It is unclear what percent of the jeopardies related to facilities issues and what percent related to non-fatal errors on orders which should have been caught pre-FOC.(85) Similarly, Qwest's commercial performance reports indicate that due date changes on wholesale orders have exceeded the number of due date changes on retail orders.(86) These disparities are not sufficient in themselves to dictate a conclusion that Qwest's OSS are inadequate, but they suggest that further analysis is warranted.
KMPG analyzed Qwest's service order accuracy and found that Qwest's performance was satisfactory. In the absence of a documented PID or standard for provisioning accuracy, KPMG compared both the submitted LSR to the resulting CSR to ascertain whether the ordered services were properly reflected on the resulting CSR,(87) and the LSR to the switch translations record to determine whether the ordered features were properly provisioned in the switch.(88) Qwest passed both tests.(89)
KPMG's provisioning accuracy reviews do not appear to have distinguished between electronically and manually processed orders, but during the ROC OSS test KPMG posited that Qwest's apparent need to continuously enhance training could reflect an underlying inadequacy in manual performance that could have a negative impact on a CLEC's ability to compete.(90) In response, Qwest provided KPMG with additional information about its training and monitoring.(91) KPMG reviewed and verified the information, and recommended that the Observation be designated "closed/resolved."(92) However, during a subsequent retest of a systems fix,(93) KPMG noticed errors on manually handled orders notwithstanding Qwest's previously verified training and process enhancements.(94) This unexpected result caused KPMG to review data on manually handled orders that it had collected on other portions of the test (and which post-dated the previous enhancements), and KPMG concluded that some 15 percent of these had been erroneously handled.(95) KPMG recognized that its sample size was small and recommended additional retesting to focus on the orders that fall out for manual handling; Qwest elected not to support a retest, so the Observation was designated "closed/unresolved."(96)
The lack of regularly reported commercial data on manual accuracy renders the record incomplete. The Department believes that this is a serious issue, particularly given the expert tester's carefully expressed concerns. But for the concerns expressed by KPMG at the close of the test, the positive results on the underlying test criteria would appear to support a finding that Qwest proved the overall adequacy of its processes. Since filing its application, Qwest has submitted substantial evidence regarding its own internal tracking of manual order accuracy, and, if reliable, this data could support a finding that Qwest's processes are sufficient to permit CLECs a meaningful opportunity to compete.(97) The Department remains concerned, however, that there is no process that would permit CLECs and regulators to monitor and maintain adequate performance as volumes increase. Therefore, the Department agrees with KPMG's assessment that further measures are necessary to permit continued monitoring,(98) recognizes Qwest's willingness to implement a new performance measure and make available information on internal manual accuracy tracking,(99) and believes this monitoring should be implemented promptly to ensure that Qwest continues to maintain the requisite accuracy of manual handling.(100)
V. Electronically Auditable Billing
The Commission has determined that a BOC must provide CLECs with "complete, accurate, and timely reports on the service usage of competing carriers' customers" and "complete, accurate and timely wholesale bills" in order to give CLECs a meaningful opportunity to compete.(101) This requirement, including auditability, was emphasized in the Department's Pennsylvania Evaluation, which stated that "[a]ccurate and auditable electronic bills are an important factor in making local telecommunications markets fully and irreversibly open to competition."(102) The FCC agrees with the importance the Department has placed on accurate, electronically auditable bills.(103)
One method of providing electronic auditability of wholesale bills is through billing in the Bill Output Specifications-Bill Data Tape ("BOS-BDT") format,(104) which Qwest did not implement until July 1, 2002.(105) It is unclear whether Qwest's billing system, absent reliance on BOS-BDT, satisfies the requirement of electronic auditability.(106) Qwest's billing system enables CLECs to receive their Customer Record and Information System ("CRIS") bills(107) in a variety of electronic formats, including EDI, ASCII, and, most recently, BOS-BDT.(108)
The CRIS bills that CLECs receive include master summary bills and individual bill detail for each CLEC end-user's account.(109) For each state in which a CLEC is doing business, CRIS provides separate summary bills for resale products and for each UNE product.(110) The summary bill contains sub-accounts for each CLEC-end user account number and lists the total number of all the services billed during the billing cycle.(111)
Qwest's application to the Commission described the levels of detail available in Qwest's wholesale bills,(112) but did not clearly address the issue of whether its wholesale bills are electronically auditable. Qwest in subsequent ex partes has provided detail to support its additional assertion that these electronic bills can be loaded into publicly available software to simplify the auditing process, thus proving the electronic auditability of the CRIS system.(113)
CLECs, in their filings with the Commission, have directly disputed the electronic auditability of CRIS bills, stating that Qwest's electronic bills reside in a "proprietary format that preclude their verification."(114) Unfortunately, due to the timing of ex parte filings, CLECs have not yet had an opportunity to comment on Qwest's most recent documentation of the electronic auditability of its CRIS bills.(115) The Department is not in a position to resolve this dispute, but recommends that the Commission reconcile Qwest and CLEC claims, and assure itself that Qwest is providing electronically auditable bills to CLECs before approving this application.(116)
VI. Change Management Process
Qwest must demonstrate a change management process sufficient to determine necessary changes to its OSS and to implement them appropriately.(117) Qwest relies on the record related to its new, redesigned change management process ("CMP") to meet this requirement.(118)
The CMP redesign process was an intense, collaborative effort that lasted nearly a year.(119) Although certain aspects of the redesign, particularly those governing product and process changes, were only recently agreed upon by Qwest and the participating CLECs, key provisions of the CMP have been in place for more than six months,(120) and Qwest has compiled evidence relating to its overall compliance with the new CMP.(121) Qwest has offered both PID data and data derived from its tracking of internal milestones to demonstrate the adequacy of its adherence to key elements of the CMP Framework.(122)
CLECs do not dispute the overall adequacy of the redesigned CMP, but question whether Qwest has proved that its implementation of the new process complies with checklist requirements.(123)
CMP redesign and implementation is a dynamic process.(124) Although Qwest has not offered data related to all aspects of the CMP it has provided evidence that tends to show it has been properly implementing and adhering to the CMP.(125) The Department concludes that Qwest adequately outlines its record of compliance with basic CMP elements that have been in place for more than six months, as well as those procedures that were implemented April 1, 2002.(126) Moreover, all five state commissions relevant to this application have participated in the redesign process, have concluded that the CMP and Qwest's record of implementation comply with the FCC's criteria,(127) and have evidenced an intention to actively monitor the CMP and require continued compliance by Qwest.(128)
VII. Stand Alone Test Environment
A robust test environment, one that mirrors and yet is separate from the BOC's production environment, is "critical to opening local markets,"(129) and is a "fundamental part of a change management process ensuring that competing carriers are capable of interacting smoothly and effectively with a BOC's OSS."(130) Qwest asserts its Stand Alone Test Environment ("SATE") satisfies this requirement.(131) SATE uses a front-end interconnect mediated access-electronic data interchange ("IMA-EDI") that is identical to the production copy and a "stubbed" environment to simulate back-end systems; SATE does not use live production systems for test transactions.(132) SATE was developed in August 2001 in response to CLECs' and KPMG's concerns that Qwest's "interoperability" environment was inadequate to support CLEC testing and market entry.(133) In response to concerns subsequently voiced about SATE's adequacy,(134) Qwest recently instituted automated post-order responses through the Virtual Interconnect Center Knowledge Initiator ("VICKI") on January 26, 2002, as well as flow-through capability throughout the region on May 20, 2002.(135)
CLECs allege that Qwest's SATE does not mirror the production environment and is therefore inadequate to support the testing by CLECs that is the prerequisite to successful implementation of systems to support their local entry and continuing competition.(136)
The requirement that the test environment mirror production means that they must perform the "same key functions," not that the two environments must be identical.(137) SATE's rapid evolution, including the recent implementation of functionalities such as VICKI and flow-through, cause some concern, especially because the adequacy of these new capabilities have not been verified in an independent test.(138) However, the Department believes that SATE is generally designed to -- and does -- operate similarly enough to the production environment to be an effective tool.(139) Qwest uses production copies of the IMA system in order to replicate real-world production.(140) The structure of the data in SATE mirrors the structure of the data in production.(141) SATE Version 9.0 contains all IMA-EDI generated error messages that occur in production as well as common legacy system errors.(142) Although in some instances the response received in SATE may not be identical to that which would have been received in production, Qwest documents any differences between the IMA production environment and SATE in the IMA-EDI SATE Data Document.(143)
HP, through its transaction testing of SATE in Arizona, found the accuracy and consistency of SATE test responses to be adequate to support certification.(144) Recent additions of flow-through and VICKI, and future increases to the products available to be tested in SATE will further the extent to which SATE mirrors the real-world production environment.(145) As of June, 16 CLECs (5 through software service bureaus) have successfully used Qwest's SATE.(146) Commercial performance reports indicate that SATE's accuracy throughout the region has been very close to the benchmark of 95 percent compliance with the documentation and business rules.(147)
Qwest, in collaboration with CLECs and state commissions, is modifying and disaggregating the results from PID PO-19, until now the only performance measurement for SATE accuracy, and creating PID PO-19B to measure the extent to which SATE mirrors real-world production results by comparing the results of SATE and real production transactions.(148) One large, unresolved concern is that Qwest does not yet have data for this measurement; however, at the behest of the Colorado PUC, Qwest will be required to include such a metric in its PAP and therefore report the associated data.(149) The Department supports the Colorado PUC's intent to track information relevant to ensuring that SATE permits CLECs to adequately test their interfaces.
VIII. TELRIC Pricing
The rates for unbundled elements and interconnection in each of the five states were adjusted shortly before Qwest filed this Section 271 application with the Commission. The Colorado PUC issued an initial pricing order in December 2001 and a final pricing order on June 6, 2002, and subsequently the state commissions in Idaho, Iowa, Nebraska, and North Dakota each ordered new rates effective June 7, 2002.(150)
The new prices in Idaho, Iowa, Nebraska, and North Dakota were derived through an exercise in which Qwest attempted to benchmark the rates in those states to the new ones ordered in Colorado by using the Commission's USF model.(151) The FCC has established that comparisons between states can be a useful way to set as well as analyze rates and sees no reason to "make a distinction between cost-based rates and rates that equal cost-based rates."(152) However, to be valid, the exercise must rely on accurate benchmark comparisons between the states at issue. Qwest relied on the Commission's model to determine average cost differentials between the Qwest regions in different states.(153) However, Qwest failed to adjust the model's cost estimates to reflect the sale of certain high-cost exchanges.(154) Due to this oversight, the cost estimates in regions that have lost exchanges are likely too high on average. Qwest's benchmarking process, which depended on these inflated cost differentials, will inevitably lead to UNE prices for those regions that are also too high on average. This and other alleged errors appear to have distorted the resulting rates relied upon by Qwest, so that it is unclear whether they fall within a range that a reasonable application of TELRIC would produce.(155) The Department urges the Commission to carefully consider this issue,(156) although it defers to the Commission's ultimate determination of whether the prices supporting this application are appropriately cost-based.(157)
Qwest's application demonstrates that it has succeeded in opening its local markets to competition in many respects. Although the Department cannot support Qwest's application as filed, Qwest has since submitted substantial additional evidence which, if sufficiently meaningful and reliable, would support the granting of long distance authority in Colorado, Idaho, Iowa, Nebraska, and North Dakota. Based on this record, and subject to the Commission's assuring itself that the concerns expressed in this Evaluation have been resolved, the Department recommends that the FCC approve Qwest's application.
July 23, 2002
I hereby certify that I have caused a true and accurate copy of the foregoing Evaluation of the United States Department of Justice to be served on the persons indicated on the attached service list by first class mail, overnight mail, hand delivery, or electronic mail on July 23, 2002.
1. Pub. L. No. 104-104, 110 Stat. 56 (1996) (codified as amended in scattered sections of 47 U.S.C.).
2. See DOJ Oklahoma I Evaluation at vi-vii, 36-51.
3. The Department also recommends that the Commission evaluate whether the prices in the Qwest region, particularly those in Idaho, Iowa, Nebraska, and North Dakota, which were developed recently and without input from competitors, comply with TELRIC. See infra text Section VIII.
4. AT&T Comments at 15-24; CompTel Comments at 13-15; New Edge Comments at 3-4; Touch America Comments at 24.
5. AT&T Comments at 15-17, 25-28; CompTel Comments at 14-15; New Edge Comments at 3-4. Commenters also allege that Qwest has violated Section 271 since its acquisition of US West by continuing to provide in-region long distance services notwithstanding the requirements of the Commission's merger order that Qwest divest itself of these assets. CompTel Comments at 7-13; Touch America Comments at 12-14.
6. Qwest has petitioned the Commission to declare what constitutes an interconnection agreement within the meaning of the 1996 Act, suggesting that the agreements at issue are not interconnection agreements that must be filed. Qwest Petition for FCC Declaratory Ruling at 28-37.
7. See 47 U.S.C. § 271(d)(6)(A) ("If at any time after the approval of an application . . . the Commission determines that a Bell operating company has ceased to meet any of the conditions required for such approval, the Commission may . . . (i) issue an order to such company to correct the deficiency; (ii) impose a penalty on such company pursuant to title V; or (iii) suspend or revoke such approval.").
8. The Arizona Corporation Commission ("Arizona CC"), Minnesota Public Utilities Commission ("Minnesota PUC"), and New Mexico Public Regulation Commission ("New Mexico PRC") have opened investigations into the matter. See generally Arizona CC Order; Minnesota Department of Commerce Complaint; New Mexico PRC Order. The Colorado PUC is informally investigating the matter, but declined to consider the matter relevant to Section 271 proceedings. See Colorado PUC Section 271 Order at 26-31; Colorado PUC Comments at 64. The Iowa UB has tentatively concluded that Qwest's failure to publically file several such agreements violates Sections 251 and 252 of the 1996 Act. Iowa UB Order at 16; see also Arizona CC Staff Recommendation at 1-2 (recommending Arizona CC assess fines against Qwest for noncompliance with Sections 251 and 252, and failure to publically file interconnection agreements).
9. Cf. FCC Kansas/Oklahoma Order ¶ 27 ("[T]he statute does not require that a BOC demonstrate that it has been in compliance with section 271 for some period of time before it files a section 271 application.").
10. See Arizona CC Spitzer Letter at 1 (arguing Arizona CC Section 271 docket must be suspended until allegations that parties were coerced from participating in the docket are investigated and resolved); Arizona CC Irvin Letter at 1 (same).
11. See generally KPMG CLEC Participation Report (listing test criteria for which KPMG partially and substantially relied upon information obtained from the CLECs in question).
12. See AT&T Comments at 30 (asserting KPMG test is of "no value because the results were based on input from the CLECs that received preferential secret deals treatment"); CompTel Comments at 14-15 (recommending review of CLEC-specific performance data to determine whether there is an inexplicable gap between the best reported performance and the aggregate).
13. DOJ Georgia/Louisiana I Evaluation at 34 n.115 ("In order to establish effective benchmarks that readily can be used to hold an incumbent to an appropriate level of wholesale performance, metrics must neither understate nor overstate actual performance. An incumbent should not be able to argue that it need not maintain a particular level of performance because, due to a problem with the relevant metric, past reports created an inappropriate benchmark."). Obviously, in this case, to the extent any historical discrimination in favor of certain CLECs improved Qwest's aggregate level of wholesale performance, Qwest should not be able to argue that such benchmark levels are inappropriate and need not be maintained.
14. Colorado PUC Comments at 64 ("In a 'but for' world, the potential impact of CLEC nonparticipation in the collaborative process is, at worst, close to nil.").
15. See Qwest Br. at 3 (noting the participation of eight other Qwest region state commissions as well); see generally Qwest Notarianni/Doherty Decl. (extensive description of test development, implementation, and results).
16. Qwest Notarianni/Doherty Decl. ¶ 19 & Ex. 3 (letter from ROC to then USWest, proposing regionwide, collaborative OSS test (June 23, 1999)).
17. In July 2000, KPMG was retained to serve as test administrator, HP was retained to serve as the "pseudo-CLEC" (the entity submitting transactions to Qwest), and Liberty Consulting was retained to audit the performance data. Id. ¶ 27. Later, Liberty also reconciled inconsistencies in performance data reported by Qwest and those collected by CLECs. Qwest Williams Decl. ¶¶ 47-53.
18. USWest was the product of three predecessor BOCs: Pacific Northwest Bell, Mountain Bell, and Northwestern Bell. Qwest's three distinct operating regions (Western, Central and Eastern) reflect this history, as each has a slightly different legacy, or back-office, OSS. Qwest's Western region comprises Washington and Oregon; its Central region comprises Arizona, Colorado, Idaho, Montana, New Mexico, Utah, and Wyoming; and its Eastern region comprises Iowa, Minnesota, Nebraska, North Dakota, and South Dakota. Qwest Notarianni/Doherty Decl. ¶ 34.
19. Id. ¶¶ 35-36.
20. Id. & Ex. 4 (KPMG Regional Differences Assessment (Oct. 5, 2000)).
22. See FCC Georgia/Louisiana Order ¶¶ 109-11 (describing analytical road-map FCC used in assessing the regional nature of BellSouth's OSS and determining the finding of "sameness" permitted the FCC to evaluate Georgia test and commercial data in support of the Louisiana application, as well as Louisiana commercial data in support of the Georgia application).
23. See Qwest Notarianni/Doherty Decl. ¶ 19n.17 (noting Arizona test was underway when ROC OSS test process convened).
24. Id. ¶ 36 n.36. The Arizona CC selected Cap, Gemini, Ernst and Young ("CG") as its test administrator. CG issued its final test report on December 21, 2001; post-report testing, particularly on billing, continued through April 2002. CG Final Report at 113. The Arizona CC also retained CG to audit Qwest's performance data, Qwest Williams Decl. ¶ 44, although Liberty undertook the data reconciliation for Arizona and retained HP to evaluate the adequacy of Qwest's Stand-Alone Test System ("SATE"), CG Final Report at 25-26.
25. During the course of the test, KPMG and HP issued "Observations" and "Exceptions" to document issues of concern that arose. The severity of the occurrence as well as the potential impact to the CLEC dictated whether the issue was classed as an "O" or an "E." Qwest Notarianni/Doherty Decl. ¶¶ 37-39.
26. Only one Observation and 14 Exceptions were designated "closed/unresolved" by the conclusion of the test. Id. ¶ 39 n.39.
27. See Qwest Br. at 7 (noting the participation of Montana Public Service Commission, New Mexico PRC, Utah Public Service Commission, and Wyoming Public Service Commission as well).
28. See id.
29. See, e.g., Colorado PUC Comments at 9.
30. Id. at 26-36.
31. Qwest Br. at 163-67.
32. Colorado PUC Comments at 7-8, 54-60; see also Qwest Reynolds CPAP Decl. ¶ 14 (noting state commission modified some performance measures for purposes of CPAP).
33. Idaho PUC Comments at 3-5.
34. Id. Ex. B at 1-10, Ex. D at 3-5.
35. Id. Ex. D at 6-7; see also Qwest Br. at 10.
36. Iowa UB Comments at 2-3; see also Qwest Phillips Decl. ¶¶ 51-69.
37. Iowa UB Comments at 68-70.
38. Id. at 26-27.
39. See also supra note 8 and accompanying text.
40. Nebraska PSC Comments at 1-7; see also Qwest Sandos Decl. ¶¶ 5-12, 18-19.
41. Nebraska PSC Comments at 4-5; see also Qwest Sandos Decl. ¶¶ 13-15.
42. Nebraska PSC Comments at6-7; see also Qwest Br. at 12.
43. North Dakota PSC Consultative Report at 5-7; see also Qwest Macintosh Decl. ¶¶ 8-22.
44. North Dakota PSC Consultative Report at 204-56.
45. Id. at 259-65; see also Qwest Br. at 13.
46. See DOJ Pennsylvania Evaluation at 3-4 ("The Department first looks to actual competitive entry, because the experience of competitors seeking to enter a market can provide highly probative evidence about the presence or absence of artificial barriers to entry. Of course, entry barriers can differ by types of customers or geographic areas within a state, so the Department looks for evidence relevant to each market in a state." (Footnote omitted.)).
47. See, e.g., DOJ Georgia/Louisiana I Evaluation at 7 ("Although the Department presumes that fully facilities-based competition is not hindered in a competitively significant manner based on the entry recorded in Georgia, the amount of entry does not justify extending such a presumption to other modes of entry in Georgia."); DOJ Missouri I Evaluation at 6-7 ("The Department presumes that opportunities to serve business customers by fully facilities-based carriers and resellers are available in Missouri, based on the entry efforts reflected in SBC's application. There is significantly less competition to serve residential customers. There also is less competition by firms seeking to use UNEs, including the UNE-platform, and there are some indications that a failure by SBC to satisfy all of its obligations may have constrained this type of competition." (Footnotes omitted.)).
48. See Qwest Teitzel Decl. at 20 (table of line counts as of March/April 2002) & Exs. DLT Track A/PI-CO-1 at 2, 5, -PI-ID-1 at 2, 4, -PI-IA-1 at 2, 5, -PI-NE-1 at 2, 4, -PI-ND-1 at 2, 4. The second three categories report CLEC lines as percentages of total lines, and all business and residential lines, respectively; the last six categories report percentages of business and residential lines served by CLECs by means of each mode of entry, i.e., facilities-based (service via primarily a CLEC's own fiber optic network that is either connected directly to the customer premises or connected through loops leased from the BOC), UNE-platform (a combination of loop, switch, and transport elements), and resale.
Qwest offers two sets of calculations of line estimates, see id. ¶¶ 36-41, and, as explained previously, the Department relies on the calculations based on E-911 database entries, see, e.g., DOJ Georgia/Louisiana I Evaluation at 8 n.24. Estimated market share will vary depending on the methodology used to estimate facilities-based lines; in any event, the effective difference between the two methods is small, resulting in estimates that CLECs serve between, respectively, 21.4 (E-911 Listings Method) and 20.5 (LIS Trunk Method) percent of Colorado lines, 10.6 and 8.8 percent of Idaho lines, 17.8 and 18.4 percent of Iowa lines, 29.9 and 21.5 percent of Nebraska lines, and 21.0 and 16.0 percent of North Dakota lines. Qwest Teitzel Decl. ¶¶ 55-56.
49. Figures report total lines in Qwest's service area in all states, each of which has several incumbent local exchange carriers other than Qwest. See, e.g., id. ¶ 37 & n.44.
50. Qwest UNE-Platform Calculations Ex Parte at 1-2 (explaining estimate of business and residential line counts).
52. CLECs in Iowa and North Dakota have migrated large numbers of business customers from resale to the UNE-platform, through the "UNE-P-Star" alternative. The different levels of UNE-platform business and residential entry in Iowa and North Dakota may reflect the extent to which various features of the UNE-P-Star alternative, including a degree of flat-rate pricing, may be particularly conducive to the migration of business resale customers to the UNE-platform. See Qwest DOJ Issues Ex Parte Tab 15 at A-B. Notwithstanding the apparent restriction in the UNE-P-Star agreements, Qwest has clarified that this alternative is available for serving both residential as well as business customers. Qwest UNE-P-Star Ex Parte at 1.
53. Qwest Br. at 150-62.
54. See, e.g., id. at 10; see also id. at 162-67. But see infra text Section VIII.
55. Qwest Notarianni/Doherty Decl. ¶ 195.
56. FCC Georgia/Louisiana Order ¶ 119.
57. AT&T Comments at 39; WorldCom Comments at 7; WorldCom Lichtenberg Decl. ¶¶ 21, 23.
58. Qwest Performance Results Ex Parte at 80, 73, 76, 73, and 67(PID PO-4B-1 & -2) (LSRs Rejected for EDI) (Rejected Manually and Auto-Rejected January-May 2002 data for Colorado, Idaho, Iowa, Nebraska, and North Dakota, respectively). The results for the five states are identical because the underlying data is collected on a regionwide basis. In reporting percent rejects Qwest's PIDs do not distinguish between types of orders, e.g., between orders for resale or UNE-platform service.
59. WorldCom Lichtenberg Decl. ¶ 30 (noting WorldCom experiences approximately 30 percent rejects using Qwest's systems, 8.1 percent using Verizon's, 16.3 percent using BellSouth's, and 18.9 percent using SBC's).
60. KPMG Final Report at 81 (reject rate 30 to 40 percent for EDI orders).
61. See, e.g., FCC Georgia/Louisiana Order App. B at 14-15 (Georgia reject rate on loop and port combination orders of approximately 14 percent for mechanized rejects and approximately 32 percent for partially mechanized rejects).
62. Id. ¶ 123. Qwest did not include any comparable commercial data in its application, but has since made an ex parte filing which reflects some variation in the reject rates experienced by individual CLECs. See generally Qwest CLEC Order Reject Data Confidential Ex Parte.
63. HP actually integrated a subset of the pre-order/order fields so as to transmit orders in its roles as pseudo-CLEC. Qwest Notarianni/Doherty Decl. ¶ 198. During transaction testing, HP used the "integration techniques it had built" as well as manual processing "to ensure that both types of processing were adequately tested." Id.; cf. FCC Georgia/Louisiana Order ¶ 122 (KPMG built a parser to send test orders across BellSouth's interfaces). HP also reviewed other pre-order/order fields and concluded that CLECs with appropriate "expertise and integration experience" could integrate the systems. HP Pre-Order/Order Integration Field Comparison Report at 40 ("This does not mean that there are not issues that would have to be resolved between Qwest and the CLEC, but simply that these issues are not insurmountable.").
64. Qwest Notarianni/Doherty Decl. Ex. 12 at 1 (Letter from Telcordia Technologies to Qwest), Ex. 13 at 1 (Letter from NightFire to Qwest); see FCC Georgia/Louisiana Order ¶ 122 (CLECs and software bureaus attest to having successfully integrated BellSouth's pre-order and ordering systems.).
65. Qwest Notarianni/Doherty Decl. ¶¶ 195, 197.
66. AT&T Finnegan/Connolly/Menezes Decl. ¶ 124; WorldCom Lichtenberg Decl. ¶ 22.
67. FCC Georgia/Louisiana Order ¶ 122 (relying on evidence of actual commercial usage, the results of two third-party tests, the fact that BellSouth provides TN migration (reducing the incidence of address-related rejects), the new introduction of a parsed CSR, and BellSouth's retention of Accenture and Science Application to aid CLECs in integrating their systems).
68. WorldCom Comments at 6 (discussing TN migration), 9-10 (addressing migrate "as specified" and reliance on CSRs); see also DOJ Georgia/Louisiana I Evaluation at 23-24 (discussing TN migration).
69. FCC Texas Order ¶ 177 (characterizing as adequately low a rate of fewer than 5 percent address-related rejects); WorldCom Lichtenberg Decl. ¶ 14 (describing greater than 6 percent address-related rejects in June).
70. See FCC Georgia/Louisiana Order ¶ 145 ("[A] BOC is not accountable for orders that fail to flow through due to competing carrier-caused errors."); AT&T Comments at 41; AT&T Finnegan/Connolly/Menezes Decl. ¶ 150 n.104 ("[F]low-through rates reported by Qwest exclude all orders that fall out for manual processing due to CLEC errors."). But see Qwest Notes on Performance Results Ex Parte at 4 (noting that as of March 2002 flow-through results exclude non-fatal rejects).
71. AT&T Comments at 40-42; Covad Comments at 39-41; Eschelon Comments at 6; WorldCom Lichtenberg Decl. ¶¶ 33-42.
72. See FCC Georgia/Louisiana Order ¶¶ 143-44 (evaluating the adequacy of documentation and support for CLECs, the integration of pre-order and order interfaces, the timely and accurate receipt of order status notices, and the timely and accurate provisioning of the requisite volume of orders in considering the adequacy of BellSouth's manual handling).
73. See, e.g., Qwest Performance Results Ex Parte (Colorado) at 70-77 (PID PO-2B (Electronic Flow-through for All Eligible LSRs received via IMA-GUI and IMA-EDI) (including disaggregations for POTS-Resale, Unbundled Loops, LNP, and UNE-P POTS)).
74. See, e.g., id. at 70, 72, 74, 76 (PID PO-2A (Electronic Flow-through for LSRs received via IMA-GUI and IMA-EDI) (from January through May 2002, flow-through rates between 60 and 80 percent for resale, 44 and 70 percent for unbundled loops, 60 and 70 percent for LNP EDI, 15 and 30 for LNP GUI, and 50 and 60 percent for UNE-P (POTS), respectively)).
75. Cf. FCC Texas Order ¶ 182 ("We recognize that SWBT demonstrates a high overall flow-through rate, which necessarily means that relatively few mechanically-submitted orders will fall out for manual processing. Where an applicant is unable to demonstrate a high flow-through rate, and thus manual processing of mechanized orders is more common, we may require more extensive data with respect to this issue [of manual accuracy].").
76. Qwest codes these post-FOC rejects as jeopardies, which is consistent with the documentation. Qwest Notarianni/Doherty Decl. Ex. 18 (Qwest Service Order Management Flow chart) at 2-3. The percentage of jeopardies returned that are actually post-FOC rejects, and the reason the rejects are not being processed when the order is received, remains unclear. See infra note 85and accompanying text. Analysis of the post-FOC rejects relates to evaluation of the accuracy of the FOC that was returned.
77. Due date changes may be acceptable. See FCC Georgia/Louisiana Order ¶ 139 (allowing BellSouth's practice of sending out a FOC "blind," i.e., without a facilities check, because BellSouth does not distinguish between wholesale and retail in following this practice). On the other hand, the need for due date changes may reflect the inaccuracy of the initially issued FOC.
78. Qwest Williams Decl. ¶ 123 (noting Qwest missed the FOC benchmarks for one month in Iowa and Idaho each), ¶ 120 (noting Qwest met relevant performance benchmarks for timely return of order rejections from January through April 2002).
79. KPMG Final Report at 82-90, 262-64 (stating Qwest met or exceeded the benchmarks set by 11 of the 13 tests relating directly to FOCs); id. at 81-82 (describing the timeliness of reject notices returned).
80. See, e.g., Qwest Performance Results Ex Parte (Colorado) at 107 (PID OP-4C (Installation Interval (Average Days) - No dispatch - UNE-P (POTS)) (showing wholesale intervals of approximately 2.5 days from January through May, and retail intervals rising to 3.5 days in May); see also id. at 106 (PID OP-3C (Installation Commitments Met (Percent) No dispatches -- UNE-P (POTS)) (showing parity performance on meeting promised due dates for non-dispatch UNE-platform orders).
Qwest's installation data for line-shared loops (no dispatch) appears to comply with the benchmark interval requirement. See, e.g., id. at 228 (PID OP-4C (Installation Interval (Average Days) - No dispatch - Line-sharing). Covad and WorldCom, however, both allege that this data is compromised by Qwest's practice of automatically sending the service order completion notice ("SOC") on the due date regardless of whether the work has been completed; this "fake" SOC also impedes the CLECs' ability to properly serve their end-user customers. Covad Comments at 25-28; WorldCom Nielson Decl. ¶ 4. Neither commenter quantified the extent of the problem. Qwest appears to agree that some of the SOCs for line-shared orders have been erroneous, and has implemented a new process to ensure that line-shared orders that are not completed by 4 p.m. are placed in jeopardy status so that the service orders do not auto-complete and result in the issuance of a SOC. Qwest Line-Sharing SOCs Ex Parte at 1. SOCs are key notifiers and must be timely and accurate. The lack of information regarding the extent of the problem or its impact, however, leads the Department to conclude that Qwest's performance has been adequate, if imperfect. Qwest expects its new process will also improve its line-sharing maintenance and repair performance. Id.
81. Exception 3120 is in fact related to Qwest's ability to properly account for the receipt date of UNE-platform orders, although the recording and due-date-calculation error in those cases resulted from systems problems that Qwest fixed and KPMG verified on retest. See generally Exception 3120 Disposition Report.
82. Qwest FCC Presentation Ex Parte (Manual Processing Slide) at 1.
83. The FCC has previously relied upon trouble reports after installation as an indicator of provisioning quality, e.g., FCC Texas Order ¶ 182; FCC New York Order ¶ 201, and Qwest's commercial data for UNE-platform does show fewer wholesale repair calls than retail repair calls made after installation. See, e.g., Qwest Performance Results Ex Parte (Colorado) at 108 (PID OP-5 (New Service Installation Quality (Percent) UNE-P (POTS)) (Qwest's wholesale installation quality is more than 90 percent, although its retail installation quality is only 85 percent.)). However, Qwest's regularly reported installation quality measure does not include troubles that are submitted for missing features as Qwest considers those as raising an order processing rather than a provisioning issue. Qwest DOJ Issues Ex Parte Tab 4 at 13. Thus, Qwest does not regularly report performance data on manual order processing or service order accuracy.
84. Qwest Williams Decl. ¶¶ 130-32 (claiming apparent disparity was largely result of CLEC submission of orders with shorter lead times -- i.e., less time to issue the jeopardy -- than those allowed by Qwest's retail orders). Qwest notes that its IMA 10.0 release on June 17, 2002 adds the capability to produce automated jeopardy notifications for non-design, unbundled loops, and UNE-platform (POTS). Id. ¶ 138; see also KPMG Final Report at 91-93 (concluding that Qwest's performance in providing timely jeopardy notices for both resale and UNE-platform products did not satisfy the requisite standards).
85. Qwest has listed the reasons for post-FOC jeopardy/rejects, including "error in LSR processing." Qwest DOJ Issues Ex Parte Tab 6 at 18-19. Qwest does not have the data to disaggregate the various categories of jeopardy notices. Id.; see also KMPG Final Report at 119 (5 percent post-FOC rejects were received on test orders submitted over GUI; less than 1 percent post-FOC rejects were received on test orders submitted over EDI).
86. See, e.g., Qwest Performance Results Ex Parte (Colorado) at 96 (PID PO-15 (Number of Due Date Changes per Order average days) (showing a wholesale average of .08 changes/order and a retail comparable average of .04 changes/order)).
87. KPMG Final Report at 186-87 (Test Cross-Reference 14-1-12).
88. Id. at 182-83 (Test Cross-Reference 14-1-3).
89. Id. at 182-83, 186-87 (noting retesting was necessary on both before Qwest achieved the desired 95 percent performance level).
90. Observation 3086 Second Supp. Response at 1-2.
91. Id. at 2, 9, 14.
92. Id. at 18. KMPG issued a study of the adequacy of Qwest's performance reporting to reflect issues relating to manual processing. See generally KPMG Manual Adequacy Study.
93. Exception 3120 Disposition Report at 7-8 (relating to accuracy of Qwest's PID OP-4 (Installation Intervals)calculations). KPMG noted an inconsistency in Qwest's handling of Saturday orders; Qwest agreed that it had improperly coded the receipt date of certain LSRs and agreed to retesting of a systems fix for the problem. See generally Exception 3120 Initial Report.
94. Observation 3110 Second Response at 4.
95. See id. (noting KPMG evaluated 49 orders and found errors on seven).
97. As filed, Qwest's application included the assertions that its service order accuracy was more than 97 percent, Qwest Notarianni/Doherty Decl. ¶ 354, and that its rate of erroneous rejects was less than one percent, see Observation 3086 Second Supp. Response at 13 (Qwest response listing results of internal analysis of rejects followed by a FOC). However, these assertions were unsupported by detail regarding the underlying raw data, including the defined processes by which they were gathered (such as which service order fields were included in the calculations; whether all rejects or only manual rejects were included). This lack of detail precluded meaningful analysis; therefore, the assertions could not be relied upon. In response to the Department's queries, Qwest has provided additional detail, including corrections of certain figures. See Qwest DOJ Issues Ex Parte Tab 5 at 16-17. Qwest's audit was limited to verifying the accuracy of the "APP" (date) field. Qwest Service Order Accuracy Ex Parte at 1. Qwest's proposed service order accuracy measure, PID PO-20, will monitor the accuracy of other fields as well. Qwest Williams Decl. Attach. 2 at 2. Qwest has also submitted data on its record of LSR/SO mis-matches; the one-week sample shows a 99 percent accuracy rate. Qwest DOJ Issues Ex Parte Tab 4 at 13. CLECs, however, have not yet had the opportunity to comment on this information, or to define what gaps, if any, may remain in the data.
98. KMPG Manual Adequacy Study at 1-2.
99. Qwest has agreed to implement a new measure for service order accuracy, PID PO-20, but its definition has not been finalized. Qwest Williams Decl. ¶ 355 & Attach. 2. Qwest has also agreed to track and report LSR/SO mismatches related to manual processing. Qwest DOJ Issues Ex Parte Tab 4 at 13. Although both commitments appear to be significant, their usefulness will depend on the governing definitions and implementation, matters on which CLECs' views should be taken into account.
100. See, e.g., Colorado PUC Comments at 38-39 (describing Qwest's plan to enhance flow-through as well as to pre-populate service order numbers and due dates from the LSR on manually handled orders, and requiring as well that a PID for manual service order accuracy be added to the CPAP).
101. FCC Pennsylvania Order ¶ 13; see also AT&T Finnegan/Connolly/Menezes Decl. ¶ 230 & Attach. 20 at 1 (manual auditing impracticable as volumes rise).
102. DOJ Pennsylvania Evaluation at 11.
103. See FCC New Jersey Order ¶¶ 122, 124; FCC Pennsylvania Order ¶ 22.
104. See FCC New Jersey Order ¶ 122 n.348 (discussing need for electronic billing in BOS-BDT format to facilitate computer auditing); FCC Pennsylvania Order ¶ 17 (explaining electronic auditability of wholesale bills in BOS-BDT format). The BOS-BDT format, as compared with Qwest's "retail-formatted" bill, is designed to conform to the industry standard and thus eases the burden on CLECs when auditing. Qwest DOJ Issues Ex Parte Tab 1 Attach. 1 at 1; AT&T Finnegan/Connolly/Menezes Decl. ¶ 227.
105. Qwest DOJ Issues Ex Parte Tab 1 at 1.
106. BOS-BDT, when initiated by Verizon, encountered numerous problems with its initial deployment. See FCC Pennsylvania Order ¶ 19. Qwest only recently implemented the BOS-BDT format, independent testing has not been conducted, and CLECs have not had sufficient experience with it to comment meaningfully. Qwest DOJ Issues Ex Parte Tab 1 at 2; AT&T Finnegan/Connolly/Menezes Decl. ¶ 234; WorldCom Lichtenberg Decl. ¶ 72. For these reasons, the Department does not rely on Qwest's recent implementation of BOS-BDT as demonstrating the availability of electronically auditable wholesale bills.
107. CRIS, the focus of third-party testers, including CG and KPMG, is used for billing for most resale products and UNEs. Qwest Notarianni/Doherty Decl. ¶ 491; CG Final Report at 97; KPMG Final Report at 435.
108. Qwest Notarianni/Doherty Decl. ¶ 498; Qwest DOJ Issues Ex Parte Tab 1 at 1.
109. Qwest Notarianni/Doherty Decl. ¶¶ 494-95; Qwest DOJ Issues Ex Parte Tab 1 at 2-5.
110. Qwest Notarianni/Doherty Decl. ¶ 494.
111. Id. ¶¶ 494-95; Qwest DOJ Issues Ex Parte Tab 1 at 2.
112. Some CLECs dispute whether Qwest actually provides a sufficient level of detail. See AT&T Finnegan/Connolly/Menezes Decl. ¶ 235; WorldCom Comments at 18; WorldCom Lichtenberg Decl. ¶ 69.
113. See Qwest DOJ Issues Ex Parte Tab 1 at 2-3.
114. AT&T Finnegan/Connolly/Menezes Decl. ¶ 230; see also AT&T Comments at 45-46; WorldCom Comments at 18-19; WorldCom Lichtenberg Decl. ¶¶ 67-73.
115. Neither KPMG in the ROC nor CG in Arizona assessed the electronic auditability of Qwest's wholesale CRIS bill. See KPMG Final Report at 435-80 (discussing billing test criteria, but not auditability). Moreover, the record is unclear whether the audit process by which KPMG verified the accuracy of the bills was manual or electronic.
116. The accuracy of Qwest's billing is subject to less dispute. On the whole, Qwest's commercial performance and the third-party testing has satisfied the Department that, despite limited problems, Qwest's wholesale billing meets the requirements for accuracy. See Qwest Performance Results Ex Parte at 102, 94, 98, 88 (PID BI-3A (Billing Accuracy) (Parity) (Colorado, Idaho, Iowa, and North Dakota, respectively)); see also id. at 94 (same in Nebraska, except in May 2002, when parity not achieved); cf. KPMG Final Report at 435-54 (evaluating Qwest's ability to accurately bill various charges on the appropriate bill type). Some CLECs claim to have received inaccurate bills from Qwest; however, billing inaccuracy does not seem to be a widespread problem. See AT&T Comments at 46; AT&T Finnegan/Connolly/Menezes Decl. ¶¶ 237-38; Eschelon Comments at 22-24; WorldCom Comments at 18; WorldCom Lichtenberg Decl. ¶ 73; cf. FCC Pennsylvania Order ¶ 26 (describing steady improvement in Verizon's billing accuracy so that the dollar figures in dispute declined from 26.59 percent in February to 2.21 percent in June). Additionally, the Department is satisfied with Qwest's ability to adequately report service usage to CLECs.
117. DOJ Georgia/Louisiana I Evaluation at 29; FCC Georgia/Louisiana Order ¶ 179 (listing five criteria by which the adequacy of a change management process is judged).
118. See Qwest Br. at 131-49.
119. Qwest Filip Decl. ¶¶ 9-10.
120. Qwest Br. at 143.
121. Qwest asserts it "promptly implemented every aspect of the redesigned change management plan as soon as it has been agreed upon in the joint CLEC/Qwest redesign process." Id.; see also Qwest Filip Decl. ¶¶ 9-20, 143-72 & Ex. 5.
122. Qwest Br. at 144-45. The Department is aware that this percentage record of performance may be slightly inflated by Qwest's having included a large number of internal milestones in its analysis. See AT&T Finnegan/Connolly/Menezes Decl. ¶¶ 48-50. These internal data are not inconsistent with the formally reported PID results, however. For example, Qwest met the benchmark for timely release notification in four of the last five months ending in May 2002. Qwest Performance Results Ex Parte at 97, 89, 93, 89, 83 (PID PO-16) (Colorado, Idaho, Iowa, Nebraska, and North Dakota, respectively). Qwest issued all five release notifications on time in May 2002. Qwest Notarianni/Doherty Decl. ¶ 693. More significantly, no CLEC has alleged with specificity any Qwest failure to meet a CMP-mandated, CLEC-affecting deadline since establishment of the revised CMP.
123. AT&T Comments at 29; AT&T Finnegan/Connolly/Menezes Decl. ¶¶ 36-39; CompTel Comments at 3-4; Eschelon Comments at 28; WorldCom Comments at 19; WorldCom Lichtenberg Decl. ¶¶ 74-78.
124. See DOJ Georgia/Louisiana II Evaluation at 15-18 (noting modification of BellSouth's change process in response to concerns voiced in DOJ Georgia/Louisiana I Evaluation, and Department's intent to monitor future applications for continued compliance); see also FCC Georgia/Louisiana Order ¶ 193 (observing BellSouth's implementation of CLEC-prioritized change requests and noting the importance of BellSouth's future collaboration with CLECs through the Change Control Process).
125. See generally Qwest Filip Decl. Ex. 5; see also supra note 122. Qwest asserts that, in any case, it is not required to include product and process change provisions in a CMP. Qwest Br. at 144 n.70. It is not necessary to decide this issue; however, the importance of having an adequate CMP to facilitate an efficient competitor's ability to compete on nondiscriminatory terms supports a broad, rather than limited, interpretation of requirements for both the systems and the product-and-process aspects of change management. See, e.g., CompTel Comments at 3; New Edge Comments at 4-5.
126. Qwest Filip Decl. ¶¶ 143-72 & Ex. 5 at 12-15.
127. Colorado PUC Comments at 4. The Colorado PUC agrees that Qwest's contention that " 'has in place the most comprehensive, inclusive and forward-looking change management plan in the nation.' " Id. at 45; see also Iowa UB Comments at 39 (concluding Qwest satisfied CMP requirements); Nebraska PSC Comments at 6 (noting Qwest's CMP meets the requirements of the 1996 Act); cf. Idaho PUC Comments at 11-12 (asserting that "while this [compliance] record is not as lengthy or perfect as desired, it is substantial, and perfection is not required"); North Dakota PSC Comments at 172 (stating that "[d]espite the fact that the redesign work on the CMP is continuing, the . . . record demonstrates that Qwest's [CMP] . . . satisf[ies] the FCC requirements for Section 271 checklist compliance").
128. The Colorado PUC asserts it has "some measure of confidence that Qwest will maintain an adequate Product and Process CMP," based on Qwest's recent addition of five levels of Product and Process changes to the CPAP, penalties associated with which the PUC will incorporate into the SGAT. Colorado PUC Comments at 48. The Idaho PUC also has committed to continue monitoring CMP implementation and support "the development and inclusion into the QPAP of applicable PIDs." Idaho PUC Comments at 12.
129. DOJ Georgia/Louisiana I Evaluation at 26-27.
130. FCC Georgia/Louisiana Order ¶ 187.
131. Qwest Br. at 136-41; Qwest Notarianni/Doherty ¶¶ 705-69, 779.
132. HP SATE Summary Report at 7.
133. KPMG Final Report at 580-81. KPMG, through its review of Qwest's OSS, identified deficiencies in the interoperability environment: it requires CLECs to use valid production data in their test cases, responses to the test cases are manually generated, and there is no flow-through capability. Id. at 580 (Test Cross-Reference 24.6-1-8, Exception 3029). In responding to Exception 3029 and CLEC concerns, Qwest developed SATE and decided that it would devote its resources to SATE and make no further enhancements to the interoperability environment. Qwest SATE White Paper at 4; see also KPMG Final Report at 580. Accordingly, the Department has focused on the adequacy of SATE, as have the state commissions. See, e.g., Colorado PUC Comments at 50-52; North Dakota PSC Comments at 175.
134. KPMG Final Report at 580-81 (Test Cross-Reference 24.6-1-8; Exceptions 3077 and 3095). KPMG still had significant concerns about SATE, including the lack of automated transaction responses in SATE, the lack of flow-through for transactions, restrictions on volume of order responses due to this manual response handling, lack of wholesale products available for testing, and apparent inconsistencies in the data contained within the SATE order responses that may not mirror responses that would be found in production. Id. (Test Cross-Reference 24.6-1-8; Exceptions 3077 and 3095).
135. Qwest Br. at 137, 139; Qwest Notarianni/Doherty Decl. ¶¶ 747, 755-60.
136. AT&T Finnegan/Connolly/Menezes Decl. ¶¶ 93, 103, 107; WorldCom Lichtenberg Decl. ¶¶ 84-85, 87. CLECs' main concerns include differences between the responses received in SATE and production and the fact that CLECs are required to choose a "path" in order to receive an automated response in SATE but not in production. Id.
137. FCCGeorgia/Louisiana Order ¶ 189 (quoting FCC Texas Order ¶ 138 ("The record indicates, however, that competing carriers are able to test adequately OSS changes prior to their implementation as long as testing and production environments perform the same key functions.")). In Texas, SBC did not allow for testing flow-through or response times, yet the FCC concluded that SBC's testing environment was adequate. FCC Texas Order ¶ 138 ("Thus, despite any differences between the testing and production environments, the totality of the evidence indicates that SWBT's testing environment is adequate."). The FCC also rejected the assertion that testing environments must fully test orders end-to-end and determined there to be adequate functionality. FCC Georgia/Louisiana Order ¶ 189.
138. See KPMG Final Report at 580-81; see generally HP SATE Summary Report; HP New Release Test Summary Report. In its supplemental evaluation of SATE 9.0, however, HP was able to use the newly-activated VICKI on 77 scenarios and did not raise any issues about VICKI in its report. HP SATE New Release Test Summary Report at 32.
139. But see Colorado PUC Section 271 Order at 46 (concluding that SATE complies with FCC's requirements but describing Qwest's SATE as the "closest call of the whole § 271 record").
140. Qwest Br. at 137; Qwest Notarianni/Doherty Decl. ¶¶ 734-35.
141. Qwest Notarianni/Doherty Decl. ¶ 737.
142. Id. ¶¶ 736, 762.
143. Id. ¶¶ 735, 763. This documentation is available on Qwest's web-site at www.qwest.com/wholesale/ima/edi/document.html>. See generally id. Ex. 48. Despite Qwest's acknowledgments, some CLECs believe there should be no differences. AT&T Finnegan/Connolly/Menezes Decl. ¶ 107; WorldCom Lichtenberg Decl. ¶ 89.
144. HP SATE Summary Report at 8.
145. Qwest Notarianni/Doherty Decl. ¶¶ 755, 758-60.
146. Qwest SATE Ex Parte at 5; see also Qwest SATE Confidential Ex Parte at 1-6 (listing CLECs using SATE).
147. Qwest Performance Results Ex Parte (Colorado) at 98, 90, 94, 90, 84 (PID PO-19 (Stand Alone Test Environment (SATE) Accuracy) (Colorado, Idaho, Iowa, Nebraska, and North Dakota, respectively).
148. Qwest Notarianni/Doherty Decl. ¶ 742. ArizonaCorporation Commission ("CC") staff requested such a mirroring measurement for the new releases of SATE. See Arizona CC Supp. Report on Qwest's Checklist Compliance ¶ 149.
149. Colorado PUC Comments at 51; Colorado PUC Section 271 Order at 144.
150. See Colorado PUC Comments at 26-36; Idaho PUC Comments Ex. D at 6-7; Iowa UB Comments at 26-27; Nebraska PSC Comments at 6-7; North Dakota PSC Comments at 259-65.
151. Qwest Br. at 163-67.
152. FCC Kansas/Oklahoma Order ¶ 82 n.244, ¶ 87.
153. Qwest Br. at 163-66.
154. See Qwest Pricing Ex Parte at 1-9; Qwest DOJ Issues Ex Parte Tab 16 at 34.
155. The alleged errors appear to have had the most significant effect on the calculation of the rates for North Dakota. See AT&T Comments at 51-54 (citing Qwest's inclusion of high-cost exchanges it had sold, failure to use state-specific minutes-of-use ("MOUs"), and the filing of additional rate elements -- e.g., "grooming" and "cross-connect" charges -- at the same time it had lowered the group of select UNE rates); WorldCom Comments at 29-32 (raising issues involving high-cost exchanges and state-specific MOUs).
156. In addition, Qwest's non-recurring charge ("NRC") for a hot cut may be significantly higher than those that have been presented in other Section 271 applications. See AT&T Comments at 61 (comparing Qwest's hot-cut NRC of $171.88 and "basic loop" installation NRC of $55.27 to rates in other states); cf. DOJ New Jersey I Evaluation at 7-8 (comparing New Jersey hot-cut NRC of $159.76 to lower rates in neighboring states); FCC New Jersey Order ¶ 62 (describing Verizon's reduction of the New Jersey NRC to $35). As inflated NRCs can be a barrier to competitive entry, the Department urges the FCC to carefully consider whether Qwest's NRCs are appropriately cost-based.
Qwest has offered to set a price of $0 for the high-frequency portion of the loop ("HFPL"), but the Colorado PUC rejected that in apparent confusion between economic benefit and cost. See Colorado PUC Pricing Order II at 17 ("Zero cannot be the right price for a scarce good with a positive demand."). A positive HFPL price would seem within the guidance of TELRIC for some reallocation of common costs, but it appears that Colorado is deviating from that in insisting on a positive price without adjusting any other loop prices downward to reflect such reallocation. Colorado PUC Pricing Order I at 117-18 (commission will consider adjusting loop costs in future). The Department believes that the FCC should pursue this matter directly with the state commission.
157. DOJ Rhode Island Evaluation at 6; DOJ Missouri I Evaluation at 1-2; DOJ Kansas/Oklahoma Evaluation at 11.