Full opinion text
OPINION AND ORDER JELDERKS, United States Magistrate Judge. Plaintiffs MCI Telecommunications Corp. and MCImetro Access Transmission Services, Inc. (collectively “MCI”), bring this action under 47 U.S.C. § 252(e)(6) against GTE Northwest, Inc. (“GTE”), the Oregon Public Utility Commission (“PUC”), and PUC Commissioners Roger Hamilton, Ron Eachus, and Joan Smith (“the Commissioners”). GTE has asserted counter-claims against MCI and cross-claims against the PUC. At issue are the terms of an interconnection agreement (“Agreement”) between MCI and GTE, which is intended to facilitate competition in local telephone service. Each party has moved for summary judgment. GTE also has moved to stay the proceedings and for leave to dismiss its state law claims without prejudice. SCOPE AND STANDARD OF REVIEW The Telecommunications Act of 1996 (“the Act”), Pub.L. No. 104-104, 110 Stat. 56, 47 U.S.C. § 153 et seq., requires each incumbent local exchange carrier (“ILEC”) to negotiate an interconnection agreement with any prospective competitive local exchange carrier (“CLEC”) that wishes to provide local telephone service in the ILEC’s service territory. 47 U.S.C. §§ 251 and 252. The Act contemplates that the ILEC and CLEC will first attempt to negotiate an Agreement. , 47 U.S.C. §§ 251(c)(1) and 252(a)(1). Any remaining disputes may be submitted to an arbitrator, whose decision can be appealed to the state public utility commission. 47 U.S.C. § 252(b), (c), and (e). Finally, “any party aggrieved” by a decision of a state public utility commission concerning such an agreement “may bring an action in an appropriate Federal district court to determine whether the Agreement ... meets the requirements of the Act.” 47 U.S.C. § 252(e)(6). Although the Act does not specify either the standard or scope of review, there is general agreement that review under § 252(e)(6) is confined to the administrative record. With regard to the standard of review, it is neither desirable nor practical for this court to sit as a surrogate public utilities commission to second-guess the decisions made by the state agency to which Congress has committed primary responsibility for implementing the Act in Oregon. Rather, this court’s principal task is to determine whether the PUC properly interpreted and applied the Act, which is a question of federal law that is reviewed de novo. In all other respects, review will be under the arbitrary and capricious standard. THRESHOLD ISSUES 1. Effect of the Recent Supreme Court Decision The court requested supplemental briefing regarding the effect upon this case of the Supreme Court’s recent decision in AT & T Corp. v. Iowa Util. Bd., — U.S. -, 119 S.Ct. 721, 142 L.Ed.2d 835 (1999). At issue in AT & T was the validity of numerous regulations that the Federal Communications Commission (“FCC”) promulgated in 1996 to implement the Act. The regulations were to take effect on September 30, 1996. See 61 Fed.Reg. 45,-476 (1996). On September 27, 1996, the Eighth Circuit temporarily stayed “the effective date” of all of the regulations. Iowa Util. Bd. v. Federal Communications Comm’n, 96 F.3d 1116, 1118 (8th Cir.1996). On October 15, 1996, the Eighth Circuit allowed some of the regulations to go into effect, but continued to stay the FCC’s pricing regulations (47 C.F.R. §§ 51.501-51.515, 51.601-51.611, 51.701-51.717) and the “pick-and-choose rule” (§ 51.809). Iowa Utilities v. F.C.C., 109 F.3d 418, 427 (8th Cir.1996). The Supreme Court rejected several requests (from the FCC and others) to vacate the stay. On July 18, 1997, the Eighth Circuit vacated many of the stayed FCC regulations on the ground that the FCC lacked jurisdiction to issue them. Iowa Util. Bd. v. FCC, 120 F.3d 753 (8th Cir.1997). The Eighth Circuit also vacated several other regulations on the merits (including at least one that had not been stayed), while affirming still others. Id. The stay expired once that order became effective. Id. at 820. Meanwhile, the Agreement at issue here was approved by the PUC, and signed by the parties, after various disputes were resolved through arbitration. The arbitrator issued his decision on January 3, 1997, and the PUC affirmed it, with minor modifications, on February 3, 1997. MCI’s petition for reconsideration was denied on May 30, 1997. All of these events occurred after the FCC regulations had been stayed, but before the Eighth Circuit issued its decision on the merits. The PUC treated the stayed regulations as persuasive authority that the PUC could, but was not required to, follow. On July 18, 1997, the arbitrator resolved a few lingering disputes over the contract language and directed the parties to execute and file the final agreement. As luck would have it, that was the same day the Eighth Circuit issued its decision on the merits in Iowa Utilities. The PUC withdrew its order and, after supplemental briefing, directed the parties to modify the contract to comply with the Eighth Circuit’s decision. On October 3, 1997, the PUC approved the revised agreement. On October 14, 1997, the Eighth Circuit clarified its prior decision and order in response to a petition for rehearing. Iowa Utilities, 120 F.3d 753. On November 26, 1997, MCI asked this court to review certain disputed portions of the Agreement. On January 7, 1998, GTE asked the PUC to amend the Agreement to comply with the Eighth Circuit’s October 14th decision. On January 13, 1998, GTE filed counterclaims and cross-claims asking this court to review disputed portions of the Agreement. On January 26,1998, the Supreme Court granted certiorari to review portions of the Eighth Circuit’s decision. On November 13, 1998, the PUC granted (in part) GTE’s request to amend the Agreement to comply with the Eighth Circuit’s October 14th decision. On January 25, 1999, shortly before oral argument in this case, the Supreme Court issued its decision in AT & T, — U.S. -, 119 S.Ct. 721. The Supreme Court reversed the Eighth Circuit’s decision regarding jurisdiction and held that the FCC did have jurisdiction to promulgate the regulations in question. It remanded that case to the Eighth Circuit to consider various challenges to the merits of those regulations. The Supreme Court affirmed several other regulations on the merits, while vacating at least one (and probably two) additional regulations. This court must now determine how the Supreme Court’s decision affects the instant case and, in particular, how it affects the interconnection agreement that was approved by the PUC and signed by the parties more than a year before the Supreme Court issued its decision. To the extent the Supreme Court simply explained what the Act or the FCC’s implementing regulations mean, this court must apply that interpretation when reviewing the challenged provisions in the MCI-GTE interconnection Agreement. When a federal court interprets a law-— whether it be a statute, a regulation, the common law, or the Constitution itself— the court is not creating new law but merely declaring what that law has always meant, even if this interpretation had not previously been acknowledged or conflicts with an earlier interpretation. See Rivers v. Roadway Express, Inc., 511 U.S. 298, 312-13, 114 S.Ct. 1510, 128 L.Ed.2d 274 (1994) (“A judicial construction of a statute is an authoritative statement of what the statute meant before as well as after the decision of the ease giving rise to that construction”); Harper v. Virginia Dept. of Taxation, 509 U.S. 86, 107, 113 S.Ct. 2510, 125 L.Ed.2d 74 (1993) (Scalia, J., concurring) (when overruling prior precedent a judge does “not pretend to make a new law, but to vindicate the old one from misrepresentation”). The regulations that had been vacated by the Eighth Circuit are a different matter, however. The Eighth Circuit stayed those regulations (with limited exceptions) before they ever went into effect, postponing their effective date. See Iowa Utilities, 96 F.3d at 1118 (staying “effective date”). See also 5 U.S.C. § 705 (expressly authorizing reviewing court to “postpone the effective date of an agency action ... pending conclusion of the review proceedings”). Ultimately, the Eighth Circuit vacated the regulations. The term “vacate” means “to annul; to cancel or rescind; to declare, to make, or to render, void; to defeat; to deprive of force; to make of no authority or validity; to set aside.” Action on Smoking & Health v. Civil Aeronautics Board, 713 F.2d 795, 797 (D.C.Cir.1983). As the PUC acknowledges, “[t]he upshot is that [these] regulations were not effective at the time the interconnection agreement at issue was crafted.” As of this writing, it is unclear whether the Eighth Circuit will stay those regulations anew while it considers various challenges to the merits of those rules. Even if the Eighth Circuit does not extend the stay, MCI nevertheless is asking this court to overturn portions of the MCI-GTE interconnection Agreement on the ground that the PUC failed to apply substantive regulations that were not in effect when the Agreement was approved by the PUC and signed by the parties. The court declines that invitation. The FCC regulations in question go beyond merely interpreting the statute. They create new legal obligations that have the force of law in their own right. The court perceives a crucial distinction between applying a new interpretation of a law that admittedly was in effect during the relevant time period, versus applying a substantive regulation that never was in effect to begin with. Cf. Rivers, 511 U.S. at 311, 114 S.Ct. 1510 (declining to apply new legislative enactment retroactively and reiterating general principle that statutes operate only prospectively); Bowen v. Georgetown Univ. Hosp., 488 U.S. 204, 208-09, 215, 109 S.Ct. 468, 102 L.Ed.2d 493 (1988) (refusing to apply substantive regulations retroactively). MCI essentially is asking this court to treat the stay, and the Eighth Circuit’s subsequent order vacating the FCC regulations, as if those events had never occurred and the FCC regulations were continuously in effect. MCI has not cited a single case that suggests this court has the power to do this, or that it should exercise that power in this circumstance. In an effort to bring this case within the ambit of the Harper retroactivity rule, MCI argues that the Supreme Court was interpreting a statute when it determined that the FCC had jurisdiction to issue these rules, hence the FCC always did have jurisdiction. It is a clever argument, but to no avail. The question is not whether the Eighth Circuit was correct in holding that the FCC lacked jurisdiction, but whether the Eighth Circuit’s orders staying and subsequently vacating those regulations were effective. Reversal of the orders on the merits does not negate the fact that they were valid, effective orders. MCI’s argument can carry the day only if the orders issued by the Eighth Circuit were void ab initio, which they clearly were not. Even though the Eighth Circuit was ultimately reversed on the merits, MCI has not suggested that the Eighth Circuit lacked authority to issue the stay delaying the effective date of the regulations. See 5 U.S.C. § 705. A stay would be a hazardous procedural device if liability could be premised upon violating a rule while it had been vacated or stayed. Furthermore, a prerequisite to application of the Harper retroactivity rule is that the case announcing the new rule has itself applied that rule to the litigants in that case. See Harper, 509 U.S. at 97, 113 S.Ct. 2510 (“When this Court applies a rule of federal law to the parties before it ... ”). See also Reynoldsville Casket Co. v. Hyde, 514 U.S. 749, 752, 115 S.Ct. 1745, 131 L.Ed.2d 820 (1995) (“when (1) the Court decides a case and applies the (new) legal rule of that case to the parties before it, then (2) it and other courts must treat that same (new) legal rule as ‘retroactive’ ”). In that circumstance, fairness requires that other similarly situated litigants receive equal treatment, not just the litigant whose case happened to be the one in which the rule was first announced. See James B. Beam Distilling Co. v. Georgia, 501 U.S. 529, 538, 111 S.Ct. 2439, 115 L.Ed.2d 481 (1991) (plurality opinion); Griffith v. Kentucky, 479 U.S. 314, 323, 107 S.Ct. 708, 93 L.Ed.2d 649 (1987). That prerequisite is not present here. The Supreme Court did not announce a new rule of law and apply it to a particular interconnection agreement, which arguably might entitle other litigants to seek equal treatment. Instead, the Court ruled upon an abstract facial challenge to the FCC’s jurisdiction to promulgate these regulations, and did not apply any of those regulations to a specific interconnection agreement. Nor, on remand, will the Eighth Circuit apply these regulations to any particular interconnection agreement. The litigants in the case decided by the Supreme Court will be treated no differently than the litigants in any other pending case raising the same issues. This court’s decision whether to give retroactive effect to the reinstated FCC regulations, in the context of the MCI-GTE Agreement, will likely be dispositive of a similar issue in five other interconnection cases pending before this court. For that reason, the court has reviewed the 19 supplemental briefs on this issue that were filed by the parties in those other cases. None of the authorities cited is on point. Pope v. Shalala, 998 F.2d 473, 483 (7th Cir.1993), and Pennzoil Co. v. United States Dept’s of Energy, 680 F.2d 156 (Temp.Emer.Ct.App.1982), each concerned an agency’s interpretation of its own existing regulations. The agency was not changing the law but stating what, in the agency’s opinion, the law had always meant. In Appalachian States Low-Level Radioactive Waste Comm’n v. O’Leary, 93 F.3d 103, 113 (3d Cir.1996), the rule at issue was not a substantive regulation but an interpretive rule expressing the agency’s view of what the statute meant. Id. at 112-13. Here, by contrast, the FCC was imposing new substantive requirements, which is why the effective date of the regulations matters. Thorpe v. Durham Housing Authority, 393 U.S. 268, 282, 89 S.Ct. 518, 21 L.Ed.2d 474 (1969), is inapposite as well. At issue was whether a public housing authority must inform a tenant of the reasons why she is being evicted. While the action was pending, a new regulation went into effect which specifically required such notice. Since the tenant had not yet been evicted, the Court instructed the defendant to comply with the regulation and give her notice, thereby avoiding the need to decide whether such notice is mandated by the Constitution. That is an entirely different situation than the one presented here. To the extent Thorpe implies that regulations routinely are applied retroactively, that dictum has been severely limited by subsequent decisions. See, e.g., Georgetown Univ. Hosp., 488 U.S. at 208-09, 215, 109 S.Ct. 468. Thorpe also suggests that courts must always apply the law as it presently exists, but that is an oversimplification. Cf. Landgraf v. USI Film Products, 511 U.S. 244, 114 S.Ct. 1483, 128 L.Ed.2d 229 (1994) (provisions of Civil Rights Act of 1991 did not apply to pending cases premised on events occurring before the effective date of the Act). The cases cited by Sprint are somewhat closer to the mark, but distinguishable nevertheless. In each of those cases, the Ninth Circuit held that the Sentencing Guidelines were applicable to crimes committed, pleas entered, or sentences imposed during the period between the Ninth Circuit’s decision holding the Sentencing Reform Act (“SRA”) unconstitutional, and the Supreme Court’s reversal of that decision. See, e.g., United States v. Kincaid, 898 F.2d 110, 111 (9th Cir.1990); United States v. Gonzalez-Sandoval, 894 F.2d 1043, 1053 (9th Cir.1990); United States v. Kane, 876 F.2d 734, 734-36 (9th Cir.1989). In the SRA cases, the underlying conduct was always illegal, and a statutory penalty was in place, regardless of whether the SRA was constitutional. Here, by contrast, a contract that complied with all laws in effect at the time would retroactively be deemed to have violated a regulation that was not in force at the time. Second, unlike the present case, the Ninth Circuit did not issue a nationwide stay postponing the effective date of the SRA, nor did that court vacate the statute. Rather, the SRA went into effect, and at all times remained on the books; the Ninth Circuit simply declined to apply the statute, believing it to be contrary to the Constitution. By contrast, the regulations at issue here never went into effect. The Eighth Circuit first delayed their effective date for nearly a year, and then vacated them. That is a critical distinction. The Supreme Court ultimately disagreed with the Ninth Circuit’s determination that the SRA was unconstitutional, and concluded that the Circuit had erred by not applying the law in effect at the time. As with other cases in which a lower court has failed to apply the proper rule, the judgments were reversed and remanded for application of the correct rule. That is not the situation here. The PUC did not fail to apply the law that was in effect at the time; rather, it is accused of failing to apply regulations that were not in effect. The procedural posture also is different. In the SRA cases, the Ninth Circuit was reviewing a direct appeal from a sentence imposed by the district court. By contrast, this court is reviewing the terms of a contract approved by an administrative agency. Finally, the Supreme Court’s decision affirming the constitutionality of the SRA applied the rule announced in that case to specific individuals, affecting their sentences. Thus, the Harper prerequisite (that the rule has actually been applied to the parties before the Court) was satisfied there, which also distinguishes the SRA cases from the one at hand. The bottom line is that — despite this court’s request for briefing on the issue— the combined forces of MCI, AT & T, Sprint, WorldCom, and the FCC have failed to cite a single case on point to support their contention that portions of the Agreement must be overturned because the PUC failed to apply certain FCC regulations to an Agreement that was entered into more than a year before those regulations took effect. A final consideration is that the challenged interconnection agreement has only a two year term and will expire seven months from now. If this court were to remand the entire Agreement (or even large portions of it) to the PUC to begin the process anew, this time applying an entirely new set of rules, there is little doubt that the Agreement would expire and this case would be moot before the PUC could complete its task. The net result would be a significant expenditure of resources by all involved, with very little to show for it in terms of opening local telephone markets to competition. Oregon is not alone in this dilemma. The court is informed that there are numerous similarly situated interconnection agreements around the country, of short duration, many of which undoubtedly would expire before they could be revised to comply with the new FCC rules. The only practical course in this circumstance is to generally limit this court’s review of the existing Agreement to a determination of whether any challenged provision violates the statutes and regulations in effect at the time this Agreement was approved by the PUC, excepting any regulations that have since been declared void. The Supreme Court’s interpretation of the relevant statutes and regulations is, of course, controlling. The parties will apply the reinstated FCC regulations — assuming they are not again stayed, or vacated on the merits — in negotiating the replacement contract that takes effect later this year. 2. GTE’s Motion for a Stay GTE has moved to stay these proceedings until (1) the Eighth Circuit completes its review of the merits of the challenged regulations, and (2) the FCC promulgates new rules to replace those vacated by the Supreme Court. The court denies the motion. DISCUSSION OF THE MERITS 1. Price for Network Elements GTE challenges the interim prices established for unbundled network elements, which were based on the results of the UM 351 proceeding. A. Reliance Upon UM 351 Proceeding The PUC began studying unbundled element prices in 1990, six years before the Act was enacted. GTE actively participated in those earlier proceedings, known as UM 351. When Congress enacted the Act in 1996, it established very short timelines for the adoption of interconnection agreements and the establishment of unbundled element prices. The arbitrator understandably chose to rely heavily upon the extensive administrative record and analysis that already existed, instead of beginning the process anew. i. Due Process of Law Reliance on those earlier proceedings did not deny GTE due process of law. The PUC did not apply principles of collateral estoppel to preclude GTE from contesting the element prices. Rather, the earlier proceedings were treated as evidence that could be considered in the subsequent proceeding, and which the arbitrator and the PUC ultimately found were the most reliable evidence then available. GTE was free to introduce contrary evidence and to urge a different result. ii. Prejudgment of Outcome Contrary to the allegations in GTE’s briefs, the arbitrator did not prejudge the results. Rather, during a conference with the parties to discuss how the case would proceed, the arbitrator advised the parties that “at this point” he intended to use the results of the UM 351 proceeding to set the element prices in this Agreement, but the parties could also “submit other costs studies for consideration.” The arbitrator merely did what many judges would have done in that situation, when confronted with a complex case and short deadlines, which was to share his tentative thoughts on how to proceed while still leaving the door open for the parties to persuade him to follow a different course. While the arbitrator’s language might have been chosen more carefully, the court concludes that the arbitrator did not prejudge the outcome of the proceeding. Moreover, as a practical matter, GTE has not pointed to any realistic alternatives to the path identified by the arbitrator. iii. Extra-Record Evidence The arbitrator did not err by taking administrative notice of the record established in the UM 351 proceeding, in which GTE was an active participant. The Act does not require the PUC to ignore the information already in its files, or its extensive expertise and experience in overseeing telephone pricing and service in Oregon, and its knowledge of pricing, costs, and related issues. iv. Exceeding Authority GTE contends that the PUC and the arbitrator exceeded their authority by adopting the element prices established in the UM 351 proceeding, because neither MCI nor GTE urged that result. The court disagrees. Although neither party urged adoption of the UM 351 prices, the question of “element prices” was properly before the PUC. The parties did not agree on a price list. The PUC was not limited to selecting the prices or terms proposed by one of the parties. If the PUC determines that neither side’s proposal on a particular issue is appropriate, the PUC is not required to select one of the two positions anyway. In this circumstance, the PUC could properly reject both proposals and fashion terms that, in the PUC’s judgment, comport with the legal requirements and the public interest. B. Reliance Upon U.S. West Cost Data The parties agree that the UM 351 proceeding was based on U.S. West’s cost data, not GTE’s. GTE contends the Act requires that element prices be based upon its costs, not those of some other company. i.Consent The PUC contends that GTE waived any objections by consenting to the use of U.S. West’s data in the UM 351 proceeding. The court disagrees. The UM 351 proceeding began in 1990 as an early attempt to identify the costs of various network elements, with the idea that this data might be useful in setting retail rates and experimenting with local competition. In 1993, the PUC suggested that the UM 351 proceeding focus on U.S. West’s cost data, in large part because the PUC did not have the resources to analyze data from the other carriers. GTE agreed to that proposal. However, GTE could not have known at the time that the results of the UM 351 proceeding would be used to set prices for GTE’s network elements under a law that was not enacted until three years later. Nor, given the court’s understanding of the limited purposes originally envisioned, is it proper to estop GTE from contesting the PUC’s decision to base GTE’s prices on U.S. West’s costs. ii.Mootness MCI contends that the issue is moot because, under the FCC pricing regulations, network element prices are based solely upon the costs that theoretically would be incurred by a mythical network that exists only in economic models. GTE’s actual costs, network architecture, and similar details would therefore be irrelevant. MCI’s description of the FCC pricing regulations is less than reassuring. However, the court need not address that question today because the FCC pricing regulations were not in effect when this Agreement was approved, and therefore were not binding upon the PUC. iii.Best Available Information The PUC defends its use of the UM 351 prices, even though based on U.S. West’s cost data rather than GTE’s, on grounds (1) this was the best information available at the time, (2) GTE has not demonstrated that its own costs are materially higher, and (3) this was only an interim price until the PUC could review GTE-specific cost studies. The court concludes that under the circumstances presented here, the PUC’s decision was not arbitrary and capricious or contrary to law. The Act established aggressive timetables for approval of interconnection agreements. The PUC reasonably concluded that it should use the existing UM 351 study, on an interim basis, until it could evaluate new cost studies and fix “permanent” prices specific to each carrier. The PUC has since adjusted U.S. West’s element prices based upon the UM 773 and UM 844 proceedings, and has begun a similar GTE-specific proceeding (which is designated UM 874). GTE contends the PUC had a viable alternative, which was to use the cost studies that GTE submitted. However, the PUC discounted those studies because it disagreed with the underlying methodology utilized by GTE. That decision was not arbitrary and capricious. Nor, as GTE contends, was the PUC obliged to adopt GTE’s flawed cost studies merely because MCI did not file a competing cost study. GTE also contends that the PUC violated § 252(b)(4)(B) by relying on U.S. West’s costs and the UM 351 proceeding. GTE asserts that the PUC may rely upon the “best information available to it” only if GTE “refuses or fails unreasonably to respond on a timely basis to any reasonable request from the” PUC. Id. In GTE’s view, the PUC should have asked GTE to submit additional cost studies or perhaps used portions of the studies that GTE did submit, while rejecting or modifying other portions of those studies. GTE’s proposed solution does not consider the statutory deadlines that the PUC was operating under. There was not sufficient time remaining for GTE to prepare and present new studies, and for the PUC to properly analyze those studies. In that circumstance, the data was constructively unavailable. In addition, since the cost studies GTE submitted did .not comply with the PUC’s proposed methodology, the PUC could properly have concluded that GTE had refused, or failed unreasonably to respond on a timely basis, to a reasonable request from the PUC. Consequently, the PUC was not precluded from utilizing the best information available to it at the time. Nor was the PUC required to dissect GTE’s cost studies, picking out the parts (if any) that either did comply with the PUC’s methodology or perhaps could be modified to comply. While the PUC was not precluded from attempting such a task if it was so inclined, it was not compelled to do so under the circumstances presented here. C. Challenges to Pricing Methodology GTE challenges the pricing methodology utilized by the PUC, which GTE contends does not fairly compensate it for the cost of providing network elements. GTE complains that the PUC’s methodology focuses upon the hypothetical costs of a mythical network that does not exist, while ignoring the actual costs to GTE of providing these elements. GTE also complains that the PUC’s emphasis on “forward-looking costs” improperly denies it compensation for the sizable investment GTE has already made in its existing infrastructure, which the PUC historically has not let GTE recover quickly because the PUC wanted to keep telephone rates low. GTE contends there was a commitment that GTE would eventually be permitted to recover those costs, which has now been broken. Despite some misgivings about the methodology used by the PUC, the court does not believe it is appropriate to decide the validity of that methodology today. Very similar issues are pending before the Eighth Circuit in a consolidated nationwide proceeding that challenges the substance of the FCC’s pricing regulations. The Eighth Circuit’s decision will supersede any ruling this court might make regarding this issue. In addition, at oral argument the parties confirmed that their Agreement does not contain a “true-up” clause, and agreed that even if this court determines that the PUC’s pricing methodology was flawed, MCI has no obligation to compensate GTE retroactively for the difference between the old and new prices on services that have already been sold. Finally, even if this court set aside the PUC’s pricing methodology, the court cannot simply pick numbers out of the air to replace the existing element prices. Rather, the court would have to remand the Agreement to the PUC to devise a replacement methodology and to compute new element prices. In all likelihood, before the PUC could complete that task, the present Agreement will have expired and the Eighth Circuit will have ruled on the challenge to the FCC pricing rules. At oral argument, GTE proposed to stay this issue pending the Eighth Circuit’s decision, but that would unduly complicate any appeal from the final judgment entered in this action. Instead, the court will dismiss, without prejudice, that portion of GTE’s case that specifically challenges the PUC’s unbundled network element pricing methodology. This includes ¶ 162 of Count 1(B), and Count 1(D), of GTE’s Amended Counterclaims and Cross-Claims. D. State Law Claims GTE also asserted several claims alleging that the PUC violated state law during the UM 351 proceeding. GTE later moved to dismiss those claims, without prejudice, because they are simultaneously being asserted in a state court challenge to the UM 351 proceeding. Initially, the court was inclined to deny the motion to dismiss because it believed the decision in the present case would, as a practical matter, be dispositive of the state law claims. For instance, if there was substantial evidence in the record to support the decision for purposes of federal law, then in all likelihood the equivalent state law standards also were satisfied. In addition, the Act gives the federal district courts exclusive jurisdiction to review any challenge to an interconnection agreement. § 252(e)(4) and (e)(6). During oral argument, however, the parties represented that the UM 351 proceeding includes a number of additional matters not at issue here that have independent significance apart from this case, and the state court could entertain and decide those matters without collaterally attacking the judgment entered in the present case. In addition, some of the plaintiffs in the state court case are not parties to the instant proceeding. Finally, it is preferable to let the Oregon courts decide important or novel questions of state law, particularly in areas of pervasive state regulation such as with public utilities. The court therefore grants GTE’s motion to dismiss, without prejudice, Counts VI, VII, and VIII of GTE’s Amended Counterclaims and Cross-Claims. GTE’s alternative motion to file an amended pleading is denied as moot. 2. Takings During oral argument, GTE clarified that it is not directly asserting a takings claim. It does contend that the Act must be construed in a manner that avoids a taking and that GTE must be fully compensated for any taking that occurs, which includes just compensation for the elements and services it is required to sell to MCI. In response, MCI and the PUC assert that the Fifth Amendment does not require that GTE be fully compensated for each unbundled element or other service it must sell to MCI, so long as GTE’s financial integrity as a whole is not impaired. They rely upon the “total effect” test articulated in Duquesne Light Co. v. Barasch, 488 U.S. 299, 312, 109 S.Ct. 609, 102 L.Ed.2d 646 (1989), Federal Power Comm’n v. Hope Natural Gas Co., 320 U.S. 591, 602, 64 S.Ct. 281, 88 L.Ed. 333 (1944). This court previously questioned that argument, U S West Communications, Inc. v. MFS Intelenet, Inc., 35 F.Supp.2d 1221, 1236 (D.Or.1998), and remains skeptical here. Duquesne and Hope Natural Gas applied the “total effect” test in the context of regulated monopolies, in which losses in one area could be offset by altering the rates in another. That may no longer be possible once the market for local telephone service is opened to competition. If GTE is forced to sell unbundled network elements or finished services at below cost to MCI, the PUC cannot simply compensate GTE by raising retail prices for other telephone services, because competitors could then under-price GTE. Moreover, with deaveraging of retail prices on the horizon, the PUC may soon have little control over the retail prices charged for telephone service, nor will it be able to guarantee GTE a particular rate of return. In addition, MCI has vehemently denounced implicit subsidies, by which one service is overpriced to subsidize another, insisting that such subsidies violate the Act. MCI cannot now be heard to argue in favor of such an implicit subsidy so MCI can avoid paying full compensation for the services it purchases from GTE. Finally, the Act expressly requires that network element prices be based upon cost, and that GTE otherwise receive just and reasonable compensation for the services it provides to MCI and for any other takings effected by the Act. The Act does not say that GTE’s other customers must pick up the tab. For purposes of this case, there is no need to conclusively determine whether the “total effect” test continues to have any applicability to local telephone service, or whether GTE could prevail on a takings claim if the challenged prices in the interconnection Agreement denied GTE adequate compensation. Suffice it to say that this court construes the Act to require that just and reasonable compensation be paid for the services GTE provides to MCI, and — except as specifically contemplated in the Act, see fn 8, supra — the PUC may not underprice those services in hopes that revenue from other customers or products will make up for the shortfall. 3. Deaveraging: MCI contends the PUC erred by establishing a single state-wide interim loop price, instead of “deaveraging” those prices into multiple zones (based on density or some other criteria) and charging a different price for each zone. The net effect of MCI’s proposal would be to reduce loop prices in dense urban areas, while significantly increasing loop prices in the rest of Oregon. This court declined to order immediate deaveraging in two prior decisions. MCI Telecom. Corp. v. U.S. West Communications, Inc., 31 F.Supp.2d 859, 860-61 (D.Or.1998); AT&T Communications of Pac. Northwest, Inc. v. U S West Communications, Inc., 31 F.Supp.2d 861, 865 (D.Or.1998). This court concluded that averaged loop prices are cost-based. The PUC simply chose to set a single loop price based upon the average cost of providing service rather than separately calculating the costs attributable to a particular loop and pricing each loop accordingly. MCI Telecommunications Corp. v. U.S. West, 31 F.Supp.2d at 860. The court also noted that “[t]he PUC has not refused to deaverage loop prices, but wants to do it in an orderly fashion” and planned to address deaveraging in a separate proceeding. Id. Finally, the court acknowledged concerns, voiced by both U.S. West and the PUC, that the deaveraging of loop prices needs to be coordinated with the deaveraging of retail prices for those services and the implementation of explicit universal service programs, which was lagging. Unlike CLECs such as MCI, an ILEC is legally obligated to service all customers within its territory, at the same fixed price. Consequently, MCI and other CLECs could solicit the most profitable customers, leaving the ILEC to service the unprofitable accounts the CLECs do not want. Id. at 860-61. The reasoning expressed in the earlier cases is equally applicable here. Nothing has changed during the intervening three months to convince this court to reverse its position. MCI argues that an FCC regulation, 47 C.F.R. § 51.507(f), now mandates deaver-aging into at least 3 zones. However, for the reasons discussed earlier, this court concludes that § 51.507(f) was not in effect during the time this Agreement was adopted and the loop prices were established, hence the PUC did not have to follow that regulation. Deaveraging into at least three zones is not a requirement imposed directly by the Act, which would have to be given retroactive effect. The Act does not even mention deaveraging, let alone require a minimum of three zones. Rather, this requirement is a creation of the FCC, issued pursuant to its general regulatory powers. See AT & T, — U.S. at -, 119 S.Ct. at 729-33. Whether the FCC has the authority to require deaveraging will be decided by the Eighth Circuit. Even assuming the regulation is affirmed, the PUC is not required to retroactively apply that regulation to interconnection agreements that were approved before the regulation took effect. MCI also argues that when approving this particular Agreement' — and unlike some other interconnection agreements this court has reviewed — the PUC deleted the arbitrator’s discussion of universal service and related considerations, and said the reason it was not deaveraging loop prices was because it lacked sufficient information to do so. Accordingly, MCI contends the court should focus solely on that justification. MCI says it provided the arbitrator with calculations that would have established six different loop prices, and contends this information was enough to implement deaveraging. However, MCI concedes that when the arbitrator began to review those calculations, he quickly realized that at least some of the calculations could not possibly be correct. MCI’s witness acknowledged the problem, and suggested it was likely a mathematical error that could be corrected, a position MCI asserts in this proceeding. Perhaps it was. However, the record excerpts reviewed by the court indicate that this document was furnished at the last minute to the parties and the arbitrator, and there had been little opportunity to verify the underlying assumptions and calculations. Upon discovering a glaring error in some critical calculations, the arbitrator understandably lost confidence in MCI’s data and was unwilling to adopt it on faith without a chance to thoroughly analyze the data and the assumptions behind it. That decision was not arbitrary and capricious. The court also observes that the Oregon PUC was aggressively moving toward implementing unbundled elements and local competition long before Congress passed the Act. This is not a case where the PUC had to be forced into embracing local competition. Accordingly, when the Oregon PUC concludes that deaveraging in Oregon is premature, or it does not have the information it needs to implement deaver-aging in an orderly manner, this court is inclined to give considerable deference to .the PUC’s opinion. 4. Wholesale Discounts The arbitrator fixed the interim wholesale discount at 15.95 percent, as proposed by MCI. However, the arbitrator made an exception for services already subject to a volume or term discount, for which he awarded MCI only one-half the wholesale discount (in addition to the volume or term discount). MCI and GTE each challenge portions of the decision. A. Volume and Term Discounts MCI contends it is entitled to receive the full wholesale discount in addition to the full volume or term discount. The court disagrees. In all likelihood, the existing volume or term discount already takes into account some of the cost savings that could reasonably be achieved from selling the product at wholesale (e.g., billing, marketing, etc.). MCI is not entitled to deduct the same cost savings twice. While there may be room for some additional savings, it is unlikely to be the full 15.95 percent. AT&T v. U S West, 31 F.Supp.2d at 868. The arbitrator therefore applied a lower wholesale discount rate in this situation. The PUC’s decision to allow one-half the regular wholesale discount, in addition to the volume or term discount, is a reasonable interim compromise. However, the PUC must expeditiously complete any studies or hearings needed to identify the additional costs savings that should reasonably result, and adjust the wholesale discount on these services accordingly. Id. B. Retail-Exit Methodology GTE challenges the methodology that the PUC used to compute the wholesale discount. 47 U.S.C. § 252(d)(3) provides, in relevant part, that: [A] State commission shall determine wholesale rates on the basis of retail rates charges (sic) to subscribers for the telecommunications service requested, excluding the portion thereof attributable to any marketing, billing, collection, and other costs that will be avoided by the local exchange carrier. To implement this section, the PUC asked what costs GTE would avoid if it abandoned the retail business entirely, and then assumed that 100 percent of those costs were avoidable for purposes of calculating the wholesale price. This is commonly referred to as the “retail-exit” methodology, which the PUC borrowed from the FCC. GTE protests that it is not leaving the retail business, and therefore cannot reasonably avoid all of the costs that the retail-exit method presumes are avoidable. Many of those costs are fixed. If Sam Customer transfers his residential service to MCI, which purchases it from GTE at a wholesale discount, the half-page advertisement that GTE runs in the local newspaper every Sunday won’t cost any less now that Sam is an MCI resale customer. The advertising agency and artist that created the ad are unlikely to reduce their fees either. GTE will save the cost of the stamp to mail Sam his monthly bill, but the cost to design the software program to track the charges to Sam’s account and to print his bill should not vary much. The cost of the machines to print bills, and to open envelopes, is unlikely to change dramatically either. Nor, unless a large number of customers abandon GTE for MCI, will there be a significant reduction in personnel, support staff (such as human resources), or office space requirements. GTE also questions the PUC’s assumption that retail costs will decrease in a “straight-line,” a fixed amount for each wholesale unit. More likely, many of the cost savings will be “lumpy.” For instance, if GTE needs one envelope-opening machine for each 50,000 retail customers, then a reduction of 50 or even 500 retail customers will have no impact on the number of machines it needs. The cost savings will accrue only when that critical mass is reached, if it ever is, and will accrue in “lumps” rather than on a straight-line basis. MCI offers several responses to GTE’s criticisms. First, it contends that the retail exit methodology is required by binding FCC regulations. However, those regulations were not in effect when the PUC adopted this agreement and set the wholesale prices, and therefore are not controlling here. Second, MCI argues that GTE’s actual cost savings are irrelevant. Instead, the focus should be on what someone in the wholesale business would charge for these services. Finally, MCI argues that the retail-exit and straight-line methodologies are necessary to prevent GTE from using the wholesale prices to subsidize its retail business. For instance, while GTE will continue to pay the same price for its display ad in the newspaper, MCI argues that unless the cost of that advertisement is deducted from the wholesale price, MCI will be helping to pay for the cost of that advertisement, from which it derives no benefit. Consequently, when setting the wholesale price, MCI believes the PUC should consider only the advertising, billing, collection, and similar costs that GTE directly incurs in order to provide the particular unit of service that is being sold to MCI, without regard to whether GTE’s operations (when viewed as a whole) actually realize any reduction in cost. Otherwise, the wholesale discount might not be enough for MCI to earn a profit reselling GTE services. At oral argument, the parties suggested deferring this issue until the Eighth Circuit rules on the challenges to the FCC wholesale pricing regulations. The court reluctantly concurs. The PUC’s methodology appears to be the same as the FCC’s methodology. Therefore, the Eighth Circuit’s decision will, as a practical matter, supersede any ruling by this court on the issue. As with the TSLRIC issue, the court will dismiss, without prejudice, the portion of Count 1(A) of GTE’s Amended Counterclaim and Cross-Claims that challenges the methodology used by the PUC to compute the wholesale discount (excepting those issues specifically addressed below). C. Unitary Discount GTE also challenges the PUC’s decision to set a single discount price for all telephone services made available for resale (except for volume and term discounted services), notwithstanding that there may be a wide variation in the cost savings to GTE for different services, and MCI is not required to purchase all of its services from GTE. GTE’s concern is illustrated by analogy to a store owner who sells two kinds of candy bars. Bar-A should retail for 25 cents, while Bar-B should retail for 45 cents. To simplify matters, he decides to price both bars at 35 cents. That plan works only if he sells at least as many units of Bar-A as he does of Bar-B. It does not work if customers purchase only Bar-B at his store, while purchasing Bar-A from a competing store that continues to price it at just 25 cents. With local telephone service, it is generally acknowledged that the ILEC’s potential cost savings from resale may be lower for some services (e.g., basic residential service) than for others (business, vertical features). If a unitary discount is utilized for all services, the net effect might be to artificially increase the wholesale discount for residential services while decreasing the wholesale discount for some other services. That is not a problem if MCI has to purchase all of its services under the same plan, but it does not. MCI can selectively purchase residential services at the resale price (and receive the benefit of the higher wholesale discount) while buying business and vertical features at the unbundled element price. If the actual difference in cost savings between the various products is substantial, then GTE is effectively in the position of the shopkeeper selling 45 cent candy bars for 35 cents. That would not be just and reasonable compensation, and it would be a violation of 47 U.S.C. § 252(d)(3). The PUC understandably was reluctant to complicate matters by setting too many discount rates. The court agrees with that basic premise, and is not suggesting that there must be a separate discount rate for each service GTE offers. However, in setting the discount rate(s), the PUC must consider the range of cost savings for different categories of services, as well as the potential for abuse through selective ordering tactics, and determine whether more than one rate is needed. The PUC did not give sufficient consideration to this issue or to all relevant factors, and placed undue emphasis on a single factor (simplicity). Therefore, the court remands the wholesale discount issue for reconsideration. In view of the pending Eighth Circuit challenge to other aspects of the FCC’s wholesale pricing regulations, the PUC may exercise its discretion in deciding what additional studies or proceedings to undertake at this time in furtherance of this remand. D. Costs of Wholesaling GTE contends that in computing cost savings for purposes of the resale discount, the PUC must take into account the new costs that GTE will incur as a wholesaler. The court agrees. For instance, although GTE will no longer have to bill and collect from these particular retail customers, it will now have to bill and collect from an assortment of CLECs. GTE may no longer have to be concerned about retail customers who skip'town, but it now has to be concerned about CLECs who go bankrupt. Not all prospective CLECs are as credit-worthy as MCI and AT & T. MCI’s contention that the PUC must focus only on the retail cost savings, while ignoring any offsetting increase in wholesale expenses, is equivalent to balancing a savings account by counting the deposits but not the withdrawals. The PUC responds that it did make an offset for such expenses by discounting only 90 percent of certain retailing cost categories instead of 100 percent. However, it is not clear whether 100 percent of those costs would have been avoidable anyway, or how the PUC calculated the additional costs of selling services at wholesale. The court will remand this issue to the PUC for reconsideration and to articulate a clearer explanation of the reasons for the PUC’s decision. E. Which Services are Subject to the Wholesale Discount 47 U.S.C. § 251(c)(4)(A) requires GTE to “offer for resale at wholesale rates any telecommunications service that the carrier provides at retail to subscribers who are not telecommunications carriers.... ” The parties do not agree on a list of “retail” services or even the definition. The dispositive question ordinarily is not whether a particular customer is a “retail” customer but whether the particular service in question is one that GTE “provides at retail to subscribers who are not telecommunications carriers.” That intent is evident from the language of the statute, and also from the use of retail prices as the starting point in computing a wholesale price. Most of the time there will be many retail customers for each service, a tariffed price, and little room for dispute. In a few circumstances, though, there may be only a handful of subscribers to that service, hence it will be necessary to look more closely at the specific customers involved. At oral argument, the attorneys for the parties attempted (with limited success) to explain each of the services in question. The court’s analysis of this issue is premised upon those explanations. i.Directory Assistance and Operator Services If retail subscribers are typically charged a fee for using these services, then they qualify as “retail” services for purposes of § 252(d)(3). Whether GTE actually saves any costs by providing these services to MCI at wholesale is another question, but that is something the PUC must examine when it reconsiders its decision to establish a unitary discount rate. ii.Non-Recurring Costs These include a variety of one-time only charges incurred when retail service is established, resumed, or transferred. If the non-recurring charge is customarily billed to the subscriber, then it is a retail service, though the amount of any cost savings to GTE is something the PUC needs to consider in deciding what discount rate to apply. On the other hand, if it is a cost that ordinarily is absorbed by the carrier, as part of the general cost of providing service, then it is not a retail service and need not be discounted. However, the item potentially could qualify for treatment as an unbundled network element. The parties failed to provide the court with sufficient information about the particular non-recurring costs at issue to issue a final ruling. By March 31, 1999, the parties shall advise the court which of the disputed non-recurring costs are retail services (and which are not) under the definition set forth above. If the parties are unable to agree, then the court will make that determination on the basis of the information the parties provide. iii.Pay Phone Lines The next disputed service is “public and semi-public pay phone services and COCOT coin and coin-less lines.” The parties defined these as lines that GTE sells to persons who install and operate pay telephones. GTE contends these are not retail subscribers but wholesalers because they resell the use of these lines to the public on a per-call basis. It is an interesting question, but the court ultimately is not persuaded by GTE’s argument. First, it does not appear that the FCC presently classifies pay phone providers as telecommunications carriers. Second, the line is not intended to be resold to retail customers, notwithstanding that numerous individuals may place calls on that phone. Finally, it is the court’s understanding that anyone can install a pay phone in his deli (or teenager’s bedroom), call GTE, and obtain a line for that phone at the generally applicable price. These are essentially retail customers. There are, of course, companies that operate numerous pay phones and purchase multiple lines from GTE, but that does not necessarily make them “telecommunications carriers” nor does it alter the fact that the service is one GTE makes available to retail customers. To the extent that customers who purchase multiple pay phone lines receive a volume discount from GTE, the wholesale discount may be lower, but GTE nevertheless must make these services available for resale to MCI. iv. Individual Case Basis (“ICB”) Contracts At oral argument, the parties surmised that these were probably contracts negotiated with very large business customers with specialized needs. It is unclear how many of these contracts exist, and the parties were vague about other details. It is not clear whether the contracts are for services generally available to the public, or for unique combinations of network elements designed for a particular client. Nor is it clear how these services are priced, or what costs GTE would save by wholesaling these services to MCI. To the extent the wholesale price would be premised on a contract between the customer and GTE, it also is unclear what will happen when the contract expires, or whether the customer is permitted to switch carriers prior to then. The court is reluctant to decide this issue without an adequate factual foundation, which the parties have failed to provide. The agency’s written decision provides little additional insight. By March 31,1999, the parties shall file supplemental briefs or other materials providing a more detailed explanation of ICB contracts and addressing the issues identified above. The court will then decide this issue (if there is adequate information) or else remand to the PUC to articulate a clearer explanation of the issue and of the rationale for the PUC’s decision. 5. Availability of Unbundled Elements by Tariff The PUC ordered GTE to publish a tariff listing elements that the PUC has decided must be unbundled and the prices that the PUC has fixed for those elements. Although the briefing was not fully informative, it appears that prospective CLECs can order services “off-the-rack” without entering into an interconnection agreement. GTE labels this an “end-run” around the Act, and complains that it denies GTE the opportunity to negotiate different terms with each CLEC. GTE asks the court to declare that the PUC’s tariff requirement is preempted by the Act and is therefore void pursuant to the Supremacy Clause of the United States Constitution. The court rejects most of GTE’s underlying premise. Nevertheless, the court concludes that the tariff, as presently structured, conflicts with the Act and is therefore preempted. Generally speaking, the PUC is not precluded from setting unbundled element prices and wholesale discounts for a particular ILEC, and using those same prices and rates in all interconnection agreements involving that ILEC. The court rejects GTE’s contention that GTE must be permitted to negotiate different prices with each CLEC. This court previously questioned the PUC’s decision to establish different wholesale discount rates for each CLEC. US West Communications, Inc. v. AT & T Communications of the Pac. Northwest, Inc., 31 F.Supp.2d 839, 845-46 (D.Or.1998). GTE’s cost savings should not differ markedly depending upon which CLEC is buying the services. Likewise, GTE’s costs to produce a network element should not vary materially depending upon which CLEC is purchasing that element. Several provisions in the Act reinforce this conclusion. An agreement under which GTE provides services at lower prices to its preferred CLECs might be deemed collusive and violate the non-discrimination requirement in 47 U.S.C. § 251(c)(4). In addition, once GTE provides an element or service pursuant to an interconnection agreement with one CLEC, GTE must make that element or service available to any other CLEC “upon the same terms and conditions.” 47 U.S.C. § 252(i). Finally, the requirement that the “State commission” determine the wholesale discount prices, 47 U.S.C. § 252(d)(3), negates any inference that GTE is entitled to negotiate different wholesale discounts with each CLEC. The court also acknowledges the PUC’s concern that the cost of negotiating (and possibly litigating) a custom interconnection agreement is prohibitive for many prospective CLECs. In theory, a CLEC could avoid litigation by signing a contract acceptable to GTE, but that effectively would allow GTE to dictate the terms. However, not all CLECs need a custom interconnection agreement. Some CLECs merely want to purchase services for resale, and do not plan to physically interconnect with GTE’s network. The PUC is not precluded from adopting a universal “short-form” interconnection agreement for use in this circumstance. Upon signing the agreement, a CLEC could purchase services “off-the-rack” at the established prices. A universal short-form agreement might also be appropriate for some CLECs purchasing unbundled elements, so long as there is sufficient opportunity for the parties to address any technical issues regarding that interconnection, and the PUC ensures that GTE is compensated (to the extent required by the Act) for any special costs associated with a particular interconnection agreement that are not already included within the unbundled element prices. Of course, a CLEC could still negotiate a custom agreement, if it chose. Admittedly, the Act does not specifically provide for this short-form procedure, but the Act does not forbid it either. The primary goal of the Telecommunications Act of 1996 was to open local telephone markets to competition, and this procedure furthers that goal and is not inconsistent with either the terms or the purposes of the Act. Although the court rejects GTE’s contention that wholesale discounts, unbundled element prices, and other terms invariably must be separately negotiated with each CLEC, the court nevertheless agrees with GTE that the tariff — as presently structured — conflicts with the Act an