Verizon Pennsylvania, Inc. v. Pennsylvania Public Utility Commission

380 F. Supp. 2d 627 (2005)

VERIZON PENNSYLVANIA, INC., Plaintiff,
v.
PENNSYLVANIA PUBLIC UTILITY COMMISSION, et al., Defendants.

No. Civ.A. 04-3866.

United States District Court, E.D. Pennsylvania.

August 3, 2005.

Barbara W. Mather, Joanna J. Cline, Pepper Hamilton LLP, Philadelphia, PA, for Plaintiff.

David Eric Screven, Kimberly Ann Joyce, Maryanne R. Martin, Pa. Public Utility Commission, Harrisburg, PA, for Defendants.

Robert B. Hoffman, Wolf Block Schorr & Solis-Cohen LLP, Harrisburg, PA, for Intervenor Defendant.

Daniel Mach, Jenner & Block, LLC, Washington, DC, Heather Jean Kelly, DLA Piper Rudnick Gray Cary US, LLP, Philadelphia, PA, Jeffrey A. Rackow, Worldcom, Inc., Lee S. Taylor, Michael B. DeSanctics, Robin M. Meriweather, Victoria H. Jueds, Jenner & Block, LLC, Washington, DC, for Intervenor.

MEMORANDUM & ORDER

KATZ, Senior District Judge.

Plaintiff Verizon Pennsylvania, Inc. ("Verizon") brings this claim against the Pennsylvania Public Utility Commission ("PUC") and its individual officers^[1] challenging the PUC's Order setting the rates that Verizon must charge competitors for access to components of its local telephone network. MCIMetro Access Transmission Services, LLC ("MCI") and AT & T Communications of Pennsylvania, LLC ("AT & T"),^[2] two such competing carriers who currently lease local network components from Verizon, intervened as Defendants, Counter-Claimants, and Cross-Claimants.

Verizon alleges that the PUC's Order, which represents the commission's third attempt to establish rates for these unbundled network elements ("UNEs") that comport with the requirements of the Telecommunications Act of 1996, 47 U.S.C. § 251 et seq., sets rates that are illegal, unsupported by substantial record evidence, and confiscatory.^[3] The PUC, however, maintains that the rates comply with the Act and are supported by record evidence. For its part, MCI argues that many of the commission's determinations must be reversed both because the PUC improperly based UNE rates on Verizon's actual or overstated costs and because it failed to support its conclusions with substantial record evidence. The parties have filed motions for summary judgment, all of which are presently before the court. For the reasons set forth below, the court will affirm the rates set by the PUC.

I. Background

A. Statutory and Regulatory Framework

Congress passed the Telecommunications Act of 1996 with the objective that "local service, which was previously operated as a monopoly overseen by the several states, be opened to competition according to standards established by federal law." MCI Telecomm. Corp., 271 F.3d at 497. Characterized as an "extraordinary" piece of legislation, the Act sought not merely to balance interests between sellers and buyers, but to meaningfully reorganize utilities markets by rendering monopolies vulnerable to competition. Verizon Communications, Inc. v. F.C.C., 535 U.S. 467, 488-89, 122 S. Ct. 1646, 152 L. Ed. 2d 701 (2002). See also AT & T Corp. v. Iowa Utils. Bd., 525 U.S. 366, 371, 119 S. Ct. 721, 142 L. Ed. 2d 835 (1999) (characterizing the Act as ending the long-standing regime of state-sponsored monopolies by fundamentally restructuring local telephone markets). In order to accomplish this goal of fostering competition, the 1996 Act imposes a series of affirmative duties upon incumbent local exchange carriers ("ILECs"), the "foremost" of which is to share their networks with competing local exchange carriers ("CLECs"). Iowa Utils. Bd., 525 U.S. at 371, 119 S. Ct. 721.

In particular, the Act requires ILECs to lease certain components of their local networks to CLECs on an unbundled basis. 47 U.S.C. § 251(c)(3). In fulfilling this obligation, incumbent carriers must charge rates for UNEs that are "just, reasonable, and nondiscriminatory." Id. In addition, the rates must be "based on the cost (determined without reference to a rate-of-return or other rate-based proceeding) of providing the interconnection or network element." 47 U.S.C. § 252(d)(1)(A). See also Verizon Communications, Inc., 535 U.S. at 467, 122 S. Ct. 1646 (explaining that Congress' novel, rate-setting mandate stood in stark contrast to the familiar public utility model of rate-of-return rate-setting and was set forth in order to give aspiring competitors "every possible incentive to enter local retail telephone markets, short of confiscating the incumbents' property").

Congress directed the FCC to promulgate regulations implementing the substantive requirements of the Act, and the commission responded by issuing an Order, which set forth, among other things, the methodology that state commissions must use in setting UNE rates. 47 U.S.C. § 251(d)(1); In re Implementation of the Local Competition Provisions in the Telecommunications Act of 1996, 1996 WL 452885, 11 F.C.C.R. 15499 (1996) ("Local Competition Order"). This methodology, known as the total element long run incremental cost ("TELRIC") methodology, measures the forward-looking, economic costs of providing a network element because those costs best replicate the conditions of a competitive market. Local Competition Order at ¶ 679. See also Bell Atl.-Del., Inc. v. McMahon, 80 F. Supp. 2d 218, 237 (D.Del.2000) ("[C]osts calculated according to the TELRIC methodology mimic those costs that an efficient company, constrained by competitive market forces, would incur in providing the requested network element."). Thus, rather than determining costs based on an ILEC's actual or embedded costs — which reflect past inefficiencies, older technologies, and outdated operating practices — TELRIC rates are based upon long-run costs in light of "the use of the most efficient telecommunications technology currently available and the lowest cost network configuration, given the existing location of the incumbent LEC's wire centers." 47 C.F.R. §§ 51.505(b)(1), (d)(1).

The FCC has explained that adherence to the TELRIC methodology is critical to achieving the goals behind the 1996 Act because, without it, CLECs' costs of providing local service would be greater than those of incumbents, allowing the established monopolies to remain in place and effectively eliminating any chance for meaningful competition. Local Competition Order at ¶¶ 662-706. Two features of TELRIC are most significant in this regard.

First, TELRIC measures costs in the long run, a time frame lengthy enough to allow all of an incumbent's costs to become variable and, thus, to allow all embedded costs to drop out. Id. at ¶ 677. Second, TELRIC is based not on an ILEC's actual network but instead on a hypothetical network that uses the least cost technology and most efficient design currently available, given the existing location of the ILECs' wire centers. Id. at ¶ 685; Verizon Communications, Inc., 535 U.S. at 522, 122 S. Ct. 1646. Despite these technical features, however, TELRIC is not a specific, mathematical formula but rather a framework of methodological principles that states retain flexibility to use in conjunction with local technological, environmental, regulatory, and economic conditions in order to arrive at forward-looking rates that are both just and reasonable. AT&T Corp. v. F.C.C., 220 F.3d 607, 615 (D.C.Cir.2000). See also AT&T Communications of Ill., Inc. v. Ill. Bell Tel. Co., 349 F.3d 402, 405 (7th Cir.2003) (explaining that TELRIC is a "framework rather than a formula" and that there is "considerable play in the joints").

B. Factual and Procedural History

This case arises from a long-standing dispute between Verizon, the incumbent carrier in the Pennsylvania local telephone market, and CLECs MCI and AT & T over the rates at which Verizon is legally obligated to lease its network elements to the competitors. The PUC has thrice attempted to set these rates, and it is the commission's third and most recent rate-setting Order that is presently before the court.

Shortly after the passage of the 1996 Act and subsequent to various phases of negotiation and arbitration, Verizon and MCI reached an interconnection agreement and submitted it to the PUC for review and approval. The commission conducted rate-setting proceedings, known as "MFS-III", and issued its Final Order on August 7, 1997. Therein, it directed that the rates established during the proceedings — which were based on a cost model submitted by Verizon — be incorporated into the existing agreement. MCI filed suit in the Middle District of Pennsylvania to challenge those rates as incompatible with the TELRIC methodology, and, in an unpublished opinion, that court agreed. The Middle District reached the merits only after finding that the PUC was not entitled to Eleventh Amendment immunity.

Following MFS-III, the PUC initiated a second round of rate-setting proceedings, known as the "Global Proceedings," which resulted in the commission's 1999 Global Order. This time, it was Verizon which filed suit to challenge the new rates as inconsistent with the 1996 Act. On review, this court found that the Eleventh Amendment was not a bar to suit but certified the question for appeal and declined to reach the dispute's merits. Bell Atl.-Pa. v. Pa. Pub. Util. Comm'n, 107 F. Supp. 2d 653 (E.D.Pa.2000). The Third Circuit then handed down companion opinions in the two cases, affirming the twin Eleventh Amendment rulings but reversing in part the Middle District opinion regarding TELRIC based on its finding that the trial court did not properly analyze the substance of the PUC's cost model in light of the 1996 Act. MCI Telecomm. Corp., 271 F.3d at 522-23. It then consolidated the two cases and remanded them to this court for consideration on the merits.

Approximately three months before the Court of Appeals rendered its companion decisions, the PUC began its third generation UNE rate-setting proceeding, known as the "Generic Investigation," and it issued its Final Order in that proceeding on December 11, 2003. Therein, the PUC made final decisions on all of the disputed input and modeling issues that were concurrently before this court. It also directed Verizon to make a compliance filing recalculating the rates based on the final order and offered the parties the opportunity to submit limited comments on that filing.

Eight days later, on December 19, 2003, this court held a hearing and, over the objection of MCI, issued an Order dismissing the consolidated cases as moot. The court agreed with the arguments of both Verizon and the PUC that the commission had already granted the relief sought by the competitors by reviewing the Global Order rates and issuing a Final Order that would change those rates upon compliance therewith.

On July 16, 2004, the PUC issued its final Compliance Order in the Generic Investigation. This Order both resolved the issues raised by the parties' comments to the compliance filing and attached a new schedule of UNE rates that would become effective after Verizon filed a tariff revision, which it was directed to do by August 2, 2004. Verizon complied and began charging the new rates — as directed by the July 16 Order — on October 1 on that same year.

II. Discussion

A. Standard of Review

This court exercises de novo review over whether the PUC's pricing determinations are consistent, as a matter of law, with the 1996 Act and binding FCC regulations. MCI Telecomm. Corp., 271 F.3d at 516-17 (explaining that federal courts owe no deference to the legal determinations of state commissions).

By contrast, the PUC's factual findings are subject only to substantial evidence review. Id. Under this more deferential standard, a reviewing court must affirm a state commission's factual findings if they have "substantial support in the record as a whole." Id. See also GTE S., Inc. v. Morrison, 199 F.3d 733, 745-46 (4th Cir.1999) (holding that the court does not "sit as a super public utilities commission" and that where a decision is supported by the record, "a court is not free to substitute its judgment for the agency's"). Essentially the same as the "arbitrary and capricious" standard, the substantial evidence standard is "the least demanding form of judicial review of administrative action" and requires only that an agency offer a reasoned explanation based on the record evidence for a particular outcome. Mich. Bell Tel. Co. v. Strand, 305 F.3d 580, 587 (6th Cir.2002); MCI Telecomm. Corp., 271 F.3d at 515.

B. Recurring Cost Model

The court will first consider MCI's claim that the PUC's adoption of Verizon's recurring cost model violated TELRIC because that model was illegally based on the incumbent's embedded network. The competitor asserts that even though the PUC made various adjustments to Verizon's model for the specific purpose of bringing it into compliance with federal regulations, no amount of modification could cure the model's inherent infirmities. This court disagrees. The commission's decision was consistent with federal law and is, therefore, affirmed.

Recurring costs are those costs that an incumbent incurs on a monthly basis when leasing UNEs to competitors. As with other types of costs, they must conform to the FCC's TELRIC methodology; that is, they must reflect the forward looking costs that an efficient carrier would incur using the least cost network configuration. 47 C.F.R. §§ 51.505(b)(1), (d)(1). During the Generic Investigation, both sides presented recurring cost models to the PUC. Because the commission found flaws with each submission, it ultimately chose a compromise solution by adopting Verizon's model but adjusting it in several respects with forward-looking modifications in order to bring it into compliance with TELRIC. Final Order at 22, Sealed Joint Appendix ("SJA") at 279.

Although the PUC agreed with MCI's initial argument that Verizon's model did not comport with the FCC's methodology because it was based upon the incumbent's embedded network, the commission concluded that the extensive modifications eliminated the model's inefficiencies and produced rates within the TELRIC range. Id. It explained that its choice was aided by the fact that it had significant reservations about using MCI's model. Id. (noting that it was particularly concerned with the MCI model's apparent inability to calculate costs for advanced services, which are expected to form an important part of the network in the long run).

MCI now submits to this court that despite the adjustments, the PUC's adoption of Verizon's model should be overturned because any model that is based on the incumbent's actual — as opposed to a hypothetical — network, even as a starting point, cannot possibly be modified to become TELRIC-compliant. The competitor, however, cites to no authority in support of this extreme position, and this court can find none. To the contrary, TELRIC does not mandate that state commissions rigidly apply a formula. Instead, it simply provides them with a methodology for establishing just and reasonable rates that approximate what it would cost a perfectly efficient incumbent carrier — given the actual location of that carrier's existing wire centers — to supply UNEs to competitors in a perfectly competitive market. Local Competition Order at ¶ 679. As such, TELRIC is a "framework rather than a formula," and there is "considerable play in the joints." AT&T Communications of Ill., Inc., 349 F.3d at 405.

In light of these principles, it seems clear that TELRIC would not be violated by a cost model that was adjusted to eliminate embedded inefficiencies even though, at the outset, it was based in part on pieces of the incumbent's existing network. This is especially true in light of the fact that the only other option was a model that clearly violated federal regulations by being based exclusively and uncompromisingly on existing infrastructure. Although the model adopted by the PUC started with pieces of Verizon's existing network, the commission concluded that these adjusted inputs — based upon least cost technology and efficient network design — modified that starting point in such a way as to eliminate embedded inefficiencies and produce costs consistent with those that an efficient carrier would incur in the future. Final Order at 22, SJA 279. The elimination of inefficiencies is the driving purpose behind the FCC regulations. Even though the PUC employed some creativity to explore TELRIC's capacity for flexibility, it ultimately constructed a model that was entirely consistent with that underlying purpose.

The PUC's extensive modification of the Verizon model is highlighted by the following forward-looking adjustments: (1) reducing Verizon's proposed cost of capital from 12.95% to 12.37%; (2) adopting FCC prescribed depreciation lives instead of Verizon's proposed, shorter lives; (3) rejecting Verizon's forward looking conversion factor; (4) adopting an 85.5% switch discount rate, which was higher than that submitted by Verizon; and (5) directing Verizon to recalculate its rates for port features. Based on the effectiveness of these adjustments, the court rejects MCI's argument and concludes that TELRIC was satisfied. Having done so, it will now consider the parties' specific challenges to these input adjustments.

1. Inputs

a. Cost of Capital

The PUC made its first adjustment to Verizon's recurring cost model by adopting a 12.37% cost of capital rather than the incumbent's higher proposal of 12.95%. MCI, which argued for a significantly lower 9.54%, contends that the PUC's decision to choose a percentage so close to the figure proposed by Verizon violated TELRIC principles. This court disagrees. The PUC's determination with respect to the cost of capital input complied with the FCC's regulations and was supported by substantial record evidence. Therefore, it is affirmed.

Cost of capital is the return that an investor can expect on an investment in a given enterprise as that enterprise raises capital to finance its operations. Once the cost of capital is determined, it is applied as a percentage markup to the costs of all UNEs that a competitor may obtain from an incumbent under the Act's unbundling regime. It is thus a crucial input in every rate-setting proceeding because it affects costs across the board.

Any calculation of cost of capital is comprised of three inputs: the cost of debt, the cost of equity, and the debt to equity ratio. In the proceeding below, the PUC arrived at a 12.37% cost of capital, which was comprised of a 7.86% cost of debt, a 14.75% cost of equity, and a ratio of 34.5% debt to 65.5% equity. Final Order at 62, Joint Appendix ("JA") 419. The issue before this court concerns the cost of equity, which is the most difficult factor to determine because, unlike debt, which can be measured using long-term interest rates, it requires economic modeling to forecast a company's long-term market performance. MCI argues that the cost of equity figure adopted by the PUC — 14.75% as opposed to the 10.42% that it proposed — violated federal law both because it was impermissibly based on Verizon's short-run, as opposed to long-run, cost of equity and because it did not take into account the risk assumptions required by TELRIC. For the reasons set forth below, the court disagrees.

TELRIC dictates that UNE rates be set according to a forward-looking framework that approximates an incumbent's long-run costs. Local Competition Order at ¶ 677. As such, the FCC has instructed that the cost of capital should "reflect the competitive risks associated with participating in the type of market that TELRIC assumes." In Re Review of the Section 251 Unbundling Obligations of Incumbent Local Exchange Carriers, 2003 WL 22175730, at ¶ 681, 18 F.C.C.R. 16978 (2003) ("Triennial Review Order"). Thus, in determining cost of capital, state commissions should assume a telecommunications market "in which facilities-based carriers would risk losing customers to other facilities-based carriers." Id. at 2003 WL 22175730, ¶ 680, 18 F.C.C.R. 16978. The appropriate level of risk is, therefore, not the actual competitive risk that an ILEC faces at the present time but rather the risk associated with TELRIC's competitive assumptions in the long run. Id. at 2003 WL 22175730, ¶¶ 678-80, 18 F.C.C.R. 16978. According to the FCC, increased competition leads to increased risk, which, in turn, warrants an increased cost of capital. Id. at 2003 WL 22175730, ¶ 681 18 F.C.C.R. 16978. See also Verizon Communications, Inc., 535 U.S. at 520, 122 S. Ct. 1646 ("[T]he Commission specifically permits more favorable allowances for cost of capital ... than were generally allowed under traditional ratemaking practice.").

MCI first argues that the cost of capital proposal adopted by the PUC violated TELRIC because it was not forward-looking. Although cast as a legal argument, MCI is actually contending not that the PUC failed to utilize a long-run framework but rather that its determination of what conditions will apply in the long run was incorrect. As such, this is a factual argument that the court must reject because the PUC's determination was supported by substantial record evidence.

During the proceedings below, the PUC was presented with two versions of the Discounted Cash Flow ("DCF") equation, which translates market data such as share prices and dividend yields into a cost of equity percentage. The first version, proposed by Verizon and dubbed the "single-stage model," assumed that the incumbent's current growth rate will remain constant in the long run. The second, endorsed by MCI and known as the "multi-stage model," used the projected overall growth rate of the economy to forecast a firm's current growth rate over the long term. The PUC adopted the former and explained its decision to do so in light of TELRIC principles and the record evidence. Final Order at 60-61, JA 418-19.

The essence of MCI's argument is that the PUC's adoption of the single-stage model violated TELRIC because it is inevitable that Verizon — as a telecommunications company in a relatively new and rapidly expanding industry — will not be able to sustain its current high growth rate as the market saturates in the long run. The PUC, however, found otherwise and determined Verizon could indeed sustain its current rate of growth in the type of long-run, competitive market assumed by TELRIC. Such a conclusion was supported by substantial record evidence. See, e.g., Verizon Stmt. 4.1 (Rebuttal Testimony of James H. Vander Weide) at 52, JA 2919 (offering the PUC an explanation as to why the single stage model is a reasonable approximation of reality even though firms cannot grow at the current rates forever). Indeed, there were dozens of pages of testimony devoted to the relative merits of the single-stage as opposed to the multi-stage DCF model, and the PUC considered the evidence and reasonably adopted the former. The mere fact that the commission did not agree with MCI's forecast of events in the long run is not grounds for reversal.

MCI's second argument is that the PUC's cost of capital determination violated federal law because the commission did not make the risk assumptions required by TELRIC. In particular, MCI contends that the PUC erred both by assuming that Verizon would be engaged only in the provision of UNEs and not in facilities-based competition and by adopting Verizon's proxy group for measuring forward-looking risk. Neither of these reasons, however, can convince the court to disturb the PUC's determination.

MCI contends that the PUC violated TELRIC by assuming that Verizon would be engaged exclusively in the provision of UNEs, an assumption that would inflate costs because UNE provision carries a higher risk than facilities-based competition. This argument, while plausible, must fail because it mischaracterizes the PUC's holding. Contrary to MCI's claim, the commission did not make such an assumption. Instead, the PUC properly adhered to the FCC standard and attempted to find a level of risk consistent with facilities-based competition in the type of market that TELRIC assumes.

Specifically, the PUC found that "an appropriate common equity cost rate ... should ... reflect those risks associated with a firm engaged solely in a competitive market for facilities-based telecommunications service." Final Order at 60, JA 417 (emphasis added). See Triennial Review Order at ¶¶ 678-80. The commission heard testimony that there are no publicly traded companies devoted solely to UNE provision. As such, it reasonably adopted a proxy group of companies that it found to face risks comparable to those that would confront a company in a telecommunications market with facilities-based competition. Final Order at 61, JA 418 ("The telecommunications industry is consistently evolving, and at this time the return associated with [the proxy group chosen] reflects the best proxy of the investment risk, and therefore, resulting return that a TELRIC-style competitive entity would experience."). See also Verizon Stmt. 4.0 (Direct Testimony of Dr. James H. Vander Weide) at 32, JA 2846 (explaining that there are no publicly-traded companies whose sole business is offering UNEs); Verizon Stmt. 4.2 (Surrebuttal Testimony of Dr. James H. Vander Weide) at 28-35, SJA 2053-60 (explaining why Verizon's proxy group was superior to that submitted by MCI). This approach was both consistent with the TELRIC standard and supported by substantial record evidence. It is, therefore, affirmed.

b. Depreciation Lives

The second modification made by the PUC to Verizon's recurring cost model was the adoption of FCC-developed, regulatory depreciation lives proposed by MCI, as opposed to the economic depreciation lives proposed by Verizon. The incumbent argues that the PUC's decision to adopt the former was both arbitrary and capricious and that it violated TELRIC by resulting in rates substantially lower than legally required. The court, however, finds Verizon's argument unconvincing and holds that the PUC's decision was both consistent with federal law and supported by substantial record evidence.

Like cost of capital, depreciation lives are an important input to the calculation of recurring rates because they affect an incumbent's overall costs. A depreciation life measures the time period over which a company's capital assets are assumed to have economic value and allows it to recover the cost of its investment in those assets over time. If a company is required to depreciate an asset over too long a time period — a life extending beyond the time during which the asset actually has economic value — then the company will be unable to fully recover its costs. On the other hand, if the depreciation life is too short and, therefore, assumes that a capital asset is worthless even when it retains real-world value, the company will recover the difference and be over-compensated for its investment. As such, longer lives result in lower UNE rates, while shorter lives bring about higher ones. In essence, Verizon contends that the lives adopted by the PUC were too long and resulted in rates that were too low.

As with all inputs, depreciation lives must comply with TELRIC methodology. The FCC has explained that "in calculating depreciation expense ... the rate of depreciation over the useful life [of an asset] should reflect the actual decline in value that would be anticipated in the competitive market TELRIC assumes." Triennial Review Order at ¶ 689. The commission has also advised that "state commissions continue to have discretion" in determining what depreciation lives comply with this standard, and, as such, it has declined to endorse either economic or regulatory lives as more likely to produce a TELRIC-consistent result than the other. Id. at ¶ 688. Thus, state commissions are free to adopt either, as long as they reflect the forward-looking decline in value that would result in a competitive market.

The PUC was presented with two sets of depreciation lives during the proceeding below. On the one hand, Verizon advocated the use of economic depreciation lives, which are generally used for financial accounting purposes and which the PUC had adopted during its MFS-III and Global proceedings. The incumbent argued that these lives — which were shorter and thus produced higher rates than those proposed by MCI — were TELRIC-compliant because they were based upon a competitive market in an industry characterized by rapid technological developments. On the other hand, MCI advocated the use of longer regulatory lives, which had been set by the FCC in 1995 and which resulted in substantially lower rates. Departing from both its prior conclusions and the recommended decision of the ALJ (which was based in large part on adhering to those conclusions), the PUC chose the FCC lives proposed by MCI. In so doing, the commission found that "the FCC-prescribed depreciation lives are more reflective of the forward-looking, economic criteria required by the FCC rules." Final Order at 62, JA 419. Verizon challenges this determination on both legal and factual grounds.

The incumbent's first argument is that the PUC's decision to adopt the FCC regulatory lives sponsored by MCI must be reversed because those lives are not TELRIC-compliant. In support of this position, Verizon contends that because the regulatory lives were developed in 1995 — before the passage of the 1996 Act and the adoption of the TELRIC methodology — they must be rejected as outdated. It also points to language in the FCC's Triennial Review Order suggesting that an accelerated depreciation mechanism may be a more accurate means of measuring the useful life of an asset in the type of market TELRIC assumes. The court, however, rejects these contentions and holds that the FCC lives comply with federal law.

Verizon's argument that the FCC regulatory lives are obsolete simply because they pre-date the 1996 Act must be rejected. While it is true that the commission developed its regulatory lives in 1995, it reviewed and updated those lives four years later, well after the Act's passage. Recommended Decision at 22, SJA 182 (advising the PUC that Verizon's argument that the regulatory lives established in 1995 were outdated was "less than candid" because they had been reviewed and modified by the FCC in 1998-99). Furthermore, in 2003, the FCC conducted rate-setting proceedings in Virginia^[4] and, when faced with the same arguments as those presented by Verizon here, reached the same conclusion as did the PUC. In particular, it rejected Verizon's argument that the FCC regulatory lives were not sufficiently forward-looking and adopted the 1995 lives, which were "the most recent ones prescribed by the Commission." In re Petition of WorldCom, Inc. Pursuant to Section 252(e)(5) of the Communications Act for Preemption of the Jurisdiction of the Virginia State Corporation Commission Regarding Interconnection Disputes with Verizon Virginia Inc. and for Expedited Arbitration at 2003 WL 22038242, ¶ 115, 18 F.C.C.R. 17722 (2003) ("Virginia Pricing Order"). In addition, approximately 20 other states have used these FCC regulatory lives to calculate depreciation. Tentative Order at 48, SJA 305.

In response to this evidence tending to show that the FCC lives remain acceptable to calculate depreciation, Verizon makes a host of arguments, the most significant of which relates to the Virginia Pricing Order. In essence, the ILEC contends that this court should not rely upon the Virginia Pricing Order — both because it was handed down by the FCC's Wireline Bureau and, therefore, does not constitute agency policy, and because it was legally erroneous — and that, even if it does, it should distinguish that decision from the instant case. These arguments, however, miss the mark.

The FCC opined in its Triennial Review Order that regulatory lives may be TELRIC-compliant, and it did so with the knowledge that its 1995 lives were the most recent regulatory lives it had established. Triennial Review Order at ¶ 688. Thus, the fact that the 1995 lives could comply with TELRIC under certain circumstances was implicit in the Triennial Review Order. With this FCC guidance as a backdrop,^[5] the PUC carefully considered the record evidence and found that the 1995 lives did indeed comply with TELRIC. That the Commission reviewed and modified those lives in 1999 — and that its Wireline Bureau found them to comport with TELRIC in another context in 2003 — merely supports, but was not necessary to, the PUC's determination that, based on the record evidence, the FCC regulatory lives were lawful.

Verizon also argues that the FCC's statement that "the use of an accelerated depreciation mechanism may present a more accurate method of calculating economic depreciation" supports the proposition that only lives shorter than the FCC's regulatory lives would comply with TELRIC. Id. at ¶ 688 (emphasis added). The incumbent's reliance on this statement, however, is misguided. The FCC clearly phrased this sentence not as a directive but instead as a suggestion offered to aid state commissions in the exercise of their discretion to choose any depreciation lives that comply with federal law. Furthermore, even if this court were to hold that the clearly suggestive language actually sets forth a mandate, the PUC could not have complied therewith because Verizon never even proposed that the commission adopt an accelerated depreciation mechanism. Thus, Verizon's argument based on this language from the Triennial Review does not persuade the court that the regulatory lives adopted by the PUC violated TELRIC.

In addition to arguing that the FCC lives are unlawful, Verizon also contends that the PUC's decision to adopt those lives was arbitrary and capricious. This court, however, finds otherwise and affirms the commission's decision as supported by substantial evidence. The PUC was presented with considerable evidence from both sides with respect to depreciation lives, and it concluded that the weight of that authority supported the adoption of the FCC lives. Tentative Order at 47-49, JA 304-06. The commission did not do so arbitrarily but instead specifically explained that it reversed its prior precedents and disagreed with the ALJ's recommended decision because it found, for a host of enumerated reasons, that the FCC lives were more consistent than Verizon's proposed economic lives with TELRIC's forward-looking assumptions. Id. at 49, JA 306. Because this determination was supported by substantial record evidence, it is affirmed. See, e.g., MCI Stmt. 2.2 (Surrebuttal Testimony of Terry L. Murray) at 36-38, JA 4821-23 (explaining that, contrary to Verizon's assertion, technological innovation in the telecommunications industry could lengthen depreciation lives and that, therefore, the FCC regulatory lives were appropriate inputs for the cost model); MCI Stmt. 5.1 (Direct Testimony of Richard B. Lee) at 4-7, JA 3751-54 (explaining that financial book reporting lives such as those advocated by Verizon are designed to protect investors and would, therefore, be inappropriate to calculate depreciation expense in a regulatory context).

c. Switching Rates

The third modification made by the PUC to Verizon's recurring cost model was to adopt MCI's proposed switch discount mix. Verizon argues that this decision resulted in illegally low switching rates and was unsupported by the record. The court, however, disagrees and will affirm the PUC's decision as both consistent with TELRIC and supported by substantial evidence.

Carriers can purchase switches — critical pieces of a telephone network that route calls to their intended destinations — in one of two ways: as brand-new switches or, in the alternative, as add-on, or growth switches, which either upgrade or increase the capacity of existing switches. McMahon, 80 F.Supp.2d at 236. A typical vendor offers new switches at deep discounts in order to make buyers dependent on its particular equipment so that they will upgrade their networks by purchasing its more expensive add-on switches. Tentative Order at 130, SJA 387. Thus, an efficient carrier will attempt to purchase as many new switches as possible at the deep discounts but will necessarily be forced to purchase a certain number of growth switches at the less discounted prices.

In order to set the switching rates in the proceeding below, the PUC was called upon to determine exactly what percentage of an efficient carrier's switches would be purchased at each of these discounts. Predictably, MCI advocated a mix weighted towards new switches — a proposal which would result in lower rates — while Verizon argued that it would be more realistic to assume that a carrier would be forced to purchase a greater number of growth switches — an assumption that would result in higher rates. After being presented with alternative proposals from both sides, the commission adopted an MCI-sponsored version, which contained a mix of 85.5% new switch discounts and 14.5% add-on switch discounts. Id. at 137-38, SJA 393-94. This decision both complied with TELRIC and was supported by substantial evidence.

TELRIC requires that costs be modeled on a forward-looking, long-term basis. 47 C.F.R. 51.505(b). Accordingly, UNE rates — switching rates included — should reflect the needs of an efficient carrier using the latest technology in a fully competitive environment. Id. Thus, under TELRIC methodology, the cost of an incumbent's existing network is irrelevant, and state commissions must, therefore, forecast a carrier's long-run purchase of switches without regard to previous investments. Although the FCC has not endorsed as optimal a specific numerical mixture of new and growth switches, it is clear that TELRIC assumes that an efficient carrier would incrementally build its switch network with a mixture of the two, as opposed to 100% of one or the other. See e.g., AT&T Corp., 220 F.3d at 616-18 (upholding FCC approval of switching rates based upon the assumption that carrier would purchase a mix of new and growth switches); In re Review of the Commission's Rules Regarding the Pricing of Unbundled Network Elements and the Resale of Service by Incumbent Local Exchange Carriers, 2003 WL 22119504, ¶ 77, 18 F.C.C.R. 18945 (2003) (noting that the FCC has rejected an assumption that the appropriate switching discount must be based on a purchase of 100% new switches); In re Joint Application by BellSouth Corp. et al. for Provision of In-Region, InterLATA Services in Ala., Ky., Miss., N.C., & S.C., 2002 WL 31084940, ¶ 80, 17 F.C.C.R. 17595 (2002) ("[S]witching prices may be based on a meld of new and growth discounts ... [C]ertain vendors have provided a greater discount for new switches and smaller discounts for growth or expansion of existing switches ... [S]uch discounts were only valid when an overall purchase of both new and growth equipment was made.").

During the proceeding below, both sides presented the PUC with alternative proposals, which included an MCI-sponsored model that assumed a 100% new switch discount and one submitted by Verizon that was comprised of 97% growth switches. The PUC, however, rejected both of these extremes and settled on a model in between the two. Proposed by MCI, the compromise model adopted by the commission assumed that an efficient carrier entering the market would purchase a partial network of all new switches and would then add on to that network with growth switches as the original switches became ineffective or obsolete. Although the original purchase at the moment of market entry would be of 100% new switches, those new switches would comprise only 85.5% of the carrier's switching network in the long-run. The model assumed that the competitor would make up the balance of its long-run network by purchasing growth switches.

Verizon attacks the PUC's adoption of this model as inconsistent with TELRIC and advances two main arguments in support of its position. First, the incumbent argues that because the model assumed that an efficient market entrant would purchase all new switches at the outset, it violated TELRIC because that methodology mandates that carriers purchase a mix of new and growth switches. This argument, however, confuses the issue. Although the model assumes 100% of a carrier's initial purchase would be comprised of new switches, that initial purchase would make up only 85.5% of the long-run switching network. Thus, the authority cited by Verizon in support of the proposition that a carrier cannot be assumed to purchase 100% new switches is inapposite. The model selected by the PUC does indeed contemplate a mix of new and growth switch purchases over the long run and, therefore, is consistent with TELRIC principles. Tentative Order at 137, SJA at 394 (adopting an approach of melding new switch and growth discounts by using a figure based upon the discounts expected for the purchase of an average switch and growth additions over the switch life).

Second, Verizon contends that the model violates TELRIC because its assumption that a carrier could purchase new switches that comprise 85.5% of its long-run network at deeply discounted rates is unrealistic. It argues that because vendors typically offer large discounts on new switches in order to lock carriers in such that they will be forced to purchase growth switches at less discounted rates, no rational vendor would be able to offer these deep discounts on such a large percentage of the overall anticipated purchase. See AT&T Corp., 220 F.3d at 618 (referencing the FCC's argument that "growth additions to existing switches cost more than new switches only because vendors offer substantial new switch discounts in order to make telephone companies dependent on the vendors' technology to update the switches"). Verizon bases its argument on the claim that the largest discounts it can expect to receive are based on an average of 50% new and 50% growth switches. This projection, regardless of its accuracy, is irrelevant to the court's analysis.

TELRIC contemplates not what prices an existing carrier — bound by its embedded inefficiencies and previous investments — could actually receive, but instead what vendors would charge an efficient carrier constructing a new, cost-effective network using the most efficient technology available. 47 C.F.R. 51.505(b)(1). The PUC complied with this standard when it adopted MCI's switch discount model. Tentative Order at 139-40, SJA 396-97 (explaining that the touchstone of forward-looking pricing is not Verizon's existing network but what an efficient provider would do if unconstrained by previous investments and concluding that MCI's alternative proposal is consistent with TELRIC). Its determination is, therefore, affirmed as consistent with federal law.

In addition to attacking the PUC's decision on switching rates as illegal, Verizon also argues that it was arbitrary and capricious. This contention, however, lacks merit because the commission's decision was supported by substantial record evidence. Indeed, MCI specifically addressed the merits of its model and explained why the PUC should adopt it over Verizon's proposals. Joint Exceptions of AT & T and MCI at 49-52, JA 3485-88. In addition, the commission considered testimony that criticized Verizon's proposed models as neither plausible in practice nor compliant with TELRIC principles. MCI Stmt. 2.1 (Rebuttal Testimony of Terry L. Murray) at 33-35, SJA 3840-42 (demonstrating the problems with Verizon's proposals and explaining that it is implausible that an efficient carrier would enter a competitive market by purchasing bulk quantities of growth switches instead of more heavily discounted new switches). As such, the court affirms the PUC's decision to adopt MCI's proposal in lieu of Verizon's because it was supported by substantial record evidence.

d. Common Overhead Costs

The PUC's fourth modification to Verizon's recurring cost model was to reject the incumbent's Forward-Looking Conversion Factor ("FLC"), a mathematical device designed to increase costs and, thereby, increase UNE rates. Although this adjustment would theoretically benefit MCI by decreasing costs, it is the competitor which raises a challenge here because it contends that Verizon employed the FLC in its calculations despite the PUC's clear directive to the contrary. It argues both that the FLC was illegal because it drove costs up to their embedded levels and that the PUC's determination was arbitrary and capricious. This court disagrees. For the reasons set forth below, it will reject those arguments and affirm the PUC's decision.

MCI first claims that the FLC was illegal because it impermissibly inflated the cost of Verizon's Common Overhead Factor ("COH"), an Annual Cost Factor ("ACF")^[6] that measures overhead costs for executive salaries, external relations, human resources, legal expenses, and other general and administrative functions. This argument, however, lacks merit. Although application of the FLC did increase Verizon's COH, it was nonetheless consistent with TELRIC methodology.

During the proceeding below, the PUC considered extensive evidence with respect to the application of the FLC to all ACFs. On the one hand, Verizon argued that because its current annual expenses were already made forward-looking by application of productive and inflation factors, the cost-inflating FLC was necessary to avoid a double TELRIC adjustment. On the other, MCI contended that Verizon's original expenses were not forward-looking and that they, therefore, needed to be adjusted downward — and not upward — to comply with TELRIC. The PUC agreed with MCI and directed Verizon to rerun its cost study after applying the FLC in a different manner. Final Order at 40-41, JA 397-98 (explaining that multiplying costs by the .692 FLC was permissible but that later dividing them by that same number was error because it canceled the effects of the multiplication). Noting that it was not rejecting the FLC "in all respects" but instead as implemented in Verizon's COH calculations, the PUC concluded that the incumbent's compliance with its order would result in "an overall reduction" of Verizon's UNE rates. Id. at 38, 39, JA 395, 96.

Verizon then returned to the PUC with new calculations that it maintained were consistent with the commission's instructions, and, as the PUC had predicted, this new math resulted in an overall reduction of the other ACFs by between 12% and 18%. Id. at 39, JA 396; Verizon's Reply Comments Regarding Compliance Filing at 2, JA 588. Nevertheless, MCI argued that Verizon failed to comply with the PUC's directive because it still employed the FLC in calculating its COH, which increased by almost 2% with the new calculations. Compliance Order at 8-9, JA 462-63. Although the competitor contended that this increase violated TELRIC because the FLC was based on Verizon's actual costs and, therefore, drove the COH up to its embedded levels, the PUC rejected this argument. This court agrees and will affirm the PUC's determination as consistent with the applicable law.

The commission concluded that Verizon complied with its instructions and properly applied the FLC based on its finding that "the estimation of overhead expense in a TELRIC environment should not reasonably differ from those expenses in the present environment." Id. at 8, JA 462. For example, the PUC explained, because common costs such as those for executive salaries and corporate offices do not necessarily decrease with the advent of newer technologies, basing overhead expenses on existing costs was indeed forward-looking. Id. This reasoning was sound.

Contrary to MCI's suggestion, the fact that the PUC based its COH input on Verizon's actual costs does not automatically render its decision legally deficient. Although TELRIC methodology mandates that costs be based on those that would be incurred by a hypothetical, most efficient carrier, that requirement is logically satisfied where the actual ILEC's costs are found to be the same as those of the hypothetical carrier. Local Competition Order at ¶ 685. Here, the PUC found that an efficient competitor's hypothetical, forward-looking overhead costs would be approximately the same as those actually incurred by Verizon. Thus, TELRIC was satisfied. Furthermore, the fact that the ACFs were reduced in aggregate after Verizon completed its new calculations corroborates the PUC's finding that Verizon did indeed follow its instructions and produce rates within the TELRIC range.

MCI argues in the alternative that even it this court finds that the FLC complied with TELRIC, it should nevertheless conclude that the common cost calculation violated federal law because it impermissibly included common costs that can be attributed in part to Verizon's retail operations. See 47 C.F.R. § 51.505(2)(d)(2) (prohibiting inclusion of any retail costs in calculation of common costs). In support of this argument, MCI relies on AT & T Communications of Cal., Inc. v. Pac. Bell Tel. Co., in which the Ninth Circuit reversed the California Public Utilities Commission's ("CPUC") calculation of common costs as violating § 51.505(2)(d)(2) even though it excluded retail-only common costs because it nevertheless included overhead costs common to both retail and wholesale operations. 375 F.3d 894 (9th Cir.2004). The incumbent essentially argues that the PUC committed the same error in the present case. This court, however, declines to follow the Ninth Circuit's decision because it rested upon a faulty interpretation of TELRIC.^[7] As such, it will reject MCI's argument.

In AT & T Communications of Cal., the Ninth Circuit explained that a common cost mark-up — which allows incumbents to recover all costs properly attributable to the provision of a given UNE — should be calculated by dividing the firm's costs common to all of its wholesale operations by its direct costs associated with providing that UNE on a wholesale basis. 375 F.3d at 905. This "apples to apples" calculation, with wholesale common costs in the numerator and wholesale direct costs in the denominator, the court opined, is consistent with TELRIC. Id. Despite the fact that the CPUC employed such a formula, the Ninth Circuit overturned its decision because the commission assumed that the incumbent engaged only in wholesale — as opposed to retail — operations. Id. at 906. The court explained that this assumption was not required by — and indeed violated — TELRIC because it inflated the numerator by attributing to wholesale operations common costs that were, in reality, partly attributable to the firm's retail business, which was assumed by the commission to be non-existent. Id.

While this court agrees with the Ninth Circuit that TELRIC does not require a hypothetical, wholesale-only environment, it disagrees that such an assumption is necessarily unlawful. To the contrary, federal law seems to suggest that it is at least permissible. In particular, the FCC's regulations provide that the marked-up cost of a UNE shall not exceed the total forward-looking costs that would be incurred by an efficient firm that produced only that element. 47 C.F.R. § 51.505(c)(2)(ii). Because a firm producing one and only one UNE would naturally provide that element to other carriers on a wholesale basis — as opposed to a firm that provided various telephone services to individual and business customers on retail bases — this court cannot conclude that the assumption of a wholesale-only environment violates TELRIC. As such, it declines to follow the Ninth Circuit's decision.

Furthermore, after its review of the record, this court is satisfied that the approach taken by the PUC — an approach which applied a "retail-avoided cost percentage" to effectively eliminate retail costs from the common overhead figure — was consistent with federal law. See 47 C.F.R. § 51.505(2)(d)(2) (retail costs not to be considered in calculation of common costs); Verizon Stmt. 1.0 (Direct Panel Testimony on Recurring Costs) Attachment B at 17, SJA 982 (explaining the method used for the "avoidance of retail-related costs" in the FLC). Thus, the commission's determination is affirmed.

In addition to attacking the legality of the FLC, MCI also attacks the PUC's decision as arbitrary and capricious. The competitor argues that the commission's approval of Verizon's use of the FLC factor in its Compliance Order should be reversed both because it was inconsistent with the PUC's conclusion in its Final Order that the factor violated TELRIC by raising costs to their embedded levels, and because it was based on a theory that the PUC articulated for the first time in the Compliance Order. The court disagrees. As the court has already explained, the PUC in its Final Order rejected the FLC factor as implemented in Verizon's filing, and it set forth instructions as to how the incumbent could cure the infirmity by rerunning the cost model to arrive at an overall decrease in the ACFs. Verizon, in turn, reran its model and claimed to have followed those instructions, and the PUC determined after careful consideration that it had indeed done so. That it based its finding of compliance on what MCI characterizes as a novel theory is irrelevant to the analysis because the PUC explained its rationale, and this court has upheld that rationale TELRIC-compliant. The commission's determinations with respect to the FLC were neither arbitrary nor capricious and are, therefore, affirmed.

e. Port Features

Fifth and finally,^[8] the PUC modified Verizon's recurring cost model by rejecting the incumbent's proposal to change the port feature structure from that set by the 1999 Global Order. Final Order at 49, JA 406. Verizon argues that this decision violated federal law because it prevented the incumbent from recovering all of the costs that it would incur in providing certain port features to competitors. 47 U.S.C. § 252(d)(1)(A)(i) (requiring UNE rates to be "based on the cost" of providing the network element). The court, however, rejects this contention based on its finding that the PUC's decision was supported by substantial evidence.

Although set forth as a legal argument, the essence of Verizon's claim is not that the PUC applied an improper framework that would preclude the incumbent from recovering its costs, but instead that the Commission arbitrarily rejected Verizon's submission as to what its actual costs of providing the port features would be. As such, this is a factual argument, and the court will treat it accordingly. See MCI Telecomm. Corp., 271 F.3d at 517 (factual findings of state commissions must be affirmed if they have "substantial support in the record as a whole").

Ports are line terminations in a carrier's central office switch that provide user customers with dial tones. As technology has advanced, carriers have added various features to ports — well-known examples of which include Call Waiting and Caller ID — to increase their functionality. Incumbents can either package some or all of these features and sell them to competitors along with the ports themselves, or they can offer the ports and features for sale separately. The structure dictated by the Global Order was a two-tiered approach: Verizon was to offer a full feature port at $2.67 per month and a limited feature port — which excluded four features^[9] that were offered in the full feature edition — at the monthly rate of $1.90. Final Order at 48, JA 405.

Despite the fact that the two-tiered structure set by the Global Order remained in effect at the time of the Generic Investigation — and despite the fact that the purpose of the new proceeding was to review the existing rates as opposed to the existing structure — Verizon's initial filing therein contained a proposal for a substantially different port structure. In particular, the incumbent suggested a lower port rate that reflected only the cost of the port itself and then sought to offer every additional feature for sale separately based on its individual cost. Verizon Recurring Cost Summary at 28, 53 JA 3793, 3818. The incumbent, however, failed to highlight this unexpected change, and both the PUC and MCI claim that they did not notice it during the early stages of the Generic Investigation. PUC Opposition Brief at 23; MCI Opposition Brief at 50.

Indeed, MCI claims that it did not become aware of the revised port structure — which consisted of a full feature port nearly twice as expensive as the Global Order rate and a limited feature port that excluded a host of features beyond those left off of the Global Order limited port offering — until Verizon's December 4, 2002 filing. The competitor then submitted objections to the PUC, which, after finding that Verizon's proposed full and limited port rates represented 77% and 138% respective increases over the current rates, directed Verizon to retain the same port feature structure as set forth in the Global Order. Final Order at 48-50, JA 405-07.

In response, Verizon submitted two more filings — one on January 26, 2004, which provided for a limited port that excluded six more features^[10] than had been excluded by the Global Order's limited port offering, and one of March 8 of that same year, which revised the language for the full feature port and removed those same six features from that offering as well. Compliance Order at 13, JA 467. MCI objected and argued that Verizon had failed to comply with the PUC's directive to revert to the Global Order structure and rates.

Verizon, however, maintained that it had indeed complied with the PUC's mandate. In its comments to the March 8 filing, the incumbent explained that despite its excluding the six features from both its full and limited feature ports, its filing was nevertheless consistent with the Global Order because "at the time of the Global Order these features were quite new and available only in limited areas." Verizon's Reply Comments Regarding Compliance Filing at 6, JA 592. As such, Verizon argued, their costs were not included in the Global Order rates and, in order to comply with federal law, those costs had to be recovered by charging for the new features on an a la carte basis. Id. at 7, JA 593. See 47 U.S.C. § 252(d)(1)(A)(i) (requiring UNE rates to be "based on the cost" of providing the network element).

The PUC rejected Verizon's argument and found that MCI's contrary assertion that the excluded features were not new but instead had been offered for years in Pennsylvania "raised substantial questions of fact" and concluded that, because the issue had been raised during the late stages of the proceedings, Verizon's exclusion of the six additional factors was improper. Compliance Order at 15-16, JA at 469-70. It explained that despite Verizon's argument that it used a new port rate structure early on in the proceedings, "it appears from our review of the record in these proceedings that the port rate discussions all proceeded on the assumption that the two-tiered structure of the Global Order was not modified." Id. at 16, JA 470. Based on that assumption, the PUC rejected Verizon's argument that it was unlawfully prohibited from recovering its costs for the six features at issue. Although Verizon argues that the PUC's decision was arbitrary, this court concludes otherwise. The commission's rejection of Verizon's contention was both supported by substantial evidence and reasonable on the record before it.

Despite Verizon's argument that it presented clear and detailed evidence on the costs of the six excluded features throughout the proceeding, this court's review of the record supports the PUC's finding that the port rate discussions proceeded on the assumption that the Global Order structure would remain intact. Indeed, the testimony that Verizon cites in support of its position is somewhat misleading. Although the incumbent characterizes that testimony as specifically explaining that it "had proposed a different port structure from that in its existing tariffs," the testimony on record provides that the Generic Investigation structure was "consistent but not identical" to that of the Global Order and that it was "consistent with the rate structure used in New York that was cited in the Global Order." Compare Verizon Reply Brief at 40, with Verizon Stmt. 1.0 (Direct Testimony Panel Testimony on Recurring Costs) at 83-84, JA 871-872. The testimony suggests that there existed only minor, almost inconsequential differences between the two structures, and the PUC cannot be faulted for relying on Verizon's representations in the context of such a large and complex proceeding. The fact that the specific purpose of the proceeding was to challenge the Global Order rates — and not the port structure — only underscores the reasonableness of the PUC's reliance.

Because the PUC acted reasonably in not detecting Verizon's proposed changes to the port structure until late in the proceedings, the incumbent's argument that the commission arbitrarily rejected its representation that it was legally required to charge separately for the six additional port features must fail. The PUC found that the determination of Verizon's costs for those features turned on a disputed factual issue — i.e., whether the features were "new" at the time of the Global Order or whether they were already included in the Global Order rates — and it reasonably refused to delay the proceedings and reopen the evidentiary record to resolve that dispute. Given the circumstances, this decision to reject Verizon's argument in favor of the Global Order rates that represented the status quo was neither arbitrary nor capricious. It is, therefore, affirmed.

2. Outputs

In addition to challenging the PUC's modification of various inputs in its recurring cost model, Verizon also claims that the commission erred by rejecting the rates that resulted from the running of that adjusted model for certain UNEs.^[11] The incumbent argues that the PUC's rejection of these outputs in favor of the existing rates was arbitrary and capricious and should, therefore, be reversed. The court disagrees. The PUC acted reasonably and within its authority in requiring Verizon to explain the large rate variances that resulted from running the cost model, and its decision to accept some of the ILEC's explanations and reject others was supported by substantial record evidence. In addition, its approach adhered to TELRIC methodology.

After Verizon ran its cost model, the PUC reviewed the results and found that a number of the outputs varied considerably from the existing Global Order rates. Because of the large and unexplained discrepancies, the commission questioned the accuracy of these outliers and directed Verizon to provide an explanation supported by documentation for each UNE rate that varied by 50% or more, in either direction, from the rates that were then in existence. Tentative Order at 44-45, JA 401-02. See also 47 U.S.C. § 252(b)(4)(B) (granting state commissions the authority to require the parties to provide information necessary for them to reach decisions as to unresolved issues). The PUC noted that it was particularly troubled by the increases — which included between an 843% and 2028% spike in the rates for Unbundled EEL Testing — because they were atypical of a declining-cost industry. Tentative Order at 42-43, JA 399-400. The commission explained that its decision to seek explanations for the large variances instead of accepting the outputs without question was consistent with TELRIC methodology, which requires not the rigid adherence to a mathematical formula but rather the use of forward-looking principles to arrive at cost-based rates that are both just and reasonable. Id. at 44, JA 401. See also 47 U.S.C. § 252(d)(1); AT&T Corp., 220 F.3d at 615. This approach was sound and comported with federal law.

Verizon responded by letter to the PUC's request, and the commission upheld a number of the rate variances based upon its finding that the incumbent had satisfactorily justified them. Letter from Verizon to PUC Secretary dated March 8, 2004 at J.A. 4764-77; Compliance Order at 19-20, JA 473-74 (upholding, for example, the rate increase for the Network Interface Device UNE after finding that Verizon had justified it by citing specific, relevant documentation). With respect to other UNEs, however, such as those at issue here, the PUC found that Verizon's brief explanations were insufficient and lacked supporting data; therefore, it rejected the new rates and ordered that the remaining rates be kept in effect. Id. at 18, 21-22, JA 472, 475-76. After reviewing Verizon's explanations in light of the PUC's requests, this court concludes that the PUC's assessment of their persuasiveness evidence was supported by substantial evidence.^[12] The commission's determinations are, therefore, affirmed.

Having considered the parties' challenges to the recurring cost model, the court will now turn to Verizon's challenge to the PUC's adoption of the MCI model for nonrecurring costs.

C. Nonrecurring Cost Model

Verizon contends that the PUC's adoption of MCI's model for nonrecurring costs should be overturned because it was arbitrary and produced illegally low rates. In support of this argument, Verizon first argues that the commission erred in rejecting its own proposed model. In the alternative, the incumbent submits that the PUC's adoption of MCI's model was illegal and unsupported by substantial record evidence. For the reasons set forth below, the court disagrees. The commission's decision was supported by the record and consistent with TELRIC. It is, therefore, affirmed.

Nonrecurring costs are those one-time costs associated with the duties that an incumbent must perform to provide UNEs to competitors and otherwise comply with the 1996 Act. As such, they are generally incurred when a competitor requests a service and are comprised of the incumbent's out-of-pocket labor costs necessary to meet those requests. As with all UNE rates, nonrecurring costs must comply with TELRIC methodology and be "based on the use of the most efficient telecommunications technology currently available and the lowest cost network configuration." 47 C.F.R. § 51.505(b)(1).

During the proceeding below, the parties presented the commission with competing nonrecurring costs models, and, after considering each of these submissions, the PUC ultimately chose to adopt MCI's model instead of its Verizon-sponsored counterpart. Tentative Order at 173-180, SJA 430-37. The commission reasoned that although MCI's model was less than ideal, the Verizon model was "so deficient as to preclude its use for establishing TELRIC-compliant rates in this proceeding." Id. at 177, SJA 434. In making this determination, the PUC noted that it was persuaded that the costs proposed by Verizon were excessive and would present a barrier to competitive entry by other providers. Id. at 178, SJA 435. It further explained its concern that the Verizon model failed to reflect the industry trend towards mechanized processes as a way to limit costs. Id. at 179, SJA 436.

Verizon first argues that the commission's rejection of its own recurring cost model was arbitrary and capricious. It supports this position by noting that other state commissions have adopted the very same model that the PUC rejected. This argument, however, is unconvincing. The PUC's rejection of the Verizon model as flawed was amply supported by the record. Furthermore, the state commission orders that Verizon cites as standing for the proposition that its model was superior to MCI's are distinguishable and not binding on this court.

In choosing a rate model, the PUC found a wide discrepancy in the respective methodologies employed by MCI and Verizon in calculating their nonrecurring costs. Verizon assessed its one-time activity costs largely on the basis of employee survey results; MCI did so based on expert evaluations. Verizon attacks MCI's cost projections by arguing that because the CLEC's experts were not employees who had personally provisioned UNEs, their determinations were purely hypothetical speculation. The incumbent further alleges that because MCI's experts were retained solely to ascertain costs for the purposes of regulatory proceedings — and because those experts knew that MCI wanted to limit costs — their evaluations were inherently biased. The PUC, however, found otherwise, and that determination was a reasonable exercise of the commission's discretion.

MCI's experts were identified and made available to Verizon during the PUC proceeding for cross-examination to challenge their credibility and testimony. Furthermore, the competitor's methodology was explained in the record and in the Recommended Decision. The question of the experts' purported biases or lack of experience was, therefore, open to exploration. As the finder of fact, the PUC was free to credit the experts' testimony, and the record was more than sufficient to justify doing so. The court's role here is not to reconsider the credibility of witnesses that the PUC found acceptable. See United States v. Loney, 219 F.3d 281, 288 (3d Cir.2000) (emphasizing that "factfinders routinely, and permissibly, draw inferences when they are evaluating a witness's credibility").

By contrast, Verizon's employee survey participants were neither identified nor available for cross-examination. Recommended Decision at 75, SJA 235. As a result, the PUC was unable to explore on the record any possible biases that may have affected the employees' calculation of task times. Naturally, this inability rendered the Verizon model much more vulnerable to suspicions of unreliability. Indeed, Verizon's workers may well have had incentive to overestimate to their employer their own efficiency and understate the amount of time required to complete tasks. The record as it stands leaves no way for the PUC or this court to discern such incentive.

In addition to the potential biases and the lack of testimony before the PUC, Verizon's survey itself conflicted with the goal of reaching TELRIC-compliant rates. Respondents were asked to report the "actual time it does take to perform the activity in its entirety, not the time that it should take." MCI Stmt. 1.0 (Nonrecurring Costs Panel Rebuttal) at 85, SJA 3479 (emphasis in Verizon's original). Thus, employees' responses were geared not towards a most efficient network standard as required by TELRIC, but towards maintaining the status quo. See Verizon Communications, Inc., 535 U.S. at 511-12, 122 S. Ct. 1646 (prohibiting pricing that relies on the costs incumbents claim to actually incur in providing UNEs). While Verizon's experts did adjust the reported results for the purposes of ascertaining forward-looking task times, the experts were not available for cross-examination on the record, and it remains unclear how the adjustments were impacted by the structure of the original survey questions. Recommended Decision at 69, SJA 229 (concluding that Verizon had not demonstrated that it had corrected any embedded inefficiencies).

Furthermore, the record reflects wide disparities in times reported by different employees for the same task and other inconsistencies that raise further questions as to the Verizon model's reliability.^[13] Although Verizon maintains that its statistical review controlled the survey for errors and that its methodology was sound,^[14] the statistical reviewers themselves also remained unidentified and unavailable for cross-examination. Thus, given the potential biases, the conflict between Verizon's employee survey and TELRIC methodology, and the inconsistent responses to that survey, the PUC reasonably rejected the Verizon model.

In addition to arguing that the PUC acted arbitrarily and contrary to the record evidence in rejecting Verizon's model, the incumbent looks to the FCC's implementation of that same model in Delaware and New Hampshire — as well as its adoption by other state commissions — as evidence that it is not deficient and is in fact preferable to MCI's model. See In re Application of Verizon, New England, 2002 WL 31130421, ¶ 86, 17 F.C.C.R. 18660 (2002) ("Delaware Order") (holding that "Verizon's nonrecurring cost model ... produced [costs] that fall within the reasonable range that TELRIC principles would produce"). The state commissions that accepted a Verizon-proposed model for nonrecurring costs include the New York Public Service Commission, the Delaware Public Service Commission, the Massachusetts Department of Telecommunications and Energy, the Rhode Island Public Utilities Commission, and the New Jersey Board of Public Utilities. Since its model was accepted by these panels, Verizon's argument goes, it should have been accepted by the PUC. The court, however, rejects that argument because these orders are inapposite and not binding on the current proceeding.

First, Verizon's reliance on the FCC and state commission orders in support of the proposition that its model was not inherently flawed is misplaced. Although each of those panels adopted a Verizon-sponsored model for nonrecurring costs, they did so — as Verizon admits — only after making appropriate alterations. Verizon Memorandum at 43. See, e.g., Virginia Pricing Order at ¶ 580 (noting that "every state commission has recognized various significant upward biases [in nonrecurring cost calculations]" and that "in most states, Verizon's was the only model submitted on the record, and thus the state commission relied upon it, but made downward adjustments to offset observed biases"); Delaware Order at ¶ 86 (acknowledging that Verizon's model met TELRIC standards only after "the Delaware Commission made reasonable adjustments"). Because each of these panels adjusted a Verizon-sponsored model before adopting it, these orders do not convince this court that the model proposed by the incumbent to the PUC was flawless. To the contrary, this authority actually seems to support the PUC's finding that Verizon's unaltered model, as presented, was problematic.

Second, with respect to the FCC order in particular, even if the commission had adopted a Verizon model without modification — and even if it had been the same exact model as that presented to the PUC — that order would still not be binding on the instant case because it was issued in the context of a Section 271 proceeding. 47 U.S.C. § 271. In such proceedings, the FCC employs a "benchmark analysis" to determine whether an incumbent's rates are reasonably comparable to rates in other states that were already found to be TELRIC-compliant. Delaware Order at ¶ 37. Because this standard differs materially from a district court's de novo review of interconnection agreements for compliance with federal law under 47 U.S.C. § 252(e)(6), the normal rule that courts should defer to an agency's interpretation of its own regulations does not apply. See Bowles v. Seminole Rock & Sand Co., 325 U.S. 410, 414, 65 S. Ct. 1215, 89 L. Ed. 1700 (1945) (holding that an agency's statutory interpretation is controlling "unless it is plainly erroneous or inconsistent with the regulation"); Applebaum v. Nissan Motor Acceptance Corp., 226 F.3d 214, 218 n. 4 (3d Cir.2000) (noting that the Bowles standard applies when an agency is interpreting its own regulations). But see WorldCom, Inc. v. FCC, 308 F.3d 1, 5 (D.C.Cir.2002) (explaining that the lax standard of review and other circumstances surrounding a Section 271 render it impossible for the FCC to independently determine whether an incumbent's rates comply with TELRIC); In re Petition of WorldCom, Inc., 2002 WL 1576912, 17 F.C.C.R. 27039 at ¶ 53 n. 123 (2002) ("Virginia Arbitration Order") (noting the limitations of a Section 271 hearing).

In addition to challenging the PUC's rejection of its own model, Verizon also contends that the commission's adoption of the MCI model was illegal and unsupported by the record. This court, however, finds that the PUC acted reasonably in selecting the MCI's model and determining that it complied with TELRIC principles. As such, the commission's determination is affirmed.

Verizon asserts that MCI's model ignores or underestimates significant nonrecurring costs that it would incur in provisioning UNEs to its competitors and that these errors violate TELRIC by compelling the incumbent to provide services to CLECs for free. Specifically, Verizon finds the MCI model unacceptable because it assumes that Verizon will employ cost-cutting technologies or processes that are not currently available. It contends that this assumption conflicts with the principle that rates "are to be set based on the costs of elements using the most efficient technology `currently available for purchase.' "Verizon Communications, Inc., 535 U.S. at 506 n. 22, 122 S. Ct. 1646 (internal citation omitted). See also 47 C.F.R. § 51.505(b)(1).

After review of the record, the court finds that Verizon's argument lacks merit. Indeed, in each instance where Verizon accused MCI's model of ignoring or omitting expenses, the PUC expressly addressed the question and found otherwise. Final Order at 28, JA 385 ("We emphasize that we do not require Verizon PA to perform any services without compensation. Rather, we conclude that the functions Verizon PA claims are not addressed in the [MCI] model are functions which do not reflect a TELRIC network."). This finding was reasonable because the record demonstrates that the MCI model adopted by the PUC employed the most efficient, forward-looking processes and was superior to Verizon's model, which included embedded inefficiencies.

For example, the parties set forth competing proposals for the manual processing fallout rate — which measures the percentage of CLEC orders that require manual intervention to correct errors — and the PUC settled on MCI's 2% figure. Although Verizon argues that such a high level of efficiency would be impossible to achieve without undeveloped technologies, the evidentiary record reflects that Verizon's greater proposed fallout rate is "due largely to the inadequacies of its existing processes [and that] those processes are not consistent with a forward-looking network." Recommended Decision at 73-74, SJA 234-35. Indeed, the PUC heard considerable testimony that Verizon's model did not distinguish between fallout actually caused by CLEC order errors and that which resulted from pre-existing inefficiencies in Verizon's network. See, e.g., MCI Stmt. 1.0 (Nonrecurring Costs Panel Rebuttal) at 33-34, 53-54, SJA 3427-28, 3447-48; MCI Stmt. 1.1 (Nonrecurring Costs Panel Surrebuttal) at 3, SJA 3591. Based on such evidence, the PUC chose to adopt MCI's model because it calls for the most efficient, forward-looking network processes, in lieu of Verizon's model, which contained embedded inefficiencies. Because this determination was both consistent with TELRIC and reasonable on the record, it is upheld. In addition to arguing that the PUC's decision to adopt MCI's model instead of Verizon's was unsupported by the record, the incumbent also contends that the PUC erred by noting that its decision was consistent with the FCC's reasoning in the Virginia Arbitration. The incumbent first claims that the PUC's reliance on this decision was improper because — as a decision handed down by the FCC's Wireline Bureau instead of the full federal commission itself — it does not constitute agency policy. As such, the argument goes, the general rule that a court should defer to an agency's interpretation of its own regulations should not apply. Second, Verizon argues that even if this court were to defer to that decision, it should nevertheless disregard it because it was based upon legal error. The court rejects these arguments.

First, the PUC did not defer to the Wireline Bureau's opinion but rather noted "parenthetically" that its choice was supported by that decision, which considered and rejected a Verizon-sponsored model that used substantially the same approach as the model at issue in the instant case. Final Order at 72, JA 429. The PUC considered the record evidence, reasonably concluded that MCI's forward-looking model was superior to Verizon's inefficient submission, explained its decision, and merely bolstered that decision by noting parenthetically that its approach was consistent with that taken by an arm of the FCC in a nearly identical situation. Indeed, there is nothing in the PUC's decision to suggest that it improperly deferred to the Wireline Bureau.

Regardless of the Virginia Pricing Order's precedential weight, however, it is worthwhile and acceptable for this court to consider the Wireline Bureau's reasoning. That order makes clear that the Verizon model failed because it was based largely on Verizon's existing processes and technology and that, on the other hand, the MCI model succeeded because it complied with TELRIC by accounting for efficiencies that would be achieved by the most effective processes and technology currently available. While the MCI model was not without flaw, there was substantial evidence on the record to support the PUC's rejection of the Verizon model; such rejection was neither arbitrary and capricious nor unsupported. This reasoning applies in Pennsylvania as well as Virginia, as the MCI model is nearly identical in both cases.^[15]

Second, Verizon argues that the PUC erred in relying on the Virginia Pricing Order because that decision was based upon an incorrect interpretation of TELRIC's requirement that rates be based on the cost UNE provision aided by the most efficient technology "currently available for purchase." Verizon Communications, Inc., 535 U.S. at 506 n. 22. In Virginia, the Wireline Bureau found that the MCI model interpreted "currently available" as "any technology that is theoretically feasible, even if it has not been implemented by any carrier." Virginia Pricing Order at ¶ 568. By contrast, the Supreme Court and the full FCC have held "currently available" to mean "currently available for a new entrant to purchase today as an alternative to the incumbent's technology, and not some technology that `may' make the efficiency feasible in the future." Verizon Communications, Inc., 535 U.S. at 506, 122 S. Ct. 1646. Despite Verizon's argument, however, these two interpretations are not irreconcilable. Indeed, it is possible for technology to be "available for a new entrant to purchase today as an alternative to the incumbent's technology" and yet not implemented by any carrier. The fact that carriers such as Verizon may choose not to implement certain new technologies does not mean that they are unable to do so. As such, the court cannot conclude that the Wireline Bureau applied TELRIC incorrectly.

For the foregoing reasons, this court upholds the PUC's decision to use MCI's model for nonrecurring cost model as opposed to that submitted by Verizon.

D. Other Challenges

MCI sets forth three additional challenges to the PUC's rate-setting methods. The court finds each of these to be without merit and will address each in turn.

1. OSS Access Charge

MCI claims that the PUC committed reversible error by adopting an Operations Support Systems ("OSS") access charge^[16] that was based on Verizon's actual, embedded expenses because TELRIC requires that UNE rates be based on the cost of provision that would be incurred by a hypothetical, most efficient, least-cost provider. 47 U.S.C. § 252(d)(1); 47 C.F.R. §§ 51.505(b)(1), 51.505(d)(1). This argument, however, is unconvincing. The PUC properly concluded that Verizon was entitled to recover its costs of providing OSS access, and the record contained substantial evidentiary support for the proposition that Verizon's actual costs of doing so were indeed based on the most efficient technology currently available. Tentative Order at 170, SJA 427 ("Verizon is entitled to recover its costs to provide CLECs access to OSS."); 47 U.S.C. § 252(d)(1)(A) (UNE rates to be "based on the cost" of provision). The commission's determination is, therefore, affirmed.

An OSS access charge is designed to allow incumbents to recover the costs incurred when they permit competitors to interface with their OSS — background software systems that contain both essential network information and programs to manage billing, repair ordering, maintenance, and other functions. Iowa Utils. Bd., 525 U.S. at 387, 119 S. Ct. 721. As such, it is comprised of two parts: the initial costs of developing such a system and the costs of ongoing maintenance and provision.

MCI argues that the former are generally prohibited under TELRIC because they represent the cost to the incumbent of modifying its network to comply with the 1996 Act and, as a result, pass on existing inefficiencies to the competitor. Similarly, it contends that the latter are specifically unlawful in this case because they are based upon Verizon's actual costs. The court, however, rejects these arguments and holds that an incumbent may lawfully charge for both the development and provision of a system that allows competitors to interface with OSS, as long as the costs are forward-looking and consistent with those that would be incurred by a hypothetical, least-cost provider. See 47 C.F.R. §§ 51.505(b)(1), 51.505(d)(1).

While the FCC regulations dictate that incumbents must cooperate with competitors and provide them with access to OSS based on the cost of provision, it does not follow, as MCI seems to suggest, that such access must be completely subsidized by incumbents. Indeed, the District of Kentucky rejected a similar argument advanced by a competitor, AT & T, which claimed that the state commission erred by requiring it to pay for the development of systems that would enable it to interface with the incumbent's OSS. The court explained that "AT & T is the cost causer, and it should be the one bearing all the costs." AT&T Communications of the S. Cent., States, Inc. v. BellSouth Telecomm., Inc., 20 F. Supp. 2d 1097, 1105 (E.D.Ky.1998). This court agrees. TELRIC permits incumbents to charge competitors for the development of systems that help them comply with the Act by facilitating access to OSS, as long as those costs are forward-looking and contemplate the use of least-cost technology to eliminate existing inefficiencies. The PUC complied with that standard here.

During the proceeding below, Verizon set forth substantial evidence that its actual costs of deploying the equipment necessary to assess its OSS did not contain embedded inefficiencies but instead were consistent with those that would be incurred by a hypothetical, least-cost provider using the most efficient technology currently available. Verizon Stmt. 1.0 (Panel Testimony on Recurring Costs) at 131-33, SJA 919-21 (explaining that the costs are forward-looking, have no relation to any of Verizon's historic business activities, and are based on the most efficient technology currently available). See also Virginia Pricing Order at ¶ 541 (explaining that although it may not generally be the case that past expenses are a valid estimate of forward-looking costs, such a conclusion can nevertheless be lawful when it is supported by the record evidence). Thus, the PUC's decision was amply supported by the record and is, therefore, affirmed.

2. DUF Charge

MCI next challenges the PUC's adoption of a Daily Usage File ("DUF") charge based on the contention that the fee improperly raises rates by adding $.30 per month to Verizon's cost of servicing an average customer. The CLEC contends that the commission's decision to adopt this charge was both illegal and unsupported by substantial evidence. These arguments, however, lack merit, and for the reasons set forth below, the court will affirm the PUC's determination.

Verizon's DUF is a service that provides competitors with critical information about their customers' usage detail by recording information about each call placed in a message. Competitors then review these messages and use them for billing purposes. During the proceeding below, the PUC approved the ALJ's recommendation to adopt Verizon's proposal to charge competitors a per-message charge of $.0153. Tentative Order at 172, SJA 429. MCI challenges the PUC's adoption of this rate and raises two main arguments in support of its claim.

First, MCI contends that the DUF rate is too high and contrary to TELRIC because the PUC relied on two improper assumptions — one depressing the denominator and the other inflating the numerator — in setting up the fraction used in its calculations. Although cast as legal arguments, MCI's claims are factual in nature, and this court will affirm the PUC's determinations as supported by substantial evidence. See MCI Telecomm. Corp., 271 F.3d at 517 (factual findings of state commissions must be affirmed if they have "substantial support in the record as a whole").

TELRIC defines the forward-looking per unit cost of a UNE as the total cost of the element divided by a reasonable projection of the sum of the total number of units that the incumbent is likely to provide to competitors and the total number that it is likely to use in offering its own services. 47 C.F.R. § 51.511(a). In 1996, a DUF rate was calculated based on the projection that competitors would demand approximately 11 billion messages from incumbents in the year 2000. This large projected demand figure — which seemed reasonable given the Act's purpose of fostering widespread competition in the industry — was inserted into the denominator of the DUF fraction and resulted in a relatively low overall rate. In reality, however, the projection was overly optimistic, and the actual demand in 2000 was only a fraction thereof, or 0.8 billion messages.

In its rate-setting, the PUC abandoned the obsolete 11 billion figure and calculated the DUF rate by beginning with the actual demand figure of 0.8 billion and forecasting nominal growth over the ensuing several years. This dramatic decrease in the denominator resulted in a large increase in the overall DUF rate.

MCI claims that the PUC erred by making this assumption because demand for all UNEs by competitive entrants into the telecommunications industry should be expected to increase dramatically, instead of nominally, as a result of the 1996 Act. This, however, was not the case between 1996 and 2000, and the PUC reasonably concluded that it would likewise not be true from this point forward. The commission considered MCI's argument, but it nonetheless decided to credit Verizon's testimony that the incumbent's actual experience in providing the DUF service to CLECs was a much more accurate estimate of forward-looking demand than the numbers submitted by MCI. Tentative Order at 172, SJA 429 ("[W]e find that Verizon has sufficiently demonstrated on the record that the increase is due to a smaller anticipated DUF usage."). This determination was reasonable and supported by substantial evidence. See, e.g., Verizon Stmt. 1.1 (Recurring Panel Surrebuttal) at 212-13, SJA 1249-50 (explaining that the higher DUF rate was consistent with the fact that the earlier rate had been based on a projected demand of 11 billion messages for 2000 when the actual demand for that year was only 0.8 billion).

In addition to challenging the 0.8 billion demand figure used in the denominator, MCI also challenges the figure that the PUC plugged into the numerator, which represents Verizon's total costs of providing DUF access based on the projection of 11 billion users. The competitor argues that it was error for the PUC to use a figure representing total costs based on Verizon's construction of a system capable of servicing 11 billion expected users and then spread those exorbitant costs over only the 0.8 billion actual users. The PUC, however, rejected this rationale and instead adopted the ALJ's finding that because certain DUF costs are fixed, Verizon could not simply downsize its facilities to correspond to the lower demand once it realized that the original projection was overly optimistic. Recommended Decision at 66, SJA 226. This finding is consistent with economic principles and, given the ALJ's disposition, was reasonable on the record.^[17]

Second, MCI attacks the PUC's adoption of the per-message DUF charge of $.0153 on the basis that it included some $1.1 million in costs that were double-counted. In particular, the CLEC claims that the costs for support labor — a force comprised of 15 employees, who manually handle errors in the automated DUF process — were included in Verizon's adjustments to its ACFs and that it was, therefore, improper to also include them in the DUF charge. The court disagrees. The PUC considered testimony from both sides on this issue and decided to give more weight to Verizon's contention that it had "carefully ensured that appropriate adjustments were made to the ACFs in order to avoid any double-counting of expenses, including DUF expenses." Tentative Order at 172, SJA 429; Verizon Opposition Brief at 28 (citing Verizon Stmt. 1.1 (Recurring Surrebuttal) at 214, SJA 1251; Verizon Ex. 4 (Minion Testimony) at 4007, SJA 3136). The commission's determination was supported by the record and is, therefore, affirmed.

3. Switching Rate Structure

Third, MCI challenges the PUC's decision to allow Verizon to charge competitors for access to switching facilities on a per-minute-of-use basis as opposed to a flat rate. The competitor contends that this determination violated TELRIC for two reasons, but the court disagrees. The commission's decision — which was consistent with past pricing of switches in Pennsylvania — was supported by substantial evidence and comported with federal law. It is, therefore, affirmed.

MCI first argues that the PUC's adoption of Verizon's proposal to utilize a per-minute charge violates TELRIC's mandate that incumbents recover costs "in a manner that reflects the way they are incurred" because, according to the competitor, Verizon does not incur costs on the basis of usage but instead according to a flat rate that depends on the number of users connected to each switch. Local Competition Order at ¶ 622. Although set forth as a legal argument, MCI is actually challenging the PUC's factual finding that "Verizon has presented substantial evidence that usage can and does lead to exhaustion of non-port resources and such considerations are impacted by anticipated usage." Tentative Order at 146, SJA 403. After reviewing the record, the court agrees that the PUC's determination was indeed substantially supported. See, e.g., Verizon Stmt. 3.1 (Rebuttal Testimony of Dr. Howard Shelanski and Dr. Timothy Tardiff) at 22-24, JA 2761-63 (explaining that switching charges should be recovered in the same way they accrue and that many of Verizon's switching costs are incurred on the basis of usage); Verizon Stmt. 6.0 (Rebuttal Testimony of Harold E. West III) at 1-9, JA 2967-75 (explaining that a flat-rate structure would be inappropriate and could even harm competing carriers that served lower volume users than MCI and AT & T).

Secondly, MCI argues that the alleged discrepancy between the manner in which Verizon incurs it switching costs and that in which it charges competitors for access to its switching facilities violates the Act's prohibition against discrimination. 47 C.F.R. § 51.311(b) (mandating that the quality of access to UNEs that an incumbent provides to a competitor be at least equal in quality to that which the incumbent provides to itself). Essentially, the CLEC claims that because Verizon incurs its costs based on a flat rate, it can pass that low rate on to its retail customers, and that such a transfer results in discrimination against competitors — whom the PUC ordered to pay based on a per-minute-of-usage basis — by pricing them out of the market. This argument, however, both confuses the requirement that rates be non-discriminatory and is predicated upon a factual assumption — that Verizon incurs its costs on a flat rate — that the PUC reasonably rejected.

Contrary to MCI's suggestion, the prohibition against discrimination does not mean that Verizon must charge competitors under the same terms and conditions that it charges its retail customers.^[18] Instead, as the language of § 51.311(b) makes clear, an incumbent must charge competitors under the same terms and conditions that it would charge itself, i.e., its costs. That is exactly what Verizon has done here. As discussed, the PUC reasonably concluded based on the record evidence that Verizon's switching costs are traffic-sensitive, and it directed Verizon to pass those costs on to MCI in the same manner — according to the amount of usage as opposed to the number of users — in which they were incurred. That is all § 51.311(b) requires. Thus, the PUC's determination is affirmed.

E. Takings

Finally, the court will address Verizon's argument that the UNE rates set by the PUC are so low as to amount to an unconstitutional taking of its network and workforce without adequate compensation, in violation of the Fifth and Fourteenth Amendments of the United States Constitution and Article 1, § 10 of the Pennsylvania Constitution. The incumbent asks the court to remand the takings question back to the PUC, which Verizon alleges refused to consider evidence on the issue. The court, however, declines to do so because it has already affirmed the PUC's rates as consistent with the 1996 Act and the FCC's TELRIC methodology. In addition, Verizon has not made a prima facie showing of any confiscation that would justify further proceedings in front of this court or the PUC.

Verizon argues that the PUC's Order setting rates is confiscatory because it would cause the incumbent to lose well over $690 million in projected revenues in order to cover costs associated with providing UNEs to competitors. As such, the rates would not be "sufficient to assure confidence in the financial integrity" of Verizon's Pennsylvania business, and would render the company unable "to maintain its credit and to attract capital." FPC v. Hope Natural Gas Co., 320 U.S. 591, 603, 64 S. Ct. 281, 88 L. Ed. 333 (1944). The PUC counters by pointing out that Verizon never raised a takings claim before it, and, as a result of this omission, it was not required to consider the matter. In addition, the commission claims that Verizon's failure to raise the takings claim in the proceeding below constitutes a waiver and that, therefore, this court should refuse to consider that claim. The court disagrees.

In its Comments to the Tentative Order, Verizon argued that the PUC's rates would risk the financial safety of its UNE business in Pennsylvania and impair its ability to invest further in its network, a hindrance that would result in its being forced to offer a lower quality service to its customers. Verizon Comments to Tentative Order at 6-7, SJA 1675-76. It also alleged that "the Commission's tentative rates would not even meet Verizon PA's 2001 cash operating expenses ... [and that, therefore,] the Commission must now consider Verizon PA's evidence demonstrating that the tentative rates are confiscatory, and adjust those rates to avoid a taking." Id. at 5-7, SJA 1674-76 (emphasis added). Because Verizon raised these arguments prior to the PUC's issuing of its Final Order, the court cannot conclude that the incumbent waived its takings claim.

Nevertheless, the court finds that it would be futile to remand the incumbent's Fifth Amendment claim to the PUC for consideration. As discussed, this court has already concluded that the UNE rates adopted by the PUC comply with TELRIC and do not exceed the forward-looking economic cost per unit of provision. 47 C.F.R. § 51.505(e). In addition, it agrees with the commission's finding that to impose the higher rates Verizon seeks would be to violate federal law. Final Order at 14 (JA 371). Although UNE rates adopted by a state commission rates are indeed subject to challenge on the basis that they fail to provide adequate compensation, Verizon Communications, 535 U.S. at 524, 122 S. Ct. 1646, it is likewise true that once those rates are determined to be reasonable under TELRIC by district court, the commission can hardly consider them to be an unconstitutional taking.

The court is not holding that Verizon cannot attempt to prove a takings claim but instead that such a claim simply requires a different forum. The PUC lacks discretion and is bound by the FCC to apply TELRIC standards, and if any unconstitutional confiscation resulted from its application of those standards, it is the FCC — and not the PUC — that can cure the alleged infirmity. Indeed, an ILEC may petition the FCC to protest TELRIC as applied by the state commission and affirmed by the district court, if it can "provide specific information to show that the pricing methodology as applied to [the incumbent] will result in confiscatory rates." Verizon Communications, Inc., 535 U.S. at 528 n. 39, 122 S. Ct. 1646.

Moreover, Verizon has failed to make a prima facie showing of any confiscatory nature that would justify further proceedings in front of this court or the PUC. The incumbent insists that the rates set by the PUC undermine its financial integrity, its opportunity to attract capital, and its ability to compensate investors for risks assumed. Hope Natural Gas, 320 U.S. at 605, 64 S. Ct. 281 (setting forth the standard for establishing an unconstitutional taking). But Verizon does not present evidence to support these claims. Although rate-setting may reduce the value of the property which is being regulated, it does not necessarily follow that the regulation is invalid. Hope Natural Gas, 320 U.S. at 601, 64 S. Ct. 281. Here, Verizon fails to offer evidence that its revenue losses eviscerate its ability to turn a profit and provide customer service. As such, this court must reject its takings claim.

F. Conclusion

For the foregoing reasons, the court affirms the rates adopted by the PUC in the Generic Investigation.

NOTES

[1] The PUC asserts that the doctrine of sovereign immunity protects the commission itself from being sued in federal court and that, therefore, only its individual officers are proper defendants. This assertion, however, is contrary to binding Third Circuit authority holding that, as a state commission that has accepted the gratuity of Congress' invitation to participate in regulating the telecommunications industry, the PUC has waived its sovereign immunity and subjected itself to suit in federal court. MCI Telecomm, Corp. v. Bell Atl-Pa., 271 F.3d 491, 510 (3d Cir.2001). In the absence of an explicit disavowal or overruling of this controlling precedent, the court must reject the PUC's argument that it should decline to decide the jurisdictional issue on waiver grounds simply because that was the approach taken by the Supreme Court in Verizon Md., Inc. v. Pub. Serv. Comm'n of Md., 535 U.S. 635, 645, 122 S. Ct. 1753, 152 L. Ed. 2d 871 (2002). As such, the court concludes that the PUC, having waived its sovereign immunity, is a proper defendant.

[2] The court will hereinafter refer to the competing carriers as "MCI" for the sake of simplicity.

[3] These claims are set forth in Counts II-VII of Verizon's Second Amended Complaint. The parties have stipulated to the dismissal of Count I.

[4] The FCC conducted the proceedings in Virginia pursuant to 47 U.S.C. § 252(e)(5), a component of Congress' unique regulatory scheme that allows the federal agency to stand in the shoes of a state commission that fails to act or carry out its responsibility under the Act.

[5] Verizon argues that the PUC misinterpreted the Triennial Review Order and misapplied federal regulations when it found that "the FCC rejected the use of financial reporting lives for use in establishing depreciation expense under TELRIC." Final Order at 62, JA 419. But see Triennial Review Order at ¶ 688 (declining to mandate the use of economic lives and explaining that state commissions retain discretion to determine which lives — either regulatory or economic — comply with TELRIC). While the court agrees that this isolated sentence is inconsistent with the FCC's position, it also finds that such error was harmless because the PUC clearly considered the applicable standard and weighed the relative merits of both proposals in light of that standard. Final Order at 62, JA 419 (concluding that the FCC prescribed depreciation lives are "more reflective" than Verizon's proposal of the forward-looking, economic criteria required by the FCC rules and recognizing that "state commissions continue to have discretion in this choice") (emphasis added).

[6] ACFs are intended to recover common costs that cannot be attributed to the provision of any one particular UNE. As such, they must be factored into the cost of all network elements and are, therefore, are an integral part of any recurring cost model.

[7] Even if this court were to adopt the logic of AT & T Communications of Cal., it would still reach the same conclusion because the two cases are readily distinguishable. Unlike the decision of the CPUC, the PUC's determination here produced rates that were reasonable and within the TELRIC range. Compare Compliance Order at 9, JA 463 (explaining that the 7.91% figure adopted by the PUC compared favorably to the 8% multiplier approved in the Virginia arbitration) with AT & T Communications of Cal., 375 F.3d at 903 (challenging a 19% percent mark-up adopted by the state commission).

[8] Although the PUC made several other adjustments to Verizon's model, the decision on port features was the final one challenged by the parties. See PUC Opposition Brief at 7-8 (listing twenty modifications directed by the PUC to render the incumbent's model TELRIC-compliant).

[9] Although excluded from the limited feature offering, competitors had the option of purchasing the four additional features — Three Way Calling, Custom Ringing, Centrex Intercom, and Calling Number Delivery — on an a la carte basis. Final Order at 48, JA 405.

[10] The excluded features were Call Waiting Display Name and Number, Calling Number and Name Delivery, Anonymous Call Rejection, Centrex Loudspeaker Paging, Centrex Meet-Me Conference, and Station Message Detail Recording. Letter from Verizon to PUC Secretary dated January 26, 2004 at JA 4735.

[11] The network elements at issue are: Digital 4 Wire Customer Specified Signaling Loop, Subloop Distribution and Feeder, and Unbundled EEL Testing.

[12] For example, the PUC rejected Verizon's contention that the 50% or more increase in loop rates could be explained simply by the fact that the Global Order loop rates were too low and that if those rates had been higher, the variance would not have been so great. Compliance Order at 21, JA 475. This sort of unsupported, collateral attack on the existing rates did not comply with the PUC's request for specific, documented explanations. As such, the commission's decision to reject it, and others similar to it, was reasonable.

[13] For example, the typical time required to set up the inside of the manhole for work to be done was reported as five minutes by one Verizon respondent and four hours by another. MCI Stmt. 1.0 (Nonrecurring Costs Panel Rebuttal) at 91, SJA 3485.

[14] Verizon identified only two survey responses as "potentially troubling outliers." Recommended Decision at 69, SJA 229. See also MCI Stmt. 1.0 (Nonrecurring Costs Panel Rebuttal) at 88-89, SJA 3482-83.

[15] Verizon also contends that this court cannot rely on the Wireline Bureau's Virginia decision because the FCC staff relied significantly on the Pennsylvania PUC's adoption of the MCI model in the instant case to support its use of that model in Virginia. Virginia Pricing Order at ¶ 580 n. 1489. Although Verizon characterizes the PUC's reliance on the Virginia Pricing Order as circular, such a contention is irrelevant to this court's analysis. Because each commission considered virtually identical models, the reasoning behind one decision remains applicable to the other.

[16] The FCC has explained that incumbents must treat access to OSS as a UNE and provide it to CLECs on an unbundled basis. 47 C.F.R. § 51.319(g).

[17] Moreover, the court notes that the PUC's determination does not violate TELRIC. Verizon constructed a forward-looking DUF system based on future demand projections, and it should not be forced to bear all the costs and be penalized for its attempt to comply with TELRIC simply because those projections missed the mark. Although TELRIC prohibits incumbents from passing on embedded inefficiencies to competitors, it also allows the former to recover their forward-looking costs. 47 U.S.C. § 252(d)(1)(A). That objective was accomplished by the PUC's decision here.

[18] Indeed, the PUC realized as much when it adopted the ALJ's recommendation on the per-minute charge and accepted the conclusion that every carrier — regardless of how it incurs costs — is free to use its own business judgment in structuring retail rates. Tentative Order at 146, SJA 403.