X. Comments on Proposed and Alternate Decision

In accordance with Pub. Util. Code § 311(d) and Rule 77.1 of the Rules of Practice and Procedure, the Commission mailed the proposed decision of the ALJ in the proceeding. Comments were filed by AT&T, California Association of Competitive Telecommunications Companies (CALTEL), MCI/WorldCom, Mpower Communications Corporation (Mpower), Navigator Telecommunications LLC (Navigator), ORA/TURN, SBC-CA, Vycera Communications Inc. (Vycera), and jointly by XO and Allegiance Telecom of California, Inc. (Allegiance).⁹⁴ Reply comments were filed by AT&T, CWA, DOD/FEA, MCI/WorldCom,Mpower, ORA/TURN, SBC-CA, and XO/Allegiance.

The comments cover myriad technical details and suggest corrections to the Commission's UNE modeling. Detailed technical comments are addressed in the section of the order pertinent to the specific topic. In Section V.D, we list the specific corrections made to the Commission's model runs of both HM 5.3 and the SBC-CA models in response to comments. Where comments merely reargued earlier positions, they are not discussed.

In this section, we discuss the more general comments on a few major themes that are repeated throughout the parties' various comments. These themes are as follows:

· Selecting rates based on the midpoint of HM 5.3 and the SBC-CA models is arbitrary and capricious

· The Commission should use only HM 5.3 to set UNE rates

· The Commission should use only the SBC-CA models to set rates

· Corrections to the SBC-CA models were ignored

· The true-up of interim rates to permanent rates is too large and causes competitive harm

· The new rates create a price squeeze

We address these key comment issues briefly below. The revised proposed decision addressing these comments was mailed to the parties to allow them an additional opportunity to comment, based on the substantial changes from the original proposed decision.

A. Rates Based on the Midpoint of HM 5.3 and SBC-CA Models

Several parties dispute the Commission's initial plan to adopt UNE rates derived from the midpoint of the results of both models. MCI/WorldCom claims this approach is arbitrary and capricious because the record does not support rejection of HM 5.3 and fails to articulate a connection between the cost models' flaws and the decision to weight them on a 50/50 basis. MCI/WorldCom suggests the Commission use the HM 5.3 model entirely or give it more weight.

XO, Navigator, CALTEL, and Mpower join MCI/WorldCom in suggesting that the Commission should use HM 5.3 alone rather than adopt rates based on the midpoint. They suggest that using the midpoint is not justified if one model produces results that are too high. ORA/TURN maintain the Commission errs in setting rates using the LoopCAT model because it does not comply with TELRIC and the flaws in HM 5.3 pale in comparison to the fatal defects inherent in LoopCAT. The main criticism of HM 5.3 relates to a perceived flaw with the clustering process, which was demonstrated to have a negligible impact. HM 5.3 is TELRIC compliant and should be used, rather than the SBC-CA models, to set UNE rates.

SBC responds that the Commission's 50/50 weighting of HM 5.3 and the SBC-CA models is lawful. According to SBC-CA, the Commission can split the difference between competing options if the decision is rationally articulated and supported by record evidence. SBC-CA disputes comments that the flaws in the SBC-CA models are more serious than those found in HM 5.3

The comments alone do not convince us to abandon the approach of adopting rates based on the midpoint of both models. We agree with SBC-CA that adopting the midpoint of two models is a supportable outcome if rationally articulated and supported by the record. MCI/WorldCom, XO and Navigator imply the Proposed Decision applied no judgment to the models before "splitting the baby." In our view, the initial Proposed Decision exercised a great deal of judgment in reviewing the models' flaws, correcting them where possible, and selecting numerous modeling inputs.

Nevertheless, our own attempts to work with the SBC-CA models and correct the errors we made in the Proposed Decision's modeling runs now convince us that the SBC-CA models are not only flawed, but also fail our modeling criteria to such a significant extent that we cannot reasonably rely on them to set UNE rates. In other words, we would have maintained the approach of adopting the midpoint of model results, despite the comments, except for the fact that additional flaws and modeling difficulties arose with the SBC-CA models as we corrected our earlier runs. Our decision to abandon use of the SBC-CA models is discussed more fully in Section V.D.

B. Adopt Rates Based on the HM 5.3 Model

MCI/WorldCom argues that the Proposed Decision errs in rejecting HM 5.3 because the record does not support the finding that HM 5.3 understates forward-looking costs. MCI/WorldCom states that the three primary flaws identified in HM 5.3 -- related to transport modeling, the customer location database, and labor rates - do not lead to understated rates as the PD suggests. Rather, MCI/WorldCom maintains that transport modeling and customer location issues would only lower rates if corrected. Further, MCI/WorldCom suggests labor costs can be modified through an outboard adjustment that it includes in its comments.

We disagree with MCI/WorldCom's comments that the record proves HM 5.3 does not underestimate rates. We have modified the decision to indicate that it is unclear of the effect on rates in HM 5.3 if the transport and customer location flaws we identified could be corrected. Further, we clarified the decision regarding the labor rates that we could not change in HM 5.3. Even though we do not adopt the outboard labor adjustments suggested by MCI/WorldCom, we analyzed the effect of speculative labor changes and found they were minimal and most likely counterbalanced by conservative inputs in other areas of the model.

ORA/TURN, Navigator, CALTEL and XO also urge the Commission to adopt HM 5.3 rather than adopting the midpoint of the two models. With regard to customer location flaws in HM 5.3, ORA/TURN states that "a perceived flaw that has been demonstrated to have a negligible impact on costs should not be sufficient grounds to reject the model." (ORA/TURN, 6/1/04, p. 6.) In the final decision, we have set rates based solely on HM 5.3, but not because the comments above urged us to do so. Rather, we select HM 5.3 because after correcting what we agree are errors in the Proposed Decision, we find the SBC-CA models fail our modeling criteria to such a significant extent that we cannot reasonably rely on them.

C. Adopt Rates Based on the SBC-CA Models

SBC-CA presents several arguments in its comments that its models should be used exclusively to set UNE rates rather than HM 5.3. First, SBC-CA argues that the Proposed Decision relies on a legally impermissible interpretation of TELRIC that prohibits consideration of SBC-CA's actual costs, requires a hypothetical network that is instantly built, and prohibits modeling based on ultimate demand. We find that SBC-CA, for the most part, reiterates its earlier arguments on these topics and we will not repeat our discussion here. Nevertheless, we note that the modeling runs supporting the UNE rates adopted in this decision do use SBC-CA's actual costs or suggested modeling inputs in several areas. Further, we have described how SBC-CA's models, for the most part, merely calculate the costs to build the network it has in place today rather than the costs of a reconstructed network.

Second, SBC-CA argues that the Proposed Decision erroneously concluded that the SBC-CA models could not be modified in several areas, including expenses and cost factors, the design point, and cabling inventories. We address these suggested corrections in the body of the decision. In short, we found that most of the changes suggested by SBC-CA could not reasonably be made, such as to the design point and cabling inventory. In the few areas where we agreed with SBC-CA and attempted to make changes to cost factors and expenses in the SBC-CA models, the result was our finding the SBC-CA models fail our modeling criteria to such a significant extent that we cannot reasonably rely on them.

Third, SBC-CA comments that customer location and clustering problems in HM 5.3 outweigh problems in the SBC-CA models. We disagree, particularly given our experience as described in Section V.D when we attempted to modify the SBC-CA models. SBC-CA again reargues its objections to the HM 5.3 customer location process which are addressed at length in Section V.B.2 of this decision.

D. Corrections Were Ignored

MCI/WorldCom, SBC-CA, ORA/TURN, and XO claim that the Commission ignored suggestions to correct various portions of the SBC-CA models, and may have never even reviewed the parties' original filings explaining these corrections. Specifically, parties ask the Commission to reconsider corrections in areas such as the modeling of multiple dwelling units, factor models, expenses related to shared and common costs, affiliate transactions, and unregulated businesses.

We address these comments in the body of the decision. The decision explains that the suggested corrections, or "restatements," were reviewed but they were lengthy, unclear and often unsupported, and disputed. These restatements of the SBC-CA models could not be accepted without substantial further review that was not reasonable to undertake. Instead, the Commission's analysis focused on what it considered key flaws and modeling inputs rather than all of the areas outlined by the parties. In a few limited areas, we did attempt to apply these additional corrections, particularly with regard to expenses in the SBC-CA models. Ultimately, these suggested changes to the SBC-CA models became moot when we abandoned use of the SBC-CA models to set UNE rates.

MCI/WorldCom and SBC-CA also suggest modifications to various labor calculations in HM 5.3 to remedy understatements in labor rates. Both parties suggest "outboard calculations" to approximate the higher labor rates used in the SBC-CA models, which the Commission found were difficult to transport into HM 5.3. As discussed in Section V.B.3, we find these adjustments unreasonably speculative and untimely and will not make them.

E. The True-Up of Interim Rates

CALTEL, Vycera, Navigator, and Mpower all comment that because the rates in the Proposed Decision are substantially higher than the interim rates adopted in D.02-05-042 and D.02-09-052, the size of the adjustment, or true-up, resulting from SBC-CA's new permanent UNE rates will hurt the level of competition in California's local exchange telephone market and drive CLCs into bankruptcy. These parties urge the Commission to consider the effects of the true-up in limiting consumer choice by driving competitors out of the market and to take steps to mitigate these negative effects.

For example, Mpower suggests limiting the amount of any back payments owed to SBC-CA to the prior OANAD rates set in D.99-11-050, rather than the lower interim rates adopted in D.02-05-042. ORA/TURN suggest the Commission should offer CLCs the option of a phased true-up payment plan to even out the cash flow consequences of an unexpected increase in UNE rates. They also suggest that interest accrual on amounts CLCs owe to SBC-CA should cease with the effective date of a decision ordering final UNE prices. (ORA/TURN, 6/7/04, p. 9.)

We note that the size of the true-up in this decision differs dramatically from the true-up that might have occurred had the Proposed Decision been adopted unchanged. Nevertheless, even though the true-up now appears to be significantly less than it might have been, we are persuaded that further proceedings are necessary to consider true-up effects. We have modified the decision to stay the effectiveness of the payment of any true-up resulting from these new permanent UNE rates, pending a review of the size of the actual true-up and the outcome of further proceedings to consider the necessity of mitigation for any true-up payments. Furthermore, we note the recent order of the Ninth Circuit Court of Appeals regarding the shared and common cost mark-up requires us to reconsider that component of UNE rates. Further proceedings on the true-up, which we intend to conduct expeditiously, will allow us to consider whether and how to implement any mark-up changes, as ordered by the court, along with the true-up. Therefore, although the rates in this order will go into effect immediately on a prospective basis, payments to implement the true-up of interim rates will only occur following further Commission action.

F. Price Squeeze Arguments

Mpower contends that the proposed rate increases for UNE loops push costs for CLCs who purchase loops well above the price CLCs can competitively charge for their service, creating a price squeeze and hindering CLCs ability to compete with SBC-CA. Mpower notes that JA raised price squeeze issues and the Proposed Decision fails to include a thorough analysis of this subject.

SBC-CA responds that UNEs must be set at TELRIC rates and any consideration of CLC profitability or an alleged price squeeze in setting a TELRIC price is legal error.

We agree with SBC-CA that this decision is not the proper place to consider price squeeze arguments because we must price UNEs based on TELRIC and not arguments of CLCs concerning their profitability.