|This document is available in two formats: this web page (for browsing content) and PDF (comparable to original document formatting). To view the PDF you will need Acrobat Reader, which may be downloaded from the Adobe site. For an official signed copy, please contact the Antitrust Documents Group.|
FEDERAL COMMUNICATIONS COMMISSION
Washington, D.C. 20554
EVALUATION OF THE
UNITED STATES DEPARTMENT OF JUSTICE _______________________________________________________
FEDERAL COMMUNICATIONS COMMISSION
Washington, D.C. 20554
EVALUATION OF THE
UNITED STATES DEPARTMENT OF JUSTICE _______________________________________________________
Introduction and Summary
SBC's application to offer long distance service in Missouri follows on the heels of its successful applications for Texas, Kansas, and Oklahoma. Many of the terms and conditions governing the availability of interconnection and unbundled network elements ("UNEs") match those that have already been approved in those applications. Moreover, the operational processes by which these wholesale inputs are made available to competitive entrants are alleged to be the same.
This Evaluation focuses on the prices at which SBC offers UNEs in Missouri.(1) The implementation of proper pricing is a key prerequisite to section 271 approval. Although the Department of Justice continues to rely upon the Commission for its ultimate determination of whether the prices supporting this application are appropriately cost-based,(2) we urge the Commission to undertake an independent scrutiny of the prices at issue rather than rely on the Missouri Public Service Commission's ("PSC") price-setting decisions. Prices in Missouri are higher than those in neighboring states which the Commission has found to be compliant with TELRIC, and this disparity does not appear to be accounted for by cost differences between states. Although the Missouri PSC appears to have focused on many forward-looking principles in its pricing docket, the record suggests that its actual decisions on several key questions of method and inputs may not comply with the Commission's forward-looking requirements. Taken together, these factors suggest that non-cost-based considerations may have resulted in prices outside the range that the reasonable application of TELRIC would produce. Moreover, competitive entry using UNEs to reach residential customers is almost nonexistent, suggesting that entry may have been impeded by above-cost rates.
In addition, the Department recognizes that there are serious concerns pertaining to SBC's resale of advanced services, namely, whether SBC is offering digital subscriber line ("DSL") services to end users without making those services available for resale at a wholesale discount. The Department urges the Commission to thoroughly investigate whether SBC is complying with its resale obligations.
To determine whether SBC has fully and irreversibly opened the local telecommunications market in Missouri to competition for both business and residential customers, the Department examines the three modes of entry contemplated by the Telecommunications Act of 1996 ("the 1996 Act")(3): facilities-based entry, which is least dependent on the incumbent's wholesale inputs and cooperation; use of the incumbent's UNEs; and resale of the incumbent's services.(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.(5) The degree to which such existing competition is broad-based determines the weight the Department places on it as evidence.(6)
For those entry modes where competitively significant entry is reasonably foreseeable but broad-based commercial entry is absent, the Department examines whether new technical and operational arrangements are available and working to support the entry mode, and whether performance benchmarks have been established to detect backsliding by the incumbent after long distance entry.(7) Small market shares held by competitors or even the absence of entry, standing alone, are neither conclusive evidence that a market remains closed to competition nor a basis for denying an application under section 271.(8)
SBC figures indicate that competitive local exchange carriers ("CLECs") serve at least 265,000 lines, or approximately 9 percent of the total lines, in SBC's Missouri service territory.(9) This aggregate level of CLEC penetration is approximately the same as those in New York, Texas, and Kansas at the time applications pertaining to those states were filed.(10) CLECs provide at least 22 percent of business lines in the SBC service territory in Missouri; only 3 percent of residential lines are provided by CLECs.(11)
CLECs provide service over their own facilities, in whole or in part, to approximately 4 percent of the lines in SBC's Missouri service territory ("facilities-based lines"). Such lines may constitute almost half of the lines served by CLECs in Missouri.(12) The great majority of facilities-based lines serve business customers; CLECs provide facilities-based local phone service for approximately 1 percent of residential lines.(13) AT&T provides the substantial majority of these lines using its own facilities, including its cable systems in Missouri; however, AT&T is planning on selling the Missouri cable systems over which it provides cable telephony services.(14) The development of cable telephony for residential customers is encouraging, but the extent to which it will be an economically feasible alternative for local phone service in the future is unknown.
Resellers also are active in Missouri.(15) Resale accounts for most CLEC service to residential customers in Missouri, although CLEC resale penetration is approximately 2 percent of residential lines.(16) CLEC resale reaches approximately 7 percent of business lines.(17)
CLECs have made limited use of the UNE-platform in Missouri, almost none to reach the residential market.(18) This use of the UNE-platform contrasts sharply with the New York and Texas markets, where its use accounts for rapid CLEC expansion into the residential market.(19)
DSL entry appears modest: there are approximately 4,500 CLEC DSL lines in Missouri.(20)
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.(21) There is significantly less competition to serve residential customers. There also is less competition by firms seeking to use UNEs, including the UNE-platform,(22) and there are some indications that a failure by SBC to satisfy all of its obligations may have constrained this type of competition.(23)
The Department has consistently stressed the importance of forward-looking cost-based pricing for the development of local competition.
Local telecommunications markets cannot be fully and irreversibly open to competition unless the prices for the interconnection and UNEs are properly based on costs. The FCC has established the basic principles that must be followed in establishing these prices, requiring that the prices "must be based on an incumbent LEC's forward-looking, long-run incremental costs for each network element." Prices which are not properly cost-based act as a barrier to entry; such prices may prevent entry entirely, or limit entry in type or scale.(24)
In a number of instances, the FCC has reasonably concluded that it will rely heavily on the pricing decisions made by the state commission. This deference, however, has its limits. The FCC may find that particular prices violate section 252 of the 1996 Act if "basic TELRIC principles are violated or the state commission makes clear errors in factual findings on matters so substantial that the end result falls outside the range that the reasonable application of TELRIC principles would produce."(25) Prices outside that range would preclude the approval of a section 271 application.(26) Ultimately, the FCC has the responsibility to make its own, independent finding regarding the propriety of the prices that are at issue in a section 271 application.(27)
Analysis of both the price-setting process and the level of the resulting prices is relevant to determining whether to subject a state's prices to further scrutiny.(28) The Department has previously noted that factors to consider may include a comparison of the prices at issue with those set in other states (as well as a comparison of the states' respective costs, to the extent that information is available),(29) the extent of entry pursuant to the prices in question,(30) and an examination of the state commission's record for significant departures from the prescribed TELRIC principles in determining costs and prices.(31) Not only must the state commission have endorsed forward-looking cost principles, but the state must also have properly selected forward-looking methodologies and inputs for the resulting rates to be judged TELRIC-compliant.(32)
In sum, the factors outlined above, when applied to the record in this application, suggest that the Missouri rates may fall outside the range that the reasonable application of TELRIC principles would produce. For this reason, the Department recommends that the Commission independently determine whether the prices conform to the requirements of the 1996 Act and the Commission's rules.
The Missouri PSC set permanent prices for many of the basic UNEs, such as loops, ports, and switches, in a July 1997 Order ("MPSC Final Pricing Order") in Docket No. 97-40, an SBC-AT&T arbitration docket.(33) The MPSC Final Pricing Order incorporated and attached a lengthy staff recommendation that reviewed both SBC's originally proposed cost studies and the modifications to those studies requested by staff.
The UNE recurring rates set in Docket No. 97-40 exceed by a significant margin those rates set in states in which SBC has already obtained section 271 approval.(34) The disparity is particularly striking for switch rates: prices in Missouri are higher than in other SBC-region states, and exceed switch prices for both Texas and Kansas by 22 to 60 percent.(35) Loop rates for Missouri also exceed those in other SWBT states, averaging approximately $3, or 20 percent, higher.(36)
Certain non-recurring rates ("NRCs") in Missouri, in particular those associated with analog line ports, also are significantly higher than those in other states.(37)
The Missouri PSC set interim prices for additional UNEs in a December 1997 Order ("MPSC Interim Pricing Order") in Docket No. 98-115, a follow-on SBC-AT&T arbitration. The MPSC Interim Pricing Order adopted SBC's proposed rates, without modification, as interim prices for some of the items in dispute, pending more thorough analysis and permanent resolution "no later than July 1, 1998."(38) These prices remain interim to date. Not until February 15, 2001, more than three years after its MPSC Interim Pricing Order, did the Missouri PSC open a new proceeding, Docket No. 2001-438, to set permanent rates for these and other UNEs.(39)
The rates set in Docket No. 98-115 exceed by a vast margin the rates for similar UNEs set in states in which SBC has already obtained section 271 approval. Missouri monthly recurring charges exceed Texas, Kansas, and Oklahoma rates by two to six times; Missouri NRCs exceed Texas, Kansas, and Oklahoma rates by two to thirteen times.(40)
Differences in prices between two states may legitimately arise from differences in costs between those states. In both its Kansas/Oklahoma Order and Massachusetts Order, the FCC used its Universal Service Fund ("USF") cost model as a "reasonable basis for comparing cost differences between states."(41) A comparison of USF costs for Missouri with those of Texas and Kansas, however, suggests that the difference in the tariffed prices described above exceeds any cost differences between the states. The comparison of Missouri and Kansas is particularly telling as these are adjacent states(42) with nearly identical costs, according to the USF model.(43) Despite this apparently close cost relationship, Missouri average loop rates exceed Kansas rates by 20 to 25 percent, and Missouri switch usage rates exceed those in Kansas by more than 50 percent.(44) This significant price differential, which is greater than the apparent cost differential, compels further scrutiny of the Missouri rates.
An analysis of the differences between Missouri and Texas or Missouri and Oklahoma prices also suggests that non-cost-based considerations may account for the differences in prices between the states: although the USF model lists Missouri costs as exceeding Texas costs, the tariffed UNE price differential between the two states is greater than this USF-suggested cost differential,(45) and although the USF model suggests that some Missouri rates should be lower than those in Oklahoma, these tariffed Missouri rates are higher than those in Oklahoma.(46)
While the Missouri PSC appears to have endorsed many of the appropriate principles in the record of its pricing docket, a review of its actual decisions on several key questions of method and inputs raise a number of serious questions as to compliance with forward-looking cost principles. It is difficult to assign any one of these possible errors a particular dollar effect,(47) and yet, in sum, they cast doubt on the compliance of the resulting rates with the forward-looking cost requirement.
Roughly speaking, the questionable decisions can be grouped as those which relate to the development of switch costs (e.g., switch discounts), those which relate to the development of loop costs (e.g., fill factors), and those which relate to the development of all UNE costs (e.g., depreciation and common cost allocation).
Switch discounts are a "key lever" in the proper development of forward-looking switch costs.(48) SBC's Missouri switch cost models did not fully reflect the switch discounts that it was receiving. Missouri staff required SBC to modify its models to reflect discounts for growth jobs, which "are typically less than the discounts for new switches,"(49) despite recognizing that this limited discount left SBC with a stated "investment per line [that] is still greater than that which staff believes is standard in the industry."(50) This "conservative"(51) decision appears not to have reflected forward-looking cost principles applied to an efficient firm.(52) Other possible errors that may have affected the Missouri switch prices include SBC's particular application of the "hardware factor,"an additive factor applied to SBC's switching investment which has a "substantial effect on switching costs."(53) Missouri staff expressed concern that the data to develop the factor may have been gathered from old technology (recognizing that this possibly overstates costs) and that maintenance port costs may have been double-counted, but staff resolved neither concern in the record.(54)
Fill factors are a key input in the proper development of forward-looking loop costs. SBC's Missouri loop cost models used its actual, historical fill factors as proxies for forward-looking fill factors.(55) Although Missouri staff recognized that "the use of actual fill factors is not forward-looking," the PSC only adjusted SBC's proposed distribution fill to 40 percent,(56) a level that the Commission specifically questioned in its Massachusetts Order.(57) Other possible errors that may have affected Missouri loop prices include the failure to allow for tapering of feeder cable,(58) and the allocation of all conduit costs to active, rather than dark, fiber.(59)
SBC's original cost models appear to have included aggressively short asset lives, which would have resulted in high depreciation costs. Although the Missouri PSC modified SBC's proposals, the asset lives chosen are still significantly shorter than those used (both by the Missouri PSC and other state commissions) in other proceedings.(60) States are not required to use the FCC's depreciation rates in setting TELRIC prices, but they must use "economic," i.e., forward-looking, depreciation rates.(61) Missouri staff appeared to acknowledge this requirement, but its stated goal was less clear: "to recommend depreciation rates based on parameters that SWBT is likely to experience for financial purposes so as to fully recover its long run capital costs in a timely fashion and be fair to customers."(62) Staff constructed its own benchmark range by figuring the depreciation rates based on the financial statements for nineteen different companies(63); staff concluded that, as adjusted, SBC's depreciation rates fell within the low end of the calculated benchmark range, so were reasonable.(64) The record does not reflect any determination of whether the benchmark range of rates was consistent with forward-looking principles or was a reflection of financial depreciation rates.(65)
SBC's original cost models included a common cost allocator of approximately 16 percent, which is significantly higher than the allocators used in Texas (13 percent) and Kansas (10 percent).(66) Missouri staff did not recommend any changes specific to SBC's common cost model, and therefore discussed the model little in its recommendation.(67) This seemingly high figure may be accounted for by SBC's reliance on historical overhead costs as its starting point for calculating the amount of common costs that need to be allocated,(68) or SBC's particular method of calculating the percentage allocator(69) (rather than any legitimate forward-looking cost concern), or both.
It is not possible to pinpoint TELRIC errors in Docket No. 98-115 because there is no detailed Missouri PSC order discussing the cost models and inputs that produced those interim rates. The MPSC Interim Pricing Order issued in December 1997 adopted SBC's proposed rates as interim rates, subject to true up, without discussing their specific cost basis. It appears that the models used to generate the rates set in Docket No. 98-115 were the same as those SBC had originally proposed for use in Docket No. 97-40, and did not include the modifications that were required by Missouri staff in that docket.(70) This blanket adoption of SBC's proposed rates suggests that not only are the rates set in Docket No. 98-115 possibly tainted by the TELRIC errors, as discussed supra, but that they may be even more significantly flawed.
While the existence of interim rates in and of itself has not led the Commission to deny any section 271 application,(71) the Commission should carefully consider the adequacy of interim rates to support the present application. SBC's Missouri section 271 application is predicated upon a large number of rates that have remained interim for a long time. The interim rates set in Docket No. 98-115 are troublingly high and have been left as interim for years, despite concerns having been raised that the rates were not forward-looking.(72)
In addition to the UNE rates set on an interim basis in Docket No. 98-115, the Missouri PSC has yet to finalize rates for collocation as well as for certain other UNEs that had not previously been requested in Missouri. While the current level of these rates (having been borrowed from Texas for the interim) does not appear problematic, the continued uncertainty of so many rates remaining interim, coupled with doubts about pricing discussed supra, gives rise to doubts that the market is open to competition by firms that seek to use these elements.
The Department recognizes that there are serious concerns pertaining to SBC's resale of advanced services, namely, whether SBC is offering DSL services to end users without making those services available for resale at a wholesale discount.(73) If true, this refusal would raise an issue with regard to SBC's compliance with section 251(c)(4) of the 1996 Act as interpreted in Association of Communications Enterprises v. FCC.(74) Given the current state of the record, the Department is not in a position to make such a determination. The Department urges the Commission to thoroughly investigate whether SBC is complying with its resale obligations.
The Commission should independently determine whether the UNE prices in Missouri conform to the requirements of the 1996 Act and the Commission's rules, rather than rely on the decision of the Missouri PSC approving these prices. In addition, the Commission should thoroughly investigate whether SBC is complying with its resale obligations pertaining to advanced services.
Certificate of Service
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 May 9, 2001.
1. The Department expresses no view as to SBC's compliance with checklist requirements that are not specifically addressed in this Evaluation.
2. In part for this reason, the Department has not attempted to review in this Evaluation every pricing criticism raised by commenters in this docket. The Department's omission of any particular pricing issue should not be read as support for or rejection of the position of any particular commenter.
3. Pub. L. No. 104-104, 110 Stat. 56 (1996) (codified as amended in scattered sections of 47 U.S.C.).
4. See DOJ Schwartz Aff. ¶¶ 149-192; DOJ Schwartz Suppl. Aff. ¶¶ 26-60; DOJ Oklahoma I Evaluation at vi-vii, 36-51.
5. See, e.g., DOJ Oklahoma I Evaluation at vi-vii, 41-42.
6. See, e.g., DOJ Schwartz Aff. ¶ 176.
7. See, e.g., DOJ Oklahoma I Evaluation at 48-51.
8. See, e.g., id. at 29-30; DOJ Louisiana II Evaluation at 26-27.
9. See SBC Tebeau Aff. at 3 tbl.2. SBC serves 2.6 million access lines in Missouri. Id. at 14 tbl.5. Estimated market share will vary depending on the methodology used to estimate facilities-based lines. The Department relied on entries in the E-911 data base. SBC, using multiple methodologies, estimates that CLECs serve between 9.2 and 14.2 percent of Missouri access lines. SBC Brief at iii. The Missouri PSC Order approving SBC's section 271 application concluded that CLECs serve approximately 12 percent of access lines. MPSC Final 271 Order at 20.
10. In New York, CLECs served approximately 8.9 percent of total access lines. DOJ New York Evaluation at 9. The Department estimated that CLECs served approximately 8 percent of the market in Texas and 9 percent in Kansas. DOJ Texas I Evaluation at 8-9; DOJ Kansas/Oklahoma Evaluation at 7.
11. See SBC Tebeau Aff. Attachs. C-1, C-2 & C-5 (CLEC lines are 77 percent business and 23 percent residential.). The Department's calculations are based on these attachments and FCC statistics on the proportions of total switched access lines comprised by business and residential customers. FCC Common Carrier Statistics at 22 tbl.2.4.
12. See SBC Tebeau Aff. ¶ 15. Counting such lines by E-911 entries, CLECs in SBC's Missouri service territory serve 119,460 lines using their own facilities.Id. The number of such facilities-based lines includes lines served by stand-alone UNE-loops and thus is not comprised entirely of "pure" facilities-based lines.
13. See SBC Tebeau Aff. Attach. C-1 (At least 10 percent of all business lines in SBC's Missouri territory are served by CLECs using at least some of their own facilities; only 1.2 percent of residential lines are so served.); see supra note 11.
14. See id.; AT&T Comments at 56-57. AT&T has agreed to sell its cable systems in the St. Louis area to Charter Communications, Inc., and notes that Charter's web-site "does not even mention" local telephone service as one of the company's products. AT&T Comments at 56-57. Thus, according to AT&T "the prospect of any future facilities-based competition for residential service in Missouri is, at best, questionable." Id. at 57.
15. See SBC Tebeau Aff. at 15 tbl.6 (97,851 CLEC lines are resale lines, which constitute approximately 3.5 percent of total lines in SBC's Missouri service territory.); see supra note 9.
16. See SBC Tebeau Aff. at 15 tbl.6 (35,488 of residential lines in SBC's service territory are served by CLECs using resale.); see supra note 11.
17. See SBC Tebeau Aff. at 15 tbl.6 (62,363 of business lines in SBC's service territory are served by CLECs using resale.); see supra note 11.
18. Approximately 47,000 UNE-platform lines are in use in Missouri, SBC Tebeau Aff. ¶ 15, approximately 1,200 of which serve residential customers, SBC Tebeau Aff. Attach. C-2. These lines constitute approximately 1.5 percent of total lines, and less than one-tenth of 1 percent of residential lines, in the SBC Missouri service territory. FCC Common Carrier Statistics at 22 tbl.4.
19. At the time of the New York application, approximately 152,000 lines were served through the UNE-platform. FCC New York Order ¶ 14. By February 2001, CLECs in New York served approximately 1.5 million additional customers over the UNE-platform. See Verizon Connecticut Brief Attach. A.4. At the time of SBC's second Texas application, approximately 244,000 lines were served through the UNE-platform. FCC Texas Order ¶ 5. By September 2000, CLECs in Texas served 569,000 customers over the UNE-platform. SBC Kansas/Oklahoma Smith/Johnson Aff. at 19 tbl.10.
20. See SBC Missouri Performance Data Ex Parte at 65c (PMs 65-08, 65-09) (Mar. 2001).
21. However, in addition to the pricing and performance issues affecting UNEs, see infra text Part II and notes 22 & 23,competitive entry may have been constrained by SBC's refusal to allow CLECs to participate in Missouri's Metropolitan Calling Area Plan ("MCA"). See generally McLeod Comments at 3-13. Several of the facilities-based CLECs (including those using UNE-loops, although not providers using solely the UNE-platform) complained to the Missouri PSC that SBC's unilateral re-rating of the CLECs' NXX codes to exclude the CLECs' end-user customers from the MCA required people calling those CLEC customers to dial extra digits and pay toll charges. See id. at 4-5. The Missouri PSC took more than a year to act on the CLECs' complaint, eventually ruling last October that CLECs were proper MCA participants. See id. at 9-10. Meanwhile, at least one CLEC, McLeod, claims to have delayed its launch of facilities-based service in Missouri pending final resolution of this matter. Id. at 6-7.
22. The allegation that SBC has failed to provide nondiscriminatory access to its maintenance and repair functions due to the failure to correctly update its records of resale and UNE-P circuit ownership, see AT&T Comments at 44-47, deserves careful attention. This problem may affect SBC's ability to provide CLECs with parity performance and may call into question the reliability of some of SBC's reported performance measures. The Department, however, has not been able to determine the scope or competitive impact of the problem because it was not addressed in SBC's Application. In particular, although SBC has apparently implemented a means of correcting new records, it has agreed to correct old records (those pertaining to orders processed before March 2001) only on a case-by-case, manual basis. See id. at 45-46.
23. SBC's performance in provisioning high-capacity (DS-1) loops has been poor on certain measures and, in some cases, worse for CLECs than for its retail customers. Despite some improvement, during the past three months SBC has missed, on average, nearly one-quarter of installation commitments for CLECs. See SBC Missouri Performance Data Ex Parte at 58b (PM 58-06). In March, average delay days caused by such missed due dates for CLECs were four times as long as those for SBC's retail customers. Id. at 62b (PM 62-06). Finally, the rate of repeat trouble reports for CLECs has deteriorated from 7 percent in January, to 14.6 percent in February, to 21.6 percent in March. Id. at 69b (PM 69-05). By contrast, the repeat trouble report rate for SBC's retail customers in March was only 8.3 percent. Id.
Poor performance in this area is particularly troubling given the unique attributes of high-capacity loops, which are key inputs for CLECs competing for business customers. See NuVox Comments at 13 (DS-1 service is superior alternative for serving business customers and therefore at core of business plan); cf. FCC UNE Remand Order ¶ 184 (recognizing commenters' statements noting "the call concentration and revenue potential of 'high-capacity' lines (DS[-]1 and higher)"). Thus, the fact that DS-1 loops account for a small percentage of orders for all loop types may understate their competitive significance. The Commission had previously warned that it "will be actively monitoring SWBT's performance in this area and . . . will take swift and appropriate enforcement action in the event SWBT's provisioning performance for high capacity loops fails to improve." FCC Kansas/Oklahoma Order ¶ 213; cf. FCC Massachusetts Order ¶ 156 (recognizing Verizon's "poor" performance in provisioning high-capacity loops on certain measures). SBC's DS-1 performance in Missouri may be no worse than that described in the FCC Kansas/Oklahoma Order, yet the lack of significant improvement should occasion attention by the Commission.
24. DOJ Kansas/Oklahoma Evaluation at 10 (citation omitted) (emphasis added).
25. FCC New York Order ¶ 244.
26. DOJ Kansas/Oklahoma Evaluation at 11.
27. Id. (citing 47 U.S.C.A. § 271(d)(3) (West Supp. 1999)).
28. See FCC Massachusetts Order ¶¶ 38-40 (analyzing both listed inputs used to set Massachusetts loop rates as well as the level of the resulting rates for compliance with TELRIC).
29. SBC contends that such cross-state comparisons are invalid, except in cases where the state in question has failed to apply TELRIC principles. SBC Brief at 36. SBC claims that as Missouri did set TELRIC prices, comparing the resulting Missouri rates to those from other states for which it has obtained section 271 approval, is inappropriate. Id. (noting that the FCC compared Oklahoma UNE rates to Texas UNE rates only after finding TELRIC errors in the Oklahoma record); SBC DOJ Questions Ex Parte at Question 2, USF Chart. But see infra text Part II.C. (possible TELRIC errors in the Missouri record). Although the Department agrees that "differences in prices from one state to another do not necessarily indicate that the prices in either state are not appropriately cost based, . . . . [i]n the absence of persuasive evidence of differences in costs between states, substantial differences in prices should trigger more careful scrutiny by the Commission." DOJ Kansas/Oklahoma Evaluation at 12. The FCC appears to have endorsed this view in its recent Massachusetts Order, where it evaluated the adequacy of Massachusetts loop rates by comparing them to New York's previously approved loop rates, using data from its Universal Service Fund ("USF") model as a proxy for legitimate cost differences between the two states. See FCC Massachusetts Order ¶¶ 32, 40.
30. See supra text Part I. "As in the case of price comparisons between states, this factor does not necessarily indicate whether prices are or are not cost-based, because the level of demand may reflect factors other than price. Nonetheless this factor may be a useful indicator of whether closer scrutiny of prices by the Commission is appropriate." DOJ Oklahoma/Kansas Evaluation at 12; see also FCC Massachusetts Order ¶ 42 ("Even if competitors can gain 'efficient entry' to a market through the availability of TELRIC-based UNE rates, they may still decide not to enter based on their independent determinations that they cannot turn a sufficient profit in the market."). See also supra note 21 (regarding MCA issue).
31. DOJ Kansas/Oklahoma Evaluation at 11-13; see also FCC Massachusetts Order ¶¶ 39-40 (questioning the propriety of the low fill factor used by the Massachusetts Department of Telecommunications and Energy but concluding that "any errors made" did not result in loop "rates outside the range that a reasonable application of TELRIC principles would produce," based on an evaluation of the resulting Massachusetts loop rates as compared to New York loop rates).
32. FCC Massachusetts Order ¶ 20.
33. MPSC Final Pricing Order at 4-5.
34. In both its Kansas/Oklahoma and Massachusetts orders, the FCC compared rates between a state at issue and a state whose rates had previously been approved in a section 271 proceeding (i.e., Oklahoma/Texas and Massachusetts/New York, respectively), and explicitly granted a presumption of TELRIC compliance if a state adopted a set of previously approved rates "and could demonstrate that its costs were at or above the costs in that state whose rates it adopted." FCC Kansas/Oklahoma Order ¶ 82 & n.244; see also FCC Massachusetts Order
¶¶ 22-27. The FCC's analysis suggests that comparison to such a safe harbor or benchmark is particularly appropriate when the states are adjoining and have the same rate structures. See FCC Massachusetts Order ¶ 28.
35. AT&T Baranowski Decl. at 22 tbl.2 (comparison chart showing Missouri switch usage rates 60.8 percent higher than Texas rates and 45.9 percent higher than Kansas rates, although only 3 percent higher than Oklahoma rates); WorldCom Fentrup Decl. ¶ 7 ("The switch usage rate in Missouri is nearly 50 percent greater than the same rate in Kansas and Texas."); Z-Tel Walters Statement at 2 (comparison chart representing Z-Tel's typical payments to SBC for switching, usage, and port, and showing Missouri rates 22 percent higher than Texas rates and 48 percent higher than Kansas rates); see also SBC DOJ Questions Ex Parte at Question 1, UNE-P Chart (showing Missouri weighted average local switching MOU rate more than 50 percent higher than Kansas or Texas average rates, although slightly lower than that in Oklahoma).
36. AT&T Lieberman Decl. ¶ 22 & tbl.2 (statewide average loop rates $17.15 in Missouri, as compared to $13.76 in Kansas, $14.33 in Texas, and $15.87 in Oklahoma); WorldCom Fentrup Decl. at 4 n.3 (statewide average loop rates are $17.40 in Missouri, as compared to $14.04 in Kansas, $14.15 in Texas, and $14.84 in Oklahoma); see also SBC DOJ Questions Ex Parte at Question 1, UNE-P Chart (showing Missouri weighted average loop rate 26 percent higher than Kansas rate, 19 percent higher than Texas rate, and 7 percent higher than Oklahoma rate).
37. SBC Price Comparison Charts Ex Parte at 4 (Non-Recurring Charge Comparison Chart) (Missouri analog line port NRC of $29.53, as compared to Texas NRC of $1.27 and Oklahoma NRC of $1.20 (Kansas listed as N/A)). NRCs "would be expected to vary minimally from state to state" within a particular region. DOJ Kansas/Oklahoma Evaluation at 16 (citation omitted). Recognizing that "in most instances, Missouri NRCs were substantially more than Texas NRCs," the Missouri PSC as part of its section 271 review required SBC to reduce the NRCs by an additional 25 percent, "but not to a level below the corresponding NRC found in the Texas agreement." MPSC Final 271 Order at 33.
38. MPSC Interim Pricing Order at 52. SBC's proposed rates were adopted as interim subject to true up for multiplexing, digital cross-connect, dedicated transport cross-connect, NXX migration, and additional elements, as well as loop cross-connect to DSC, subloop cross-connect, access to directory assistance database, dark fiber cross-connect, and dark fiber record research. See id. at 19-20, 24-25. Other rates (including NRCs for UNE-platform simple migrations and for vertical features in the switch) were set at zero, pending further cost analysis by the Missouri PSC. Id. at 21-22, 24-25. Although the MPSC Interim Pricing Order did not appear to limit the true-up period, the Missouri PSC's final order approving the M2A and SBC's section 271 application determined that "the interim rates contained in the M2A are subject to a limited true-up ... a true-up period that is six months retrospectively from the date of the Commission's order establishing a permanent rate is appropriate." MPSC Final 271 Order at 35.
39. MPSC UNE Pricing Proceeding Order at 2-3.
40. NuVox Comments at 3-4; SBC Price Comparison Charts Ex Parte at 8-9 (Non-Recurring Charge Comparison Chart) (e.g., DS-1 Entrance Facility NRC of $471(first)/$342 (each additional) in Missouri as compared to $73.25/$26.68 in Texas, $165.86/$65.78 in Kansas, and $214.36/$84.56 in Oklahoma). These disparities between Missouri and Texas prices are as striking as those between the original 1998 Oklahoma and Texas or Kansas prices: Oklahoma recurring rates were as much as twice those comparable in Texas or Kansas; Oklahoma NRCs were two to thirteen times those comparable in Texas. DOJ Kansas/Oklahoma Evaluation at 15-16.
41. FCC Massachusetts Order ¶ 22.
42. See id. ¶ 28.
43. AT&T Baranowski Decl. at 22 tbl.2 (showing USF switch usage costs 2.6 percent higher in Kansas than in Missouri, and 20 percent lower in Texas than in Missouri ); AT&T Lieberman Decl. ¶ 22 & tbl.2 (showing 0 percent difference between USF loop costs for Kansas and Missouri, and costs 14 percent higher in Missouri than in Texas ); WorldCom Fentrup Decl. ¶ 9 (Regarding switch costs, "[t]he cost relationships from the [FCC's Synthesis Model] suggest that the Kansas and Missouri costs are nearly identical, and less than 20 percent above the Texas cost."); id. ¶ 11 (Regarding loop costs, the "model indicates that loop costs in Kansas and Missouri are almost identical, both being about 14 percent above the cost in Texas."); SBC DOJ Question Ex Parte at Question 2, USF Chart (showing 0 percent difference between USF loop costs between Kansas and Missouri, and costs 14 percent higher in Missouri than in Texas).
44. See supra notes 35-36.
45. AT&T Baranowski Decl. at 22 tbl.2 (showing USF switch usage costs 20 percent higher in Missouri than in Texas while tariffed switch usage costs are 60 percent higher in Missouri than in Texas); AT&T Lieberman Decl. ¶ 22 & tbl.2 (showing USF average loop costs 14 percent higher in Missouri than in Texas while tariffed average loop rates are 20 percent higher in Missouri than in Texas); WorldCom Fentrup Decl. ¶¶ 8-11 (showing USF switch usage rate about 20 percent higher in Missouri than in Texas while the tariffed switch usage rate is about 50 percent higher in Missouri than in Texas; as well as a USF average loop cost about 14 percent higher in Missouri than in Texas while the tariffed average loop cost is about 20 percent higher in Missouri than in Texas); see also SBC DOJ Questions Ex Parte at Question 1, UNE-P Chart, and Question 2, USF Chart (showing USF switch usage costs 21 percent higher in Missouri than in Texas while tariffed average switch usage rates are 57 percent higher in Missouri than in Texas, and USF loop costs 14 percent higher in Missouri than in Texas while tariffed average loop rates are 19 percent higher in Missouri than in Texas).
46. AT&T Baranowski Decl. at 22 tbl.2 (showing USF switch usage costs 5 percent higher in Missouri than in Oklahoma, while tariffed switch usage costs are 3 percent higher in Missouri than in Oklahoma); AT&T Lieberman Decl. ¶ 22 & tbl.2 (showing USF average loop costs 7 percent lower in Missouri than in Oklahoma, while tariffed average loop rates are 8 percent higher in Missouri than in Oklahoma); WorldCom Fentrup Decl. ¶ 11 (USF average loop costs are about 6 percent higher in Oklahoma than in Missouri, while tariffed average loop costs are about 20 percent higher in Missouri than in Oklahoma); see also SBC DOJ Questions Ex Parte at Question 1, UNE-P Chart, and Question 2, USF Chart (showing USF switch usage costs 5 percent higher in Missouri than in Oklahoma while tariffed average switch usage rates are 15 percent less in Missouri than in Oklahoma, and USF loop costs are 7 percent less in Missouri than in Oklahoma while tariffed average loop rates are 7 percent higher in Missouri than in Oklahoma).
47. This difficulty appears to stem, at least in part, from SBC's failure to make electronic versions of its cost models available for review and rerunning by interested parties. See AT&T Baranowski Decl. ¶ 3 ("SWBT apparently claims that electronic versions of its cost studies no longer exist. For this reason, my efforts to evaluate SWBT's cost studies and to quantify the impact of certain of its TELRIC violations have been considerably hampered."); WorldCom Fentrup Decl. ¶ 12 ("Without these cost models, WorldCom cannot quantify the effect of these issues on rates."). SBC filed additional cost model spreadsheets with the FCC the day before third-party comments were due in this proceeding. SBC Cost Model Ex Parte (Apr. 23, 2001).
48. MPSC Final Pricing Order at 31.
49. Id. at 32.
50. Id.; see also id. at 53 ("Staff believes that SWBT's reported discount for SCP equipment may be less than the discounts actually received. Based on information discovered while attempting to determine the SCIS/MO discounts, Staff has reason to suspect that SWBT may be receiving additional discounts. Staff does not have data to propose an alternative discount. . . . Staff notes that discount levels were not verified and could very well be incorrect.").
51. Id. at 32.
52. See FCC Massachusetts Order ¶ 251 (expressing concerns that repetition of some of the assumptions, including low switch discounts, incorporated into the original Massachusetts switching prices may result in rates outside the range that reasonable application of TELRIC principles would produce).
53. MPSC Final Pricing Order at 43.
54. Id. at 42-43.
55. See id. at 13, 23. SBC's use of actual fill as a substitute for forward-looking fill factors was specifically rejected in the FCC Kansas/Oklahoma Order, which determines that SBC's 30 percent fill factor used in Oklahoma was too low to be considered forward-looking. FCC Kansas/Oklahoma Order ¶ 80.
56. MPSC Final Pricing Order at 13-14.
57. FCC Massachusetts Order ¶¶ 39-40 (expressing concern about Verizon's 40 percent distribution fill factor in the absence of any state-specific justification, as compared to the 53 percent distribution fill factor adopted by Kansas and a 50 percent distribution fill factor adopted by New York).
58. SBC's Missouri loop cost model did not reflect cable tapering, whereby "a feeder segment may originate as a very large cable and taper as the cable terminates to multiple [feeder distribution interfaces]." MPSC Final Pricing Order at 18. Although staff recognized that this assumption "increase[s] the cost of the feeder segment," SBC maintained that it could not modify its cost model to revise this assumption. Id. Staff listed the lack of cable tapering as a concern that could not be addressed at that time, adding "[i]t is important to remember that SWBT's assumption of a single feeder cable terminating to an FDI will overstate the cable costs and overstate the cost of the loop." Id. at 18-19.
59. SBC's Missouri loop cost model assigned all of the cost of conduit to the active fiber within it. Missouri staff verified that SBC's dark fiber cost model did not include any conduit cost "so the issue of double recovery does not apply." Id. at 18. AT&T points out that this allocates the entire cost of conduit to the active (or "lit") fiber rather than between active and dark fiber (which is a separate UNE, separately priced). AT&T Baranowski Decl. ¶ 35.
60. AT&T Baranowski Decl. ¶ 20 & tbl.1 (showing depreciation lives used by Missouri PSC in setting UNE rates "as little as 1/2" of those used in other Missouri rate-making proceedings and in other section 271 approved states).
61. FCC Kansas/Oklahoma Order ¶ 76 (rejecting AT&T's challenge to the Oklahoma depreciation rates, based on the conclusion that although the Oklahoma Administrative Law Judge did not explicitly adjust SBC's proposed lives, he found that issues regarding the propriety of those lives were "amply addressed within the stipulation results which reduce recurring costs").
62. MPSC Final Pricing Order at 99. Missouri staff rejected the use of the FCC depreciation ranges because of concern that they did not reflect "true plant mortality experience," but rather were the result of "expediency, sometimes involving compromise." Id. at 99-102.
63. Id. at 102.
64. Id. at 104.
65. Id. at 102-04. In fact, Missouri staff noted that "[t]he resulting range of implied [depreciation] rates is puzzling and begs the reviewer to search for an answer. Unfortunately, no actual explanation is available other than to state that for the most part, each company chooses its own depreciation rates for the particular type of assets in the particular market and the industry it is in. There is no requirement to report details of how depreciation is calculated." Id. at 103-04.
66. AT&T Baranowski Decl. ¶¶ 22, 25; see also SBC Smith Aff. ¶¶ 23-27 (describing development of common cost factor).
67. MPSC Final Pricing Order at 125-26 ("Staff has no specific concerns or proposed modifications to this study other than Staff's proposed modifications affecting all studies (Cost of Money, Depreciation, etc.).").
68. SBC Smith Aff. ¶¶ 23-24 (SBC used 1995 actual overhead costs, less certain retail, executive, planning, general, and administrative expenses, plus new general network supervision and wholesaling expenses, to calculate the total amount of common costs that needed to be allocated.); see also AT&T Baranowski Decl. ¶ 23 (SBC's 16.47 percent common cost factor "is based entirely on SWBT's pre-1996 Act monopoly level of common costs and is, therefore, not reflective of the forward-looking common costs that an efficient provider would incur. SWBT has since conceded that it has become more efficient.").
69. AT&T Baranowski Decl. ¶ 24 (alleging a "fundamental mismatch" between the numerator and denominator SBC used to calculate the per-unit percentage markup common cost allocator).
70. With regard to NRCs in Docket No. 97-40, for example, Missouri staff had expressed strong concern that SBC's proposed NRCs may have been based on unreliable labor estimates and may have reflected double-recovery of an amount of labor costs, and had attempted to remedy these flaws by reducing SBC's proposed NRCs by 50 percent. MPSC Final Pricing Order at 121, 123-24. Eight months after SBC's prices had been adopted by the Missouri PSC as interim rates for Docket No. 98-115, Missouri staff apparently repeated the recommendation to reduce by 50 percent the NRCs generated by SBC's cost model in Docket No. 98-115. See SBC Comments on Costing and Pricing Rep. at 1, 3; see also id. Supp'g Affs.
71. The Commission has accepted interim rates where the interim solution to a particular rate dispute is reasonable and the state commission has demonstrated a commitment to proper pricing rules. Provision for true ups once permanent rates are set may give additional comfort regarding interim rates. See FCC Kansas/Oklahoma Order ¶ 238; FCC Massachusetts Order ¶ 34.
72. The Missouri PSC apparently decided to leave these rates in place still longer because they already "have been used by SBC and some of those [competitive] carriers for a substantial period of time." NuVox Comments at 8. This determination may not be an adequate proxy for a forward-looking cost analysis, even on an interim basis.
73. See generally AT&T Comments at 32-38.
74. 235 F.3d 662 (D.C. Cir. 2001) (An incumbent local exchange provider cannot avoid obligations pursuant to section 251(c) of the Telecommunications Act of 1996 with respect to resale of advanced services by providing them through a subsidiary.).