Given our intent to set interim rates for at least some of Verizon's UNEs, we now turn to the three interim pricing proposals presented by Verizon, Joint Commenters and TURN.
We shall first examine Verizon's proposal, which proposes a percentage adjustment to Verizon's current UNE rates based on a trend analysis of costs for Verizon in Florida.
Verizon believes that the Commission does not need to set interim rates, but should keep Verizon's current UNE rates in place pending the completion of this proceeding. Verizon argues that the Commission set these rates in D.97-01-022 after extensive proceedings. Nevertheless, Verizon provided an interim pricing proposal in the interest of avoiding extensive litigation regarding interim rates. According to Verizon, its proposals for interim UNE loop and switching rates are voluntary and due process requires that the Commission may not reduce rates beyond these proposals without a full and fair proceeding that would undoubtedly overlap with the permanent phase of this case. Verizon's proposal entails a 35% to 52% reduction in its UNE switching rates and no change in its UNE loop rates. On September 9, Verizon modified its original position and proposed a reduction in its UNE loop rates of 15.1%.
Verizon's proposed switching rate reductions are based on a trend analysis it performed of switching costs for Verizon's Florida operations. For each switching rate element, Verizon compared cost studies it filed with Florida regulators in 1996 and 2001 and calculated a percentage difference between the two filings. Verizon then applied this percentage difference, which amounted to a 35% to 52% reduction, to its current California switching rate elements to arrive at an interim price.14
Verizon does not propose interim rates for switch features because they are not included in its trend analysis using Florida. According to Verizon, its current prices and volumes for switch features are very low and it would be too great a burden to change its billing systems to track these elements for possible later true-up.
Verizon justifies its use of a Florida trend analysis for switching by arguing that trends in Florida switching costs over the 1996-2001 time period represent a good estimate of the trend that could be expected in California over the same period. First, Verizon notes that the 1996 Florida proposals were prepared in the same year and using the same methodology as Verizon's 1996 study used as the basis for its current California rates. Second, Verizon asserts the 2001 Florida study reflects current, forward-looking switching information. Third, Verizon contends that California and Florida have a similar (but not identical) switching infrastructure, including switch mixes. (Collins Declaration, 7/30/02, paragraph (para.) 7.) Finally, Verizon contends that its trend analysis is similar to the approach used by the Commission recently to set interim UNE prices for Pacific Bell,15 but is actually superior because it is "based on real ILEC data and more accurately reflects Verizon's forward-looking costs." (Verizon Proposal, 7/30/02, p. 4.)
With regard to interim loop rates, Verizon initially proposed no change in its UNE loop rates and justified this by claiming that its current California loop rates should be increased, not decreased. As support for this contention, Verizon states that its current loop rates are below the costs it submitted in California in its 1997 cost studies and are below the loop rates it has recently proposed in Florida. Further, Verizon stated that several significant loop cost drivers, such as material costs for copper and fiber cable, have increased since 1996. (Collins Declaration, 7/30/02, para. 12.)
On September 9, Verizon updated its interim UNE loop proposal and proposed to reduce its existing loop rates by the same 15.1% discount applied to Pacific Bell in its interim rate proceeding, subject to adjustment once final rates are set in the permanent phase of this case. This translates into a reduction in the current statewide average 2-wire loop rate from $16.81 to $14.27. Verizon explains that it makes this updated offer solely to bring this interim phase of the case to an expeditious conclusion so that the permanent phase involving review of cost studies can begin. Verizon notes that it does not agree with the Commission's basis for reducing Pacific Bell's loop rates by 15.1%, but it would allow this reduction to its current rates as an interim measure to avoid extensive litigation on interim rates.
Verizon opposes deaveraging its loop rates at this time because it has not been able to make a detailed proposal for deaveraging due to the expedited schedule for this interim phase of the case. (Verizon Reply Comments, 9/20/02, p. 7.) If the Commission chooses to deaverage loop rates for the interim, Verizon suggests that the rate for the former Contel areas that Verizon now serves should be $40.37 based on the data contained in GTEC's September 1997 cost filing. (Id., p. 8, n. 9.)
Joint Commenters request that the Commission set interim UNE rates for Verizon based on the interim prices adopted for Pacific in D.02-05-042, with appropriate adjustments to reflect cost differences between Verizon and Pacific as indicated by the FCC's Synthesis Model.16 They maintain that interim relief is warranted because Verizon's current prices are based on TSLRIC cost studies, which rely heavily on 1994 base year data, and were filed by GTEC in OANAD prior to the adoption of the Telecommunications Act of 1996. They also ask that the Commission take official notice of the evidence of cost reductions for key inputs for unbundled loops and switching since the 1994 time frame, as found in D.02-05-042, because these cost reductions impact Verizon as well as Pacific. Further, they highlight the merger of GTEC and Bell Atlantic in 2000 that led to the formation of Verizon, and contend that this merger provides ample reason to anticipate Verizon has experienced cost reductions and greater economies of scale since the time that its rates were set by the Commission.
For unbundled loops, Joint Commenters recommend interim prices based on two geographic zones - one corresponding to the former GTEC service territory and the other corresponding to the former Contel of California Inc. (Contel) service territory, which GTEC acquired in mid-1996.17 Joint Commenters propose that the loop price in each of these zones should equal the current loop price for Pacific, adjusted by the percentage difference in costs identified by the FCC's Synthesis Model. Specifically, the 2-wire loop price for the former GTEC territory should equal Pacific's interim statewide average price multiplied by the ratio of the Synthesis Model cost result for the former GTEC zone to the Synthesis Model cost result for Pacific. Joint Commenters then remove the 19% shared and common cost markup incorporated into Pacific's rates,18 and replace it with 8%.19 The resulting rate for the former GTEC territories is $9.35, or a 44.4% reduction from Verizon's current average loop rate of $16.81. A similar calculation is performed to set the 2-wire loop price for the former Contel territory. The proposed rate for that zone is $17.20, or a 2.3% increase to the current average loop rate.20
Joint Commenters support their Synthesis Model adjustment to Pacific's interim loop rates by explaining that the FCC itself has used the relative UNE costs reported by the Synthesis Model to assess the relative costs of unbundled loops and switching between ILECs in different states. Specifically, the FCC has used this test to gauge whether an ILEC's UNE rates are within a range of reasonableness for TELRIC compliance. (Murray Declaration 7/30/02, paras. 10-12.) Joint Commenters explain that although the FCC originally developed the Synthesis Model to calculate the forward-looking costs of providing universal service, the forward-looking economic cost criteria that the FCC applied are consistent with, and largely identical to, the forward-looking economic cost criteria that the FCC has adopted for UNEs. (Id., para. 10.)
For unbundled switching, Joint Commenters recommend that the Commission set prices for Verizon equal to the interim switching prices established for Pacific in D.02-05-042, translated into Verizon's current rate structure.21 Similar to their interim loop proposal, Joint Commenters also decrease the current markup in Pacific's switching rates from 19% to 8%. In contrast to Verizon, Joint Commenters include switch features in their proposal for interim switching rates because, they argue, switch features are an integral part of the unbundled switching UNE established by the FCC. Further, they contend that excluding switch features would create complications for using their methodology, which is anchored to Pacific's interim switching prices and includes switch features. (Joint Commenters' Proposal, 7/30/02, p. 3.) Joint Commenters also propose that the Commission use an analogous process to establish interim prices for all Verizon's port rate elements, mirroring any port discounts applied to Pacific's rates. (Id., p. 10, n. 27.)
To support the reasonableness of their proposal, Joint Commenters note that the Commission has twice before relied on Pacific's rates and cost models in other proceedings. In an arbitration proceeding involving Roseville Telephone Company and Covad Communications, the Commission found it did not have sufficient cost data on which to base a reasonable estimate of forward-looking costs for Roseville. Thus, the Commission relied on Pacific's adopted UNE prices as a proxy for Roseville's loops and other UNEs. (See D.01-02-042, mimeo., pps. 11-12; affirmed by D.01-06-089). Also, when the Commission set Verizon's nonrecurring costs in an earlier phase of OANAD, the Commission rejected GTEC's cost model and used Pacific's cost model, with adjustments, to set GTEC's nonrecurring costs. (See D.98-12-979, mimeo., pps. 27-29.)
Finally, Joint Commenters maintain that reliance on Pacific's interim rates is reasonable because of three main similarities between Verizon and Pacific: 1) both companies operate in California and are subject to the same economic conditions statewide and the same labor and material costs; 2) both companies have enjoyed similar increases to their economies of scale and scope as a result of recent mergers; and 3) public data shows that Pacific and Verizon are comparable with regard to line densities and certain switching variables such as switched lines per central office and the percent of host versus remote switches. (Murray Declaration, 7/30/02, paras. 25-32.) Joint Commenters contend that any network differences that do exist between Pacific and Verizon are captured by the Synthesis Model comparisons that Joint Commenters have used to adjust their proposed interim rates. In addition, Joint Commenters note that their proposal for separate, deaveraged loop rates for the former GTEC and former Contel service areas accounts for network differences between Pacific and Verizon.
Z-Tel, Mpower, and ORA support Joint Commenters' proposal to use Pacific's interim rates adjusted by the Synthesis Model.
In response to the August 23 ruling allowing interim pricing proposals based on rates in Verizon-affiliated states, Joint Commenters and TURN reiterate their view that Pacific's interim loop and switching rates, adjusted by the Synthesis Model, are appropriate as interim rates for Verizon. Nevertheless, Joint Commenters and TURN state that if the Commission prefers to use rates from another Verizon jurisdiction, the Commission should set interim loop and switching rates for Verizon based on the most recently adopted New Jersey TELRIC-based UNE rates.22 Joint Commenters further suggest that if New Jersey rates are used for California, the Commission should use the Synthesis Model to develop adjustment factors to reflect the loop and switching costs differences identified by the Synthesis Model between Verizon California and Verizon New Jersey. Specifically, Joint Commenters apply the same adjustment methodology suggested in their initial proposal based on Pacific's interim UNE rates. Finally, Joint Commenters and TURN propose that the Commission adjust the 10% shared and common cost markup embedded in Verizon New Jersey rates to the 8% level suggested previously.
With regard to interim loop rates, Joint Commenters indicate that according to the Synthesis Model, Verizon California's statewide-average loop costs are 6% higher than loop costs for Verizon New Jersey. If the comparison of loop costs is done on a deaveraged basis, loop costs in the former GTEC service territories areas are .83% lower in California than New Jersey, while in the former Contel California service territory, loops costs are 82.39% higher than New Jersey loop costs. (Murray Declaration, 9/9/02, para. 15.) Joint Commenters propose deaveraged loop rates for the former GTEC and former Contel service areas that are .83% lower and 82.39% higher than New Jersey rates, respectively. The proposed rates, including an adjustment of shared and common costs to 8%, are $9.27 and $17.05 for the former GTEC and former Contel service areas, respectively.
With regard to switching, the Joint Commenters' analysis using the Synthesis Model indicates Verizon California's costs are 3% lower than costs for Verizon New Jersey. (Murray Declaration, 9/9/02, para. 10.) Therefore, Joint Commenters propose switching rates that are 3% lower than switching rates in New Jersey.23 For switch features, Joint Commenters point out that Verizon New Jersey includes features costs in the UNE port charge and prices only a few switch features individually. Therefore, Joint Commenters propose switch features priced similarly in California, i.e., individual prices for those features that are individually priced in New Jersey, and all other features included in the port rate. (Joint Commenters' Ruling Response, 9/9/02, p. 7, n. 16.) The proposed feature rates are 3% lower than New Jersey rates.
Although UNE rates were also recently adopted for Verizon in New York, Joint Commenters and TURN propose basing interim California prices on New Jersey rates for several reasons. First, they stress that the New Jersey rates reflect the most recent TELRIC based, forward-looking cost information from another Verizon state. Newly adopted New Jersey rates reflect conditions following the GTE and Bell Atlantic merger. Second, the rates for New Jersey were fully litigated before adoption by New Jersey regulators in contrast to rates for Verizon New York, which emerged from a settlement. Third, Joint Commenters and TURN contend that New Jersey's cost modeling criteria and assumptions are more consistent with this Commission's prior rulings than New York rates.24
In comments on the various proposals, ORA prefers the proposal to use Pacific's interim rates for Verizon. Nevertheless, ORA contends it is also reasonable for the Commission to look to other Verizon jurisdictions, such as New Jersey, that have recently set forward-looking, TELRIC-compliant rates. ORA notes that Verizon's own interim proposal and supporting declarations contradict its concerns about the validity of using rates of another Verizon jurisdiction for comparison.
We find that Verizon's proposal does not withstand the criticism leveled by Joint Commenters, namely that the proposal rests on an imperfect foundation because it recommends discounts from TSLRIC-based rates that date back to 1996.
The primary problem with Verizon's proposal for interim loop and switching rates is that it involves a percent reduction from Verizon's current UNE rates. Thus, the proposal rests on the presumption that current prices reasonably reflect forward-looking economic principles. This is not the case. As explained above, Verizon's current rates are based on unsatisfactory, non-forward-looking cost studies that the Commission concluded "do not adequately conform with the TSLRIC principles adopted in D.95-12-016." (D.96-08-021, mimeo., p. 91.) It is not reasonable to leave these rates in place while we evaluate cost studies in the permanent phase of this case. Although the Commission attempted in 1996 to remedy the flaws that it found, this was a temporary measure and it is time to revisit Verizon's rates. Thus, for the same reasons that interim rates are warranted, we find that it would be improper to use these TSLRIC based rates as a starting point for any interim rates.
Verizon defends its proposal with several arguments. First, Verizon contends that its current rates are an appropriate base from which to calculate interim rates because they were extensively litigated and there is no basis to abandon them. Second, Verizon argues that the Commission's 1997 order significantly adjusted GTEC's TSLRIC cost studies, which makes Verizon's current UNE rates forward-looking. Third, Verizon argues that the Commission did not find flaws with the switching portion of GTEC's 1996 cost filing.
We disagree on all points. We do not find it reasonable to use Verizon's current rates as a basis for interim rates simply because a lot of time was spent litigating adjustments to GTEC's non-forward-looking TSLRIC model. The fact that parties and the Commission spent a lot of time litigating these adjustments does not mean we are bound to these rates. Since the time that GTEC performed its original TSLRIC analysis, the Commission and the entire country have moved to the TELRIC approach to UNE ratesetting. It is now time for Verizon's rates to be based on a TELRIC approach as well. Furthermore, we do not agree that the adjustments ordered in D.96-08-021 and implemented in D.97-01-022 suffice today to call Verizon's current rates forward-looking. These adjustments were not intended to be in place for six years, particularly when the Commission ordered new cost studies to be filed in 1997. Further, we have found ample reason to adjust Pacific's forward-looking rates, which were set later than Verizon's. The fact that Verizon's rates may have been considered forward-looking in 1996 does not mean they are still forward-looking today, particularly when Verizon does not dispute the cost declines that led us to lower Pacific's rates in D.02-05-042. Finally, we do not agree that the Commission did not find flaws with GTEC's switching model. The Commission ordered GTEC to file new cost studies in 1997 for all of its UNEs and did not exempt switching from this directive. Given all of these factors, we find sufficient justification to avoid using Verizon's current UNE rates as a basis for any interim rates that we will set in this order.
Verizon also maintains that the Commission must fully consider Verizon's proposal, and that it cannot adopt anything but the Verizon proposal without extensive further proceedings. We disagree. Given that this case involved four rounds of comment on interim pricing proposals, Verizon has been given ample opportunity to explain its interim proposal and have its due process rights protected. While we appreciate Verizon's efforts in providing an interim pricing proposal for us to consider, the Commission is not limited to Verizon's proposal.
With regard to interim switching rates, we agree with Joint Commenters that Verizon has proposed a flawed trend analysis based on proposals in Florida using two different cost models. Essentially, the 1996 and 2001 Florida studies used to indicate a trend in switching prices do not provide an apples to apples comparison. Joint Commenters detail significant problems with the Florida trend analysis including:
· The trend analysis relies on two different cost models, with different methodologies and assumptions. The 1996 cost model was prepared in the same year and using the same TSLRIC methodology as the Verizon California 1996 study (which the Commission found flawed in D.96-08-021). (Murray Declaration, 8/20/02, para. 9.) The 1996 study estimates costs based on a small subset of wire centers (Id., para. 16.), whereas the 2001 cost study is based on the TELRIC-based ICM, and it estimates switching costs for each wire center. (Id.)
· Verizon itself admits that the two Florida cost studies use different methodologies and cannot be relied on to produce a reliable trend for tandem switching rates. (Id., para. 8-17.)
· Verizon omits switch features, which Joint Commenters consider an integral part of the switching UNE and the cost analysis on which Verizon relies for its cost trend. Joint Commenters contend that Verizon's failure to include features could bias the cost trend. Further, switch feature costs should decline in parallel to other switching costs. (Id., paras. 35-36.)
We agree with Joint Commenters that Verizon's trend analysis, which is based on a comparison of Florida cost study proposals, is significantly different from the trend analysis used to set rates for Pacific Bell. In setting Pacific's interim rates, the Commission relied on the same model to compute costs at both end points of the trend analysis. This enabled the Commission to isolate the effect of changing only a few variables, namely loop and switching equipment costs and demand growth. In contrast, Verizon's trend analysis uses two different cost models, which Verizon admits contain methodology differences. (Collins Declaration, 7/30/02, para. 9.) This fact alone makes Verizon's proposed trend analysis quite different from the one we performed in calculating interim rates for Pacific.
Further, our own experience with Verizon's cost filings in California in 1996 and 1997 cast doubt over whether these models are truly comparable. Verizon admits that the 1996 model it uses for its trend analysis is similar to the model filed in 1996 in California. (Id., para. 7.) This 1996 California model is the same one that the Commission found inadequate and modified in D.96-08-021 to approximate forward-looking costs. Verizon uses a 2001 version of its ICM model for the trend analysis, which appears to be a later version of the ICM model that Verizon filed in California in 1997 to replace the 1996 studies the Commission rejected. While we agree with Verizon that cost models may evolve over time to reflect technology and network design changes, we do not find it appropriate to use admittedly different cost studies to perform a trend analysis. Therefore, the Commission cannot be assured that the "trend" identified by the models represents changes in forward-looking economic costs rather than methodological changes in the model itself that are irrelevant to cost changes. Because we have other options available and proposed in detail on the record, we will not use Verizon's trend analysis.
Verizon asks us to take official notice that on October 14, 2002, the Florida PSC adopted UNE rates derived from Verizon's proposal in that state and these rates are higher than Verizon's proposal in California. Joint Commenters do not object to this notice, but ask that the Commission disregard Verizon's claims concerning the implications of Florida's new UNE prices. We will grant Verizon's request for official notice, but as Joint Commenters suggest, we will disregard Verizon's claims regarding these rates. Likewise, we will disregard Joint Commenters' counter claims regarding the import of the Florida rates.25 At the same time, we note that Verizon does not explain the differential between its proposed UNE rates in Florida and the rates ultimately adopted by the Florida PSC. For example, the Florida PSC adopted end office switching rates per minute of use (MOU) and tandem switching per MOU that are 24 % and 16% lower than Verizon's initial Florida proposals, respectively. The Florida PSC adopted a basic port rate 29% lower than Verizon's proposal in that state. (Verizon Petition to Reopen Proceeding, 10/25/02, Attachment (Appendix A-1, p. 385.)) The fact that the Florida PSC adopted final UNE rates does not mean that Verizon's trend analysis using proposals it made in Florida is valid, particularly when the adopted rates are significantly lower than Verizon's proposed rates. Thus, we are still not persuaded to rely on Verizon's trend analysis.
Finally, Verizon's Florida proposals can be differentiated from SBC-Ameritech's proposals in Illinois, which were used to set interim rates for Pacific, because in that case the record revealed similarities between California and Illinois with regard to switching characteristics. Here, Verizon has only asserted similarities, but has provided no data showing these similarities actually exist. (Collins Declaration, 7/30/02, para. 7.)
Turning to interim loop rates, Verizon provides little explanation why it supports the 15.1% discount recently applied to Pacific. Although Verizon proposes a trend analysis using Florida cost studies for its interim switching proposal, it abandons this approach for loops. This inconsistency is largely unexplained. Verizon suggests it does not want to waste time litigating interim rates, but it does not explain why a trend analysis using Florida or another Verizon state would not be appropriate for loops.
We understand that Verizon is offering this approach as a way to expedite the proceeding and avoid prolonged litigation over interim rates. While we can appreciate that as a goal, this does not mean we should give greater weight to Verizon's proposal, or ignore inconsistencies in it, simply because it was voluntary. Verizon's loop proposal contains a reversal from its earlier positions without an adequate explanation. Verizon initially alleged that it would be inappropriate to apply Pacific's switching discount to Verizon because there are no similarities between Pacific's and Verizon's switching networks. Verizon also opposed use of Pacific's loop and switching interim rates because they resulted from sanctions on Pacific. Later, Verizon reversed positions in part and now supports using Pacific's loop discount. Verizon takes this new position with regard to interim loop rates, and in contrast to its position on switching, without showing any similarity between Verizon's and Pacific's loop networks to support using Pacific's discount. These contradictions lead us to find that Verizon has not sufficiently explained why it would be reasonable to use the percentage discount found in California for Pacific's loop costs, but not for switching. On the whole, we find Verizon's loop discount proposal unreasonable and contradictory because it proposes one methodology for its switching rates and an entirely different methodology for loop rates. Verizon's explanation that it is only trying to expedite this proceeding is not enough to support adoption of its proposal when other alternatives are available to the Commission. .
Verizon contends that although it agrees to a 15.1% decrease to its average loop rate, its loop rate should actually increase because the current rate is below cost. Verizon alleges that several loop cost drivers, such as labor, cable costs, and the cost of money, have increased since 1996. Verizon's argument for higher loop costs is not well-supported. We are not convinced that Verizon's current loop prices are below TELRIC and should be raised simply because proposals that Verizon made here in 1997, and later in Florida, are higher than current TSLRIC-based rates. The fact that Verizon has proposed higher rates does not prove that current rates are below cost, particularly when neither of Verizon's proposals in Florida or California has been adopted.26 Moreover, Verizon has not provided any analysis indicating why Florida loop costs are indicative of California loop costs. Verizon also argues that several cost indices prove loop cost drivers have increased. Verizon's contentions are not supported by an analysis of the effects of any input increases on per unit loop rates. If Verizon intends to pursue this line of reasoning, it will need to provide further support for its analysis in the permanent phase of this proceeding.
Verizon argues we should accept its proposed interim rates because its own comparison of loop and switching rates across other states in which Verizon operates, even after adjustment using the Synthesis Model, shows that its proposal is reasonable.27 Specifically, Verizon explains that its proposed average loop rate of $14.27 and its 35% to 52% switching rate discounts are well within the range of Synthesis Model adjusted rates in other Verizon states. (Verizon Response, 9/9/02, p. 8). We do not find this to be a meaningful or convincing argument. The fact that Verizon's proposed rates fall within the range of other states' UNE rates, many of which were adopted as long ago as 1996 and 1997, does not prove that Verizon's proposed rates are accurate. Verizon's multi-state comparison produces a very large range. One could just as easily argue from this analysis that rates for California should fall at an end-point for this range rather than somewhere in the middle. Moreover, Verizon's Synthesis Model analysis indicates that average California loop rates should be 7% higher than New Jersey loop rates, but Verizon proposes a loop rate of $14.27 that is 50% higher than New Jersey's current loop rate without explaining this differential. (Meny Declaration, 9/9/02, Attachment 1, p. 5.)
Indeed, Verizon's multi-state rate comparison presents further contradictions in Verizon's logic. On the one hand, Verizon contends it is inappropriate to use another Verizon state to set rates for California. At the same time, Verizon maintains that a comparison of Verizon's proposed rates to the range of UNE rates across all Verizon jurisdictions, adjusted to reflect California specific inputs using the Synthesis Model, shows the reasonableness of Verizon's own proposal. (Meny Declaration, 9/9/02, para. 2.)
In summary, we find Verizon's proposals use an inappropriate TSLRIC starting point for discounts to both loop and switching prices, and a flawed trend analysis for switching. Verizon provides little justification for using the 15.1% discount recently applied to Pacific's unbundled loop rates, particularly when it eschews this idea for switching. Verizon's Synthesis Model analysis is not convincing when it suggests that we should adopt Verizon's proposed interim rates merely because the rates fall within the large, and somewhat dated, range of rates from other Verizon states. Therefore, Verizon's interim rate proposals are not reasonable and we will not adopt them.
4. Joint Commenters' and TURN's New Jersey Proposal is Preferable to the Proposal to Use Pacific Bell Interim Rates
In contrast to Verizon's proposal to base interim UNE rates on its current TSLRIC rates, Joint Commenters propose to use Pacific Bell's interim UNE rates as a basis for interim UNE rates or, if the Commission prefers a proposal based on another Verizon jurisdiction, they propose, along with TURN, to use Verizon's New Jersey UNE rates.
We agree with Joint Commenters' and TURN's proposal to base Verizon's interim UNE rates on rates for Verizon New Jersey because these are among the most recent forward-looking, TELRIC-based UNE rates from another Verizon state.28 We agree with Joint Commenters and TURN that using the Synthesis Model to reflect loop and switching cost differences between California and New Jersey is reasonable for the purposes of interim pricing. Because the rates apply to a company within the same regional Bell operating company (RBOC), it is reasonable to use the Synthesis Model to adjust for regional and network differences between California and New Jersey. Further, because these rates are so recently adopted, they reflect conditions after GTEC's merger with Bell Atlantic in mid-2000 to form the company now known as Verizon. Moreover, we agree that using Verizon New Jersey rates is a better alternative than using Verizon New York rates because New York rates were the product of a settlement. Despite Verizon's objections, we find that deaveraging as proposed by Joint Commenters and TURN is reasonable because it is based on the differences shown by the Synthesis Model for the former GTEC and former Contel areas now served by Verizon.
We find that Joint Commenters' and TURN's proposal to use New Jersey rates is not that different from Verizon's own suggestions to use Florida rates for a trend analysis and to use other Verizon states as a reference point for California rates. For example, Verizon proposes to use Florida cost studies while providing little, if any, data on network similarities between Florida and California. We find it superior to use the New Jersey rates adjusted for the Synthesis Model rather than Verizon's Florida-based trend analysis because the New Jersey rates are final adopted rates in contrast to trends elicited from proposed rates in Florida. We also find it preferable to use New Jersey rates as a starting point for interim rates rather than Verizon's current California rates, which are derived from a 1996 TSLRIC analysis as we have already discussed. Verizon itself has suggested the reasonableness of comparing UNE rates across other Verizon jurisdictions using the Synthesis Model. Thus, we find that it is appropriate to use New Jersey rates as an interim measure, especially after adjustments based on the Synthesis Model.
Joint Commenters and TURN have proposed adjusting either the New Jersey rates or Pacific's interim rates using the FCC's Synthesis Model. We agree with Verizon that it may not be appropriate to use the Synthesis Model for inter-RBOC comparisons, particularly when the FCC has only made comparisons when two states have a common RBOC. (Verizon Comments, 8/20/02, p. 12.) For this reason, we do not think it is appropriate to compare the costs of Verizon and Pacific and adjust them using the Synthesis Model. Instead, we will base interim rates on Verizon's New Jersey rates, adjusted using the Synthesis Model. Although Joint Commenters and TURN urge us to overlook the FCC's cautions in this circumstance, we are not compelled to do so because we have the option of relying on rates recently adopted in another Verizon jurisdiction. Therefore, we find Joint Commenters' and TURN's proposal to use New Jersey rates more reasonable than using Pacific Bell's interim rates. We prefer to base interim rates on a TELRIC analysis emanating from Verizon itself, as opposed to Pacific.29
We find that using the Synthesis Model to compare relative costs between California and New Jersey is reasonably consistent with how the FCC has used the Synthesis Model in Section 271 proceedings to determine whether UNE rates are acceptable.30 We agree with Joint Commenters that the forward-looking economic cost criteria that the FCC applied in developing the Synthesis Model for universal service purposes are consistent with, and largely identical to, the forward-looking cost criteria that the FCC has adopted for UNEs.31 Further, we agree with Joint Commenters that the Synthesis Model adjusts relative switching costs based on major cost drivers, including line density per switch and the number of host and remote switches.32 For loops, the Synthesis Model takes into account population densities that drive equipment sizing, a major determinant of geographic variation in loop costs.33
When it comes to actually performing the Synthesis Model comparison of California and New Jersey rates, Joint Commenters and Verizon take different approaches. Joint Commenters used Synthesis Model results straight from the FCC's website to derive a differential between California and New Jersey UNE costs. For loops, Joint Commenters developed adjustment factors between California and New Jersey on a deaveraged basis by comparing Synthesis Model results for the former GTEC and former Contel areas now served by Verizon with the average rates for New Jersey.
Verizon claims that Joint Commenters' approach inappropriately allocates all of the fixed per-line common support expenses to loop costs instead of assigning a portion of these costs to other elements. (Collins/Meny Declaration, 8/20/02, para. 20.) Joint Commenters counter that there is no single agreed upon approach for allocating overheads in the Synthesis Model. They offer comparisons with varying overhead assumptions and claim that despite these varying assumptions, the results do not vary by that much. (Klick Declaration, 9/20/02, para. 16.)
In contrast to Joint Commenters approach, Verizon compares UNE rates across states after adjusting common support expenses and other items as described by the FCC in the Section 271 benchmarking process. (Meny Declaration, 9/9/02, para. 5.)34 Verizon's Synthesis Model analysis does not compare loop costs on a deaveraged basis, but only compares Verizon's California statewide average loop rate to New Jersey's statewide average loop rate. Verizon's results indicate that California loop and switching costs are 7% higher than New Jersey costs. (Id., Attachment 2, p. 5.)
Joint Commenters criticize Verizon's analysis because it does not use a consistent approach between loops and switching. Specifically, Joint Commenters allege that Verizon does not include any overhead expense allocation when it compares switching costs. (Murray Declaration, 9/20/02, para. 10, n. 11.) Further, Verizon distorts its analysis by focusing on end-office switching investment while disregarding other components of total switching investment. (Klick Declaration, 9/20/02, paras. 7-12.)
We have examined both methods used for Synthesis Model comparisons and we prefer to use an approach consistent with the method most recently used by the FCC. This is the same as the approach offered by Verizon, with one exception. We will compare the Synthesis Model results for deaveraged zones in California to the statewide average Synthesis Model results for New Jersey. In other words, we will compare the results for the former GTEC territories and the former Contel territories to New Jersey results because we want to set deaveraged loop rates for Verizon in California.
Commission staff has made the adjustments noted by Verizon in order to obtain Synthesis Model results for this comparison. We have been careful to avoid the mistakes alleged by Joint Commenters regarding Verizon's switching analysis, by treating overheads consistently for loops and switching and considering the results based on total switching investment. The results show that loop costs in the former GTEC areas and in the former Contel areas are 2.2% lower and 111.9% higher than New Jersey average loop costs, respectively. For switching, the Synthesis Model as run by Verizon indicates that end-office usage costs for Verizon are 10.7 % higher than New Jersey, port costs are 3.2% lower than New Jersey, and tandem switching costs are 38.9% higher than New Jersey. The analysis and results of this Synthesis Model comparison are set forth in Appendices B and C.
Based on these results, we will set rates for Verizon equivalent to New Jersey rates, and apply the percentages shown by our Synthesis Model analysis wherever the results show that California costs are higher than New Jersey costs. In other words, we will not set interim rates any lower than New Jersey UNE rates, even where the Synthesis Model indicates that California costs may be lower than New Jersey. We prefer this approach in order to be conservative in setting interim rates. This means that, before adjusting for shared and common costs, 35 loop rates in the former GTEC areas will be set equal to the New Jersey loop rate of $9.52, and loop rates in the former Contel areas will be set 111.9% higher than the current New Jersey loop rate.36 For Interim UNE switching, end-office usage rates will be set 10.7% higher than the current New Jersey rate, port and feature charges will be set equivalent to New Jersey rates, and tandem rates will be set 38.9% higher than New Jersey rates.37
Verizon opposes any adoption of another state's UNE rates for use in California, and it specifically objects to the proposal to use New Jersey rates. Verizon contends that this Commission should not abdicate its authority to review the numerous methodological and other issues, such as network requirements, geography, and demand levels, that factor into UNE rate decisions. Further, Verizon argues that New Jersey rates have nothing to do with the costs Verizon incurs to provide UNEs in California. As an example, Verizon contends that New Jersey UNE rates are based upon different depreciation lives than used by this Commission. (Verizon Reply Comments, 9/20/02, p. 21).
We agree with Verizon that the Commission should not abdicate its obligation to set UNE rates and it should examine the multitude of issues involved with cost modeling. We will indeed pore over the multitude of cost modeling issues in the permanent phase of the case. On the other hand, the circumstances surrounding the delays in setting TELRIC-based UNE rates for Verizon are unusual and have led to an extraordinarily lengthy delay in this case. We reluctantly turn to another state at this juncture, as we found it necessary to do for Pacific, because of the unique circumstances of this case. When we used Illinois rates for Pacific's interim switching rates, it was necessitated by discovery disputes. We turn to New Jersey to set interim rates for Verizon because we can no longer justify rates for Verizon based on a 1996-era TSLRIC cost study.
We have already explained why we do not find Verizon's own proposal reasonable because it is based on an inappropriate starting point and a flawed trend analysis. Given these findings, we find the proposal to use Verizon's New Jersey rates more reasonable than using Pacific's interim rates, despite the potential methodological differences between New Jersey's cost studies and this Commission's prior rulings. The interim rates we adopt today will be subject to adjustment once final rates are set. This fact, coupled with the Synthesis Model adjustment to account for cost differences between New Jersey and California, override our concern with any methodological differences such as differing depreciation assumptions.38 For these reasons, coupled with our view that Verizon's current rates are not a legitimate starting point for calculating interim rates, we find it reasonable to turn to the work of another state in this particular circumstance. We are also satisfied that this is a sound approach because in applying rates from Verizon New Jersey, we rely on the methodology used by the FCC to compare UNE rates in order to set what we consider TELRIC-based interim rates for Verizon.
In adopting the proposal to use Verizon New Jersey rates, we must decide whether to accept Joint Commenter's proposal to convert the New Jersey switching rates into the Verizon California switching rate structure. Joint Commenters point out that Verizon New Jersey has incorporated the costs of switch features into the port charge, although a few switch features are priced individually. With regard to switch features, Joint Commenters propose eliminating charges for features for which Verizon New Jersey does not charge, and basing charges for other features on New Jersey's current rates adjusted using the Synthesis Model. Verizon has proposed leaving switch feature charges at current levels rather than setting interim rates for them.
We will use the Joint Commenters' approach to switch features because otherwise, it would be difficult to use the New Jersey rates to set separate port and feature prices for Verizon California. For simplicity and because the rates will be subject to adjustment once final rates are adopted, we will accept the Joint Commenters' proposals and adopt rates for switch features equal to the rates New Jersey has established for switch features. This means that where Verizon New Jersey charges an individual feature rate, the same rate will be charged for Verizon California. Interim rates for all other features will be set at zero because Verizon New Jersey has incorporated these into its port charge, which we adopt for the interim in California. In a later phase, if we find that it is reasonable to adopt charges for all of Verizon's switch features, carriers will owe payment to Verizon for features they have purchased from the date of this order.
With regard to DS-1 port rates, we note that when we use New Jersey rates as interim rates for Verizon, the DS-1 port rate is above the level that Verizon currently charges for this UNE in California. We have no basis on which to increase any of Verizon's current UNE rates, and indeed Verizon itself proposed a sizeable 34.5 % reduction to the DS-1 port rate. Therefore, we will leave the DS-1 port rate at its current level, subject to adjustment from the date of this order, once we establish a final DS-1 port rate in the permanent phase of this proceeding.
Joint Commenters request that we extend any interim discounts to Verizon's 2-wire and Centrex port rates to other, "non-basic" ports. Joint Commenters fail to identify the other "non-basic" ports that they want discounted and we are unclear if there are corresponding port rates for Verizon New Jersey that we could use to derive other "non-basic" port rates. We are unwilling to make a guess on what additional ports Joint Commenters are talking about. Therefore, we will deny this request.
The interim rates that we adopt in this order shall be adjusted, either up or down, once final rates are set. This is the same as the approach the Commission took when setting interim UNE rates for Pacific in D.02-05-042. In that order, the Commission noted that a true-up provision protected against later claims that interim rates were not cost-based in compliance with Section 252(d) of the Federal Telecommunications Act. (D.02-05-042, mimeo., p. 50.)
Therefore, we require Verizon to establish a balancing account to track the revenues received from the interim UNE rates adopted in this order. The balancing account should begin tracking revenues on the same date the interim rates become effective, which is the effective date of this order. Further, the balancing account should accrue interest at the three-month commercial paper rate, as is common practice for accounts of this type. When permanent UNE rates are adopted at the conclusion of this Verizon UNE phase, we will determine how to adjust the interim rates, either up or down, from the date the interim rates became effective through the date of adoption of a final rate.
a. Verizon.
After the mailing of the Draft Decision on November 14, 2002, Verizon filed comments recommending the Commission reject the Draft Decision and adopt Verizon's interim rate proposals instead. Verizon alleges that its proposals are the only ones supported by any evidence. According to Verizon, interim UNE rates must reflect the cost of providing UNEs in California and adoption of Verizon New Jersey rates results in interim UNE rates that are below TELRIC. Verizon states that the ALJ did not review New Jersey cost studies, did not make a finding that Verizon New Jersey rates reflect TELRIC costs in California, and did not account for the fact that New Jersey inputs differ from prior decisions of this Commission. Therefore, Verizon contends that the Draft Decision results in interim UNE rates below TELRIC and not supported by substantial evidence. Verizon faults the Draft Decision for dismissing the UNE rates recently adopted in Florida after submittal of the record in this matter, and suggests that these new Florida rates are more probative of UNE costs than New Jersey rates.
We disagree with Verizon's fundamental assumption that the interim rates in this order are below TELRIC and therefore not reasonable. This Commission has never established TELRIC-based rates for Verizon, as discussed at length in this order. It is difficult to understand how any interim rates set herein could be "below TELRIC" when the Commission has never actually completed a forward-looking (i.e. TELRIC) analysis of Verizon's costs.
Verizon suggests that a detailed review of New Jersey cost studies is required in order to set interim rates. We disagree because this would defeat the purpose of interim rate-setting, which is to arrive at a proxy rate quickly while a comprehensive review of cost studies continues on a reasonable schedule. In D.02-05-042, the Commission discussed in some detail its authority to set interim rates "as long as the rate is subject to refund and sufficient justification for the interim relief has been presented." (D.02-05-042, mimeo at 11, citing TURN v. CPUC (44 Cal. 3d 870,878 (1988)). Both of these criteria are met as discussed in this order. We can rely on the TELRIC analysis of the New Jersey Board of Public Utilities without an extensive review of New Jersey cost studies.39
Verizon suggests there is no credible rationale for applying New Jersey rates to California, and it presents several pages of entirely new argument on why New Jersey UNE rates should be ignored. Unfortunately, Verizon is providing new factual information in its comments after the close of the record and without an opportunity for other parties to comment adequately on the assertions therein. Rule 77.3 prohibits submittal of new factual information in comments on a draft order. Verizon had ample opportunity to present its arguments opposing the New Jersey rates and to suggest adoption of rates from a state other than New Jersey. We will not consider new information and arguments that Verizon presents in comments on the draft decision.
Furthermore, we find that the Draft Decision provides a credible rationale for adopting New Jersey rates, namely an analysis using the FCC's Synthesis Model and corresponding adjustments to New Jersey rates. This analysis was conducted using the information supplied by the parties before submittal of the case. Indeed, the decision adopts the methodology offered by Verizon for adjusting rates using the Synthesis Model. Verizon's contention that New Jersey rates are not TELRIC-based is inappropriate. We will not second-guess the work of another state commission and essentially "rehear" how New Jersey set UNE rates. The New Jersey Board of Public Utilities has found that the rates it adopted were based on the TELRIC standard.40 Verizon can always seek remedy by appealing the order of the New Jersey Board if it believes that errors were made there.
Verizon suggests that the Draft Decision dismisses the Florida rates out of hand. Yet Verizon never proposed that actual Florida rates be used to set interim rates in California. It appears that Verizon considers the adoption of final rates in Florida as supportive of its trend analysis. Once again, Verizon presents new information regarding the newly adopted Florida rates after the close of the record in this phase. This order takes official notice of the newly adopted Florida rates, but still dismisses the Verizon proposal to adopt rates based on a trend analysis. We see no connection between the adoption of rates in Florida and Verizon's trend analysis using proposals it made in Florida. The fact that the Florida Commission adopted final rates does not cure the flaw in Verizon's trend analysis, namely that it relies on two different cost models that estimate forward-looking costs using different methodologies.
b. Joint Commenters and Other Parties
Joint Commenters suggest that the Draft Decision errs in using Verizon's methodology to adjust UNE prices using the Synthesis Model. Specifically, Joint Commenters state that the Draft Decision improperly compares end-office usage costs per minute of use rather than per switched line. Covad
Communications Company (Covad) echoes these comments. We will not revise the Draft Decision in this regard because we find that the appropriate cost driver for the switching UNE is usage and not line counts. Further, Joint Commenters contend that comparison of California and New Jersey switching sub-elements (i.e. ports and usage) creates interim rates that are systematically biased upward. Again, we disagree. The Synthesis Model provides the detail to compare at the sub-element basis and we find it appropriate to use this level of detail to set interim rates for Verizon California.
29 Verizon suggests that the Commission cannot use Pacific's interim rates as a basis for Verizon's interim rates because of due process problems. Verizon has had ample opportunity to respond to this proposal and make its case regarding this approach. Pacific's interim loop rates are not based on a discovery sanction against Pacific, but are based on a trend analysis measuring changes in equipment input costs and line growth. Pacific's interim switching rates are based on a discovery sanction in part, but are also supported based on a showing of geographic similarities in addition to the discovery sanction.
30 See Joint Application by SBC Communications, Inc., Southwestern Bell Tel. Co., and Southwestern Bell Communications Services Inc., d/b/a Southwestern Bell Long Distance for provision of In-Region, InterLATA Services in Kansas and Oklahoma, Memorandum Opinion and Order, FCC 01-29, CC Docket No. 00-217, 16 FCC Rcd 6237, January 19, 2001, at para. 84. ("Our USF cost model provides a reasonable basis for comparing cost differences between states. We have previously noted that while the USF cost model should not be relied upon to set rates for UNEs, it accurately reflects the relative cost differences among states.") 31 See Murray Declaration, 7/30/02, para. 10, citing, In the Matter of Federal-State Joint Board on Universal Service, Report and Order, FCC 97-157, CC Docket No. 96-45, 12 FCC Rcd 8776, May 7, 1997, at paras. 250 and 251. 32 See In the Matter of Verizon New England, Inc., Bell Atlantic Communications, Inc., NYNEX Long Distance Company and Verizon Global Networks Inc., For Authorization to Provide In-Region Inter LATA Services in Massachusetts, Memorandum Opinion and Order, FCC 01-130, CC Docket No. 01-9, 16 FCC Rcd 8988, April 16, 2001, at para. 23. 33 See In the Matter of Federal-State Joint Broad on Universal Service; Forward-Looking Mechanism for High Cost Support for Non-Rural LECs, Tenth Report and Order, FCC 99-304, CC Docket No. 96-45 and 97-160, 14 FCC Rcd 20156, October 21, 1999, at para. 84 ("The model also uses structure cost tables that identify the per foot cost of loop structure by type (aerial, buried, or underground), loop segment (distribution or feeder), and terrain conditions (normal, soft rock, or hard rock) for each of the nine density zones.") and para.87 ("The Commission found that an efficient carrier will vary its plant mix according to the population density of an area and tentatively concluded that the assignment of plant mix defined by the model should reflect both terrain factors and line density zones"). 34 Specifically, Verizon started with the Synthesis Model's default wire center results, then it allocated fixed, per-line common support expenses across all UNEs, removed the allowance for retail uncollectible revenues in the model, and added allowances for wholesale uncollectible revenues and carrier-to-carrier customer service costs. (See In the Matter of Application of Verizon Pennsylvania Inc., Verizon TFQIy