Opinion
No. 3:10-cv-00806 (JAM)
03-31-2015
THE SOUTHERN NEW ENGLAND TELEPHONE COMPANY d/b/a AT&T CONNECTICUT, Plaintiff, v. KEVIN M. DELGOBBO, ANTHONY J. PALERMINO, JOHN W. BETKOSKI III, AMALIA VASQUEZ BZDYRA, and ANNA M. FICETO, in their official capacity as Commissioners of the Connecticut Department of Public Utility Control, Defendants.
ORDER AFFIRMING THE DECISION OF THE CONNECTICUT DEPARTMENT OF PUBLIC UTILITY CONTROL
This is a civil action arising under the Telecommunications Act of 1996 and its implementing regulations. Plaintiff, the Southern New England Telephone Company d/b/a AT&T Connecticut, seeks judicial review of a decision by defendant commissioners of the Connecticut Department of Public Utility Control (DPUC) concerning rates that plaintiff may charge competitors to use parts of plaintiff's local telephone network facilities ("interconnection" rates). The DPUC decision evaluated cost studies plaintiff had provided to support its proposed interconnection rates, found them to be sorely inadequate, and ordered plaintiff to file rates that included adjustments to portions of its studies. The resulting rates were lower than plaintiff's initial cost studies had supported.
Several of plaintiff's concerns with the DPUC decision have been resolved by related litigation. Remaining in the case are claims that DPUC acted arbitrarily and capriciously when it made adjustments to plaintiff's cost studies with regard to two of the approximately 15 elements that went into the rate-setting analysis: the "fill factor" and the "line mix." For the reasons set forth below, I affirm the decision of the DPUC.
BACKGROUND
The Telecommunications Act of 1996 imposes obligations on established, formerly monopolistic phone companies (generally referred to as "incumbent local exchange carriers" (ILECs)) to provide certain services to newer entrants in a local telecommunications market ("competitive local exchange carriers" (CLECs)), in order to foster greater competition in the market. ILECs must share their networks by "interconnecting" with CLECs' networks, and must negotiate fair contracts by which CLECs will pay for the interconnection services.
"Interconnection" is defined as "the linking of two networks for the mutual exchange of traffic" and "does not include the transport and termination of traffic." 47 C.F.R. § 51.5.
If voluntary negotiations fail, the parties may petition a state utility commission to step in and arbitrate the dispute. In order to ensure that the resulting interconnection rates are "just, reasonable, and nondiscriminatory," 47 U.S.C. § 251(c)(3), state agencies that resolve a dispute must determine the appropriate rates using "forward-looking" cost methodology, see 47 C.F.R. § 51.505, and they may not consider the historic or embedded costs that an ILEC incurred for an element in the network. See id.(d)(1); see also 47 U.S.C. §252(d)(1)(A)(i) (a state commission's determinations of the interconnection rate must be "based on the cost (determined without reference to a rate-of-return or other rate-based proceeding) of providing the [service]").
As Judge Easterbrook has explained, this forward-looking methodology means that "[i]ncumbents that have aging and inefficient equipment thus must sell for less than their historical cost; the old system that calculated rates based on actual cost of equipment plus a reasonable rate of return on capital is out the window." AT&T Commc'ns of Illinois, Inc. v. Illinois Bell Tel. Co., 349 F.3d 402, 405 (7th Cir. 2003) (Easterbrook, J.) (hereinafter Illinois Bell). The forward-looking "total element long-run incremental cost" (TELRIC) methodology (47 C.F.R. § 51.505(b)) attempts to approximate rates based on the costs that a hypothetical ILEC would incur to provide the services in an efficient market, and this methodology has been approved by the Supreme Court as an appropriate forward-looking methodology for state commissions to apply. See generally Verizon Commc'ns., Inc. v. F.C.C., 535 U.S. 467 (2002).
Connecticut's Department of Public Utility Control uses the "total services long run incremental cost" (TSLRIC) methodology, which is functionally equivalent to the TELRIC methodology for the purpose of this case. See S. New Eng. Tel. Co. v. Perlermino, 2011 WL 1750224, at *6 n.3 (D. Conn. 2011), aff'd sub nom. S. New Eng. Tel. Co. v. Comcast Phone of Conn., Inc., 718 F.3d 53 (2d Cir. 2013). I refer to Connecticut's methodology as TELRIC simply for ease of reference in this ruling.
TELRIC is not a precise algorithm; it is "a framework rather than a formula," with "considerable play in the joints." Illinois Bell, 349 F.3d at 405. For example:
Incumbent carriers may be unable to agree with would-be entrants about what the most efficient technology is, how much it would cost to construct, and what the incremental costs of a given network element would be. Moreover, even when the parties can agree on the technology, they may be unable to agree on vital details.Ibid. Ultimately, a state commission must ensure that the interconnection rate is "just, reasonable and nondiscriminatory." 47 U.S.C. § 251(c)(3). In order to do so, it will ask the ILEC to submit a cost study that shows that the rates it offers do not exceed the forward-looking economic cost of the service. See 47 C.F.R. § 51.505(e).
The present case arises from a series of disputes between an ILEC, plaintiff Southern New England Telephone Company d/b/a AT&T Connecticut, and competitor CLECs in Connecticut. In October 2008, wireless carrier Youghiogheny Communications-Northeast, LLC d/b/a Pocket Communications (Pocket) petitioned defendant, the Connecticut Department of Public Utility Control (DPUC), for arbitration of its dispute with plaintiff over rates plaintiff could charge CLECs for "reciprocal compensation" See Doc. #39 at 51. Reciprocal compensation refers to the per-minute rate plaintiff would charge a CLEC if that CLEC's customer made a local call to an AT&T Connecticut customer, thereby using plaintiff's network. Pocket argued that plaintiff's rates, which relied on a ten-year-old cost study, were out of date and unreasonably expensive. An arbitrator appointed by DPUC evaluated the parties' filings and, in March 2009, recommended that there be a new cost study for reciprocal compensation rates. See id. at 59. DPUC agreed, and in April 2009, it ordered plaintiff to file a new cost study to support its desired reciprocal compensation rates by July 2009. Id. at 63.
DPUC has recently been reorganized and renamed the Public Utilities Regulatory Authority (PURA). Because the challenged decision and all the relevant procedural history took place under the previous scheme, I refer to the state commission in this ruling as DPUC.
Meanwhile, in December 2008, Pocket filed a second petition, this time requesting declaratory judgment with regard to plaintiff's rates for the provision of transit services. See id. at 67. Such rates reflect the per-minute price plaintiff would charge a CLEC when that CLEC's customer made a local call to a customer of a third telecommunications carrier, if plaintiff's network were used as an intermediary to connect the call. Plaintiff argued, inter alia, that transit traffic did not qualify as "interconnection" under the Telecommunications Act of 1996, and that it therefore did not need to be provided via TELRIC-based rates. Id. at 71-72. DPUC disagreed. In October 2009, it entered an order stating that transiting qualified as interconnection and therefore required TELRIC-based rates. Id. at 97-101. DPUC imposed an interim rate pending the resolution of the docket assigned to the cost study for reciprocal compensation, to which at some point DPUC had also asked plaintiff to include a study of transit rates. Id. at 103-04.
Plaintiff filed its cost studies in July 2009, as required, and filed additional materials on several occasions in the subsequent months, at DPUC's request. In September 2009, plaintiff and several intervening companies (Pocket, Sprint, Comcast, Cablevision, and Cox) filed testimony regarding the new cost studies. DPUC held evidentiary hearings over several days and issued its final decision in April 2010, ordering plaintiff to file rates for both reciprocal compensation and transit services in accordance with several changes to the TELRIC calculation undertaken in its cost studies. See Doc. #29-1 at 4.
That decision made repeated references to plaintiff's failures in supporting its proposed rates. To begin with, DPUC found that plaintiff's cost studies "initially . . . could not be analyzed in any level of detail," because the filings "consisted of a single Excel spreadsheet for each service containing only summary cost information" that "failed to meet the requirement established . . . for cost studies to be documented in a manner that the source of the data can be audited." Id. at 16-17. Even after plaintiff provided additional information in response to interrogatories, DPUC found that "even then, it was difficult to understand [plaintiff's] explanation as to how these various spreadsheets should be utilized in order to understand the costs that were in the July 17, 2009 filings." Id. at 17. And plaintiff's several delays in providing information upon DPUC's request "resulted in the Department and the parties receiving [one study] only four calendar days (including a weekend) before the evidentiary hearing began." Ibid. All in all, DPUC found that "[plaintiff's] failure to provide its cost studies on a timely basis negatively impacted [DPUC] and the parties' ability to thoroughly analyze [plaintiff's] cost studies. Ibid.
DPUC also found that the problems of plaintiff's "incomplete cost studies were further exacerbated by [plaintiff's] inadequate job of supporting them during cross-examination," in that "[t]he testimony of [plaintiff's] witnesses was decidedly deficient as the basis for the validity of the cost studies." Ibid. Nevertheless, DPUC utilized plaintiff's "inadequate" studies, with several adjustments, to determine appropriate cost-based rates for reciprocal compensation and transit services.
Most of the costs in the studies related to the costs of using plaintiff's switches, which channel incoming data to the appropriate outgoing port. One of the adjustments DPUC ordered was to the proportion of "replacement" versus "growth" lines that plaintiff anticipated purchasing for its network. It is critical to this case to understand the distinctions between replacement and growth lines. Plaintiff based its switching cost studies on contracts with switch vendors, who provide per-line prices based on the number and type of lines expected to be purchased by plaintiff over the term of a contract. Replacement lines are added to a switch to fully replace another switch, while growth lines are simply added to an existing switch. Vendors charge far more for growth lines than replacement lines, because vendors recognize that once a telecommunications carrier such as plaintiff has purchased replacement lines along with a new switch from that vendor, it will have to purchase its future growth lines for that new switch from the same vendor (having already committed to that vendor's brand of switch).
Plaintiff's cost studies assumed a 50/50 line mix of replacement/growth lines, based on negotiations it had with vendors in 1999 for the 2000-2004 timeframe. DPUC rejected this assumption for several reasons, finding it too heavily weighted in favor of expensive growth lines. First, it found that plaintiff had not pointed to any evidence that the 50/50 mix was actually negotiated with its vendors or that plaintiff was in any way locked into purchasing that line mix. Id. at 25. Second, it noted that the record showed negative growth since 2000 and projected negative growth into the future, and therefore found that plaintiff's "proposed mix is neither reflective of its actual historic purchases, nor does it produce accurate forward-looking costs filed in this proceeding." Ibid.
Finally, it rejected plaintiff's argument that changing the line mix would necessitate consideration of any effects on the pricing for replacement and growth lines in its vendor contracts. DPUC found that it was "entirely unclear how adopting [plaintiff's] proposed mix would impact its relationship with its vendors," and that, although a 50/50 line mix assumption might raise plaintiff's revenues (because of the increased switching costs and therefore interconnection rates plaintiff could charge CLECs), there was "nothing in the record that suggests that vendors would benefit from [plaintiff's] increased revenues or how vendors would be harmed by lower revenues." Ibid.
Instead of accepting plaintiff's proposed 50/50 line mix, DPUC looked to an arbitration order issued by the Federal Communications Commission (FCC) with regard to a dispute in Virginia, which adopted line mixes based on an objective algorithm that took into account projected growth rate, cost of capital, and the economic life of the switch. See In Re Worldcom, Inc., 18 F.C.C. Rcd. 17722, 17881-82 (2003) (hereinafter "Virginia Arbitration Order"). DPUC found that the Virginia Arbitration Order's algorithm was superior to plaintiff's approach for determining the appropriate mix because "the underlying inputs can be verified, are [AT&T] specific, and the mix is derived via an objective algorithm rather than the subjective beliefs of an AT&T witness." Doc. #29-1 at 26.
Thus, DPUC calculated a new line mix, using the algorithm described in the Virginia Arbitration Order and the cost of capital and depreciation inputs proposed by plaintiff. DPUC's calculation assumed a 3% growth rate, which it considered conservative, given that the data showed actual and forecasted negative growth for plaintiff. The algorithm produced a mix of 85.1% replacement lines and 14.9% growth lines, which DPUC ordered plaintiff to use as the appropriate mix to input in its cost studies. The upshot of DPUC's line-mix analysis was to conclude that plaintiff's cost studies—because they included too low a ratio of less expensive replacement lines to more expensive growth lines—resulted in too high a cost element for the line-mix component of its projected cost structure; DPUC therefore ordered plaintiff to replace its proposed 50/50 line mix with the 85/15 line mix produced by DPUC's application of the methodology described in the Virginia Arbitration Order.
Apart from this line mix factor, DPUC also ordered an adjustment to plaintiff's cost studies with regard to a "fill factor," which represents the portion of switches' capacity that are anticipated to be actually used. Plaintiff's cost studies had included as one element the ratio of trunks carrying traffic on a switch to the total number of installed trunks on the switch, and plaintiff had proposed that ratio be captured by inputting in its model a fill factor of 72.5%, which fell at the midpoint of a range it considered healthy. Id. at 30. Elsewhere in plaintiff's cost studies was something called the "CCS assumption," which accounted for the expected volume of traffic during the busy hours of the day and was expressed as a percentage that indicated the amount of time that traffic was carried over a single trunk. Put another way, the CCS figure describes what percentage of the time a trunk is in use.
CCS stands for "Centum Call Seconds" and represents increments of 100 seconds. For example, there are 3600 seconds and therefore 36 CCS in an hour. See Doc. #29-1 at 30 n. 163.
DPUC found, based on filed testimony and testimony elicited during the evidentiary hearing, that the CCS assumption served to lower the overall fill/utilization of plaintiff's trunk facilities in the studies. Ibid; see also Doc. #56-1 at 16-29. DPUC therefore ordered that plaintiff re-run its studies with an "effective overall trunk fill/utilization no higher than the 'midpoint' value proposed by [plaintiff]," meaning that the "overall effective utilization/fill should take full account of the cumulative interaction" between the CCS assumption and the trunk utilization ratio previously input as the fill factor, in order to result in an "effective fill/utilization that is not higher than [plaintiff] AT&T's proposed number." Doc. #29-1 at 30. DPUC suggested that one way plaintiff could achieve this would be to change the fill factor input in plaintiff's model from 72.5% to 95% and leave the CCS values as is, so that the 95% fill factor multiplied by the CCS percentage would result in a figure not higher than 72.5% that would ostensibly reflect network usage in terms of both the proportion of trunks in use and the proportion of time they were carrying traffic. See id. at 30 & n.166.
Based on these and other modifications to plaintiff's cost studies, DPUC directed plaintiff to perform compliance runs of its cost studies with the specified modifications and to file those with DPUC for approval. DPUC also ordered the final TELRIC rates—based on approval of the compliance runs—to be made available to all CLECs and wireless carriers in Connecticut. Id. at 5, 40.
Plaintiff timely filed the present case in this Court in May 2010, claiming that the DPUC's cost-study decision violated the Telecommunications Act of 1996 in several ways. See Doc. #1. It alleged that its transit services were not "interconnection" (Count One), that TELRIC pricing did not apply to its transit services even if they were "interconnection" (Count Two), that plaintiff should not be required to offer its transit services with the option of a "bill clearinghouse" function (Count Three), and that DPUC's decision was arbitrary and capricious with regard to the two cost inputs described above: the fill factor (Count Four) and the line mix (Count Five). This case was temporarily stayed pending the resolution of judicial review of DPUC's transit-rate decision from October 2009.
In 2011, in the federal case arising out of plaintiff's DPUC proceedings with regard to the transit-rate decision, the Court (Eginton, J.) reviewed DPUC's transit-rate decision and affirmed that transit traffic qualified as interconnection and that DPUC had jurisdiction to address the issue. See S. New Eng. Tel. Co. v. Perlermino, 2011 WL 1750224, at *6 (D. Conn. 2011). It reversed the DPUC transit-rate decision with regard to one matter, holding that DPUC should not have imposed an interim TELRIC rate without giving the parties the opportunity to voluntarily renegotiate. Id. at *8. The Second Circuit affirmed Judge Eginton's decision in its entirety. S. New Eng. Tel. Co. v. Comcast Phone of Conn., Inc., 718 F.3d 53 (2d Cir. 2013).
That decision resolved Counts One and Two of the instant case, and plaintiff is no longer pursuing Count Three. I therefore consider only Counts Four and Five of plaintiff's complaint, which challenge DPUC's modifications to its cost studies with regard to the line mix and fill factor inputs. For the reasons set forth below, I affirm DPUC's decision with regard to those inputs.
Although the complaint references both reciprocal compensation and transit services, plaintiff's counsel clarified at oral argument that the remaining challenges in Counts Four and Five pertain only to rates for transit services.
DISCUSSION
This Court has jurisdiction to review state commission decisions and determinations with regard to interconnection rates. 47 U.S.C. § 252(e)(6). A state commission's interpretations of federal law, including the Telecommunications Act of 1996, are subject to de novo review. See Perlermino, 2011 WL 1750224, at *2; WorldCom, Inc. v. Conn, Dep't of Pub. Util. Control, 375 F. Supp. 2d 86, 91 (D. Conn. 2005). Otherwise, "[o]nce federal courts determine that state commissions properly interpreted the Act and its regulations, courts apply an arbitrary and capricious standard to review the remaining state commissions' determinations." WorldCom, 375 F. Supp. 2d at 92. Here, the parties have agreed in their briefing and as clarified at oral argument that plaintiff's remaining challenges to the DPUC findings under Counts Four and Five of the complaint are governed by the arbitrary-and-capricious standard of review.
A state commission's determination is "arbitrary and capricious" if it "has relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise." Islander E. Pipeline Co., LLC v. Conn. Dep't of Envtl. Prot., 482 F.3d 79, 94 (2d Cir. 2006) (quoting Motor Vehicle Mfrs. Ass'n of the U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 42-43 (1983)). Although a court "'may not supply a reasoned basis for the agency's action that the agency itself has not given,'" it nevertheless "'will . . . uphold a decision of less than ideal clarity if the agency's path may reasonably be discerned,'" as long as the court can do so "'on the basis articulated by the agency itself.'" Id. at 94-95 (quoting State Farm, 463 U.S. at 43, 50).
This highly deferential standard of review is "particularly appropriate when reviewing findings of fact made by an agency in enforcing the 1996 Act." Worldcom, 375 F. Supp. 2d at 92 n.8 (internal quotation marks and citation omitted). Nowhere are the reasons for this more salient than when a state commission applies the TELRIC methodology to determine the rates ILECs can charge CLECs for services. The TELRIC methodology is "hypothetical and prospective" and does not lend itself to any "right" answers, only "better and worse estimates." MPower Commc'ns Corp. v. Illinois Bell Tel. Co., 457 F.3d 625, 630 (7th Cir. 2006) (Easterbrook, J.). Nor does it even contain any "tried-and-true or mandatory elements." Id. at 631. And "[b]ecause rate-making is an exercise that requires a high level of technical expertise, only egregious errors by an agency will justify judicial intervention." TDS Metrocom, LLC v. Bridge, 387 F. Supp. 2d 935, 938 (W.D.Wis. 2005). Therefore, "[w]hat is important for the purpose of this review is that the commission considered the arguments of both sides, arrived at a conclusion that is supported by evidence in the record and gave an adequate explanation for its conclusion." Id. at 943.
Moreover, any factual mistake a state commission might make along the way with regard to an individual input into the analysis is not cognizable unless it can be said that the final rate required by DPUC was "unreasonable or unsupported by substantial evidence." MPower, 457 F.3d at 629. In other words, "[a]ll a court need do is determine whether the [state commission's] bottom line is supported by the record." Id. at 630; see also Qwest Corp. v. Boyle, 589 F.3d 985, 994-95 (8th Cir. 2009) (potential concerns with a state commission's findings not legally meaningful where they "do[ ] not necessarily mean that the resulting rates deviate from TELRIC").
Similarly, the FCC reviews applications under 47 U.S.C. § 271 from ILECs wishing to provide in-region long distance service. See AT&T v. F.C.C., 220 F.3d 607, 612 (D.C.Cir. 2000). One of the things the ILEC must demonstrate in its application is that it provides interconnection in accordance with the requirements of 47 U.S.C. §§ 251 & 252. Ibid. When the FCC reviews such applications, it does so in consultation with the relevant state's attorney general and state utility commission. Ibid. It does not conduct de novo review of state pricing determinations or adjust rates to conform with TELRIC, even if "isolated factual findings by a commission might be different from what we [the FCC] might have found if we were arbitrating the matter under section 252(e)(5)." Id. at 615 (internal quotation marks and citation omitted). Instead, it rejects an application "only if basic TELRIC principles are violated or the state commission makes clear errors in factual findings on matters so substantial that the end result falls outside the range that the reasonable application of TELRIC principles would produce." Id. at 616 (internal quotation marks and citation omitted).
Line Mix
Turning first to the line mix issue, I find that DPUC's decision to reject plaintiff's proposed 50/50 mix of replacement to growth lines and instead order plaintiff to use a mix assuming 85.1% replacement lines and 14.9% growth lines was not arbitrary and capricious. DPUC considered but quite reasonably declined to rely on plaintiff's estimate, which was based on old data and reflected assumptions that were simply not factually borne out: according to DPUC's findings, the AT&T network was overbuilt and as a result facing significant negative growth in the future. A 50/50 line mix would have plaintiff purchase far more growth lines than made sense for a network experiencing negative growth. Having rejected the premise of plaintiff's 50/50 proposal, it was not arbitrary and capricious for DPUC to turn to more objective methodology used by the Wireline Competition Bureau of the FCC in the Virginia Arbitration Order and adopted by at least ten other states' utility commissions to determine the appropriate line mix. See Doc. #29-1 at 26. DPUC reasonably used plaintiff's own evidence with regard to the other inputs in the line mix algorithm, and assumed a positive 3% growth rate—a fairly generous assumption for plaintiff, considering the negative growth plaintiff had been and anticipated experiencing.
Plaintiff's argument that DPUC should have relied on data from 1999 vendor contract negotiations for the 2000 through 2004 time period is without merit. The cost studies, testimony, and evidentiary hearing before DPUC all took place in 2009 before a 2010 decision. The data on which plaintiff seeks to rely was a decade old when plaintiff presented it to DPUC—not an appropriately forward-looking consideration for a TELRIC cost study. Doc. #29-1 at 28; see also 47 U.S.C. §252(d)(1)(A)(i); 47 C.F.R. §51.505(d)-(e). Even for that time period, the data was of questionable value, given that plaintiff's vendor contracts did not require it to purchase a certain number of growth lines during any time period, and plaintiff's own purchases did not reflect the mix supposedly negotiated. See Doc. #29-1 at 25 & n.125. Plaintiff's own witness stated in 2009 that the "current environment for switching is vastly different than the 2000-2004 time period." Doc. #37-5 at 2. And most importantly, the usefulness of the older data was severely undermined by plaintiff's more recent data from late 2008, which demonstrated that the then-current and forecasted growth of digital switches lines and trunks in Connecticut was negative. Id. at 2-3. It was not remotely arbitrary and capricious for DPUC to decline to rely on older, speculative data that assumed significant positive growth, when actual purchase data and more recent forecasts reflected negative growth.
Plaintiff's economic argument—that DPUC impermissibly neglected to account for any increases in vendor prices that would result in the event that a new line mix was imposed—is also unavailing. DPUC considered and rejected that very argument in its decision:
[T]he Department disagrees with AT&T's claims that any change to its proposed replacement/growth mix would invalidate its contracts or ignore the tradeoff that occurs between replacement and growth line prices. Since AT&T has not purchased the mix of facilities it advocates in its cost study and does not intend to purchase lines in the foreseeable future, it is entirely unclear how adopting [its] proposed mix would impact its relationship with its vendors. Specifically, while [its] proposed mix would raise switching costs and possibly the Recip Comp and TTS rates and revenues, there is nothing in the record that suggests the vendors would benefit from AT&T's increased revenues or how vendors would be harmed by lower revenues. Moreover, higher Recip Comp and TTS rates could possibly cause demand for such services to decrease, which may lower the use of switching facilities and, as a result, further depress AT&T's purchases of growth facilities. Under this scenario, AT&T's vendors would actually be better served if the Telco lowered its Recip Comp and TTS rates and boosted demand for switching services and facilities. Accordingly, the Department finds AT&T's claims that the use of any line mix other than the one it proposed would violate the terms of its contracts is unsupported, at odds with historic and projected purchasing patterns and internally inconsistent.Doc. #29-1 at 25. See also Verizon Pa., Inc. v. Pa. Pub. Util. Comm'n, 380 F. Supp. 2d 627, 642 (E.D.Pa. 2005) (rejecting an identical argument as "irrelevant to the court's analysis" because "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").
Plaintiff's economic argument about vendors is too simplistic. For one thing, it assumes that Connecticut is the only geographic market in which an ILEC and vendors might operate: if, for example, another state's utility commission set a line mix that was heavily or overly weighted toward pricier growth lines rather than less expensive replacement lines, both the ILEC and the vendors with which it contracted could be just fine. For another, plaintiff's argument assumes that vendors' primary goals must always revolve around maximizing profit through higher prices charged to their existing customers, rather than acquiring new market share. Nor does plaintiff provide empirical support for its argument about vendors' price-setting behavior.
DPUC's decision with regard to the line mix was supported by substantial record evidence, in the form of plaintiff's interrogatory responses indicating negative growth, plaintiff's own data regarding the cost of capital and depreciation inputs into the methodology explained in the Virginia Arbitration Order, and the Virginia Arbitration Order itself, of which DPUC took administrative notice on the record. Doc. #29-1 at 25 n.129 & 26; Doc. #37-5. I therefore find for DPUC on Count Five of the complaint.
Fill Factor
I also find that DPUC's decision was not arbitrary and capricious with regard to its analysis of the fill factor input. The purpose of the fill factor is to account for the fact that carriers must build excess capacity into their networks, so that capacity will be available for additional customers or other demands. TELRIC contains no specific algorithm for determining the fill factor, and the FCC has approved several. See Illinois Bell, 349 F.3d at 405. Fill factors are "'estimates of the proportion of a facility that will be filled with network usage.'" TDS Metrocom, 387 F. Supp. 2d at 943 (some internal quotation marks omitted) (quoting In re Implementation of the Local Competition Provisions in the Telecommunications Act of 1996, 1996 WL 452885, 11 F.C.C.R. 15499, ¶ 682 (1996)). Logically, this concept extends not only to how many trunks are in use (i.e., the trunk utilization figure plaintiff input in its model as the fill factor), but also to how long they are used (i.e., the CCS assumption).
Substantial evidence in the record supports DPUC's decision to ask plaintiff to include the CCS assumption in its calculation of the 72.5% fill factor, rather than elsewhere in the model. The 72.5% fill factor plaintiff wanted to use as an input was itself the result of a calculation of the midpoint between 65% and 85%, which were the endpoints of a range provided in the prefiled testimony of DPUC's consulting expert Hamiter. See Doc. #56-1 at 16-17. Hamiter's range was determined in order to capture the range of healthy network usage for "aggregate trunk utilization percentage," which he specifically distinguished from "trunk-fill." Id. at 18. Plaintiff's expert Mollett stated, however, that he had been using the terms interchangeably. Id. at 19. His conflation of the two resulted in plaintiff importing the 72.5% (which represented a healthy trunk utilization rate) into the individual fill factor input.
That individual fill factor input, it became clear through the course of the testimony, should properly also have taken into account the CCS assumption, which represents average annual usage in terms of minutes of use during the 328 busy-hour days in a year and plays an important role in the fill of a trunk. Id. at 23-26. In fact, Mollett even conceded three times that one could incorporate the CCS assumption information into a fill factor, though he had not done it that way Id. at 26-27. Mollett further agreed that the fill factor and CCS assumption in the model interacted such that the two numbers were ultimately multiplied together by the model, resulting in an "effective fill" of a lower number than either the original fill factor or CCS percentages. Id. at 28-29.
Mollett also specifically agreed that if the CCS figure were lowered, the trunk would be less utilized. Doc. #56-1 at 27. This, of course, lies at the very heart of what the fill factor is supposed to represent—the "proportion of a facility that will be filled with network usage." TDS Metrocom, 387 F. Supp. 2d at 943 (internal quotation marks and citation omitted).
--------
On the basis of this evidence, DPUC found that plaintiff's model was not actually including a fill factor of the healthy 72.5% it meant to incorporate, but that the fill was actually much lower as a result of the impact of the CCS assumption on it. It was not at all arbitrary and capricious for DPUC to ask plaintiff to incorporate the CCS figure in its fill factor input, so that the "effective fill" would land at the 72.5% that plaintiff and Hamiter agreed would represent healthy network usage. Nor was it arbitrary and capricious for DPUC to suggest that one simple way for plaintiff's model to do so would be to simply replace the 72.5% fill factor input in plaintiff's model with 95% and leave the CCS values as is. See Doc. #29-1 at 30 n.166. DPUC did not require plaintiff to run its calculation that way, as long as it re-ran its cost studies with a fill factor that took the CCS assumption into account and was nevertheless no higher than 72.5%. Id. at 30.
Perhaps plaintiff could have otherwise manipulated its fill factor input to incorporate the CCS value and then removed the CCS assumption from its location elsewhere in the model, to the same end. Plaintiff concedes that the CCS assumption is a permissible consideration in determining a TELRIC rate. At no point did DPUC suggest that the CCS figure be double-counted by including it both in the fill factor input and in its original location in the model. Because DPUC was not arbitrary and capricious when it asked plaintiff to ensure that the fill factor took into account the amount of usage by time as well as by number of trunks, I find for DPUC on Count Four of the complaint.
Harmless Error
Finally, even if plaintiff is correct that DPUC made an error with regard to either its line-mix or fill-factor findings, plaintiff has not shown that it suffered any harm to the final TELRIC rate as a result of such error. Although the Second Circuit has not addressed the issue, the persuasive law of other federal circuits indicates that I should affirm DPUC's decision unless the final rate required by DPUC was "unreasonable or unsupported by substantial evidence." See MPower, 457 F.3d at 629; see also Qwest Corp., 589 F.3d at 994-95; AT&T, 220 F.3d at 617-18. That is, the Court need not overturn DPUC's decision even if there were technical errors in it, because the record does not show that DPUC "made any error so large that it drew the bottom line out of whack." MPower, 457 F.3d at 632. After all, "TELRIC requires that the rate reflect the costs of efficient production, not that each ingredient of the formula do so independently." Illinois Bell, 349 F.3d at 411 (emphasis in original) (rejecting application of a state law that would require certain factors to be used in isolation). As Judge Easterbrook has explained,
Congress provided for federal judicial review of rates set by state commissions; it did not provide for review of individual factors that influence those rates. A lower fill factor, which elevates the rate, may be offset by other factors that depress it. As long as the final rate comports with TELRIC, why should it matter what role particular intermediate factors played? Any effort to analyze a factor in isolation poses a distinct risk of generating an advisory opinion, as well as a certainty of complicating review of the rate ultimately announced. A different way to put this is that review of agency action usually is limited to the agency's final decision, and the choice of one or two legal criteria that the agency will use along the way cannot be called a ''final'' decision.Id. at 408-409 (emphasis in original). In that case, Judge Easterbrook also noted that "[t]he district court's analysis may have been affected by the parties' choice to present for decision a challenge to two factors, standing alone, rather than a challenge to a promulgated rate. Both of these factors look to the present or the past; if they were the only factors, then the problem would be clear; but under TELRIC they can't be the only factors, and their propriety should not have been evaluated in isolation from the other components of a TELRIC rate." Id. at 411 (emphasis in original).
The present case similarly provides a challenge to two factors, standing alone, without consideration of the other dozen-plus factors that jointly determined the final TELRIC rates. In fact, while discussing another input factor, DPUC noted its concern with the high transit rate proposed by plaintiff and emphasized that it was not inclined to approve plaintiff's proposed cost inputs without better information than plaintiff had provided to understand the reasonableness of those costs:
When comparing [plaintiff's] proposed Connecticut [transit] rate to transit rates in
other states, [plaintiff's transit] rate significantly exceeds those in all the other states. The Connecticut rate is almost double that of the second highest AT&T [transit] rate on the record in Nevada. The Department will not approve a cost element that has such a dramatic impact on AT&T's [transit] (CLEC) rate based on information that is not transparent and cannot be audited to determine its reasonableness.Doc. #29-1 at 20.
It is clear that DPUC was mindful of the final TELRIC rates as it was evaluating each individual factor in the analysis, and it would be unwise for this Court to speculate how analysis of the unchallenged factors may have interacted with or offset analysis of the challenged factors in the determination of suitable TELRIC rates. In other words, even if plaintiff is correct that DPUC sub-optimally evaluated the line mix or fill factor inputs, plaintiff has not shown that the final TELRIC interconnection rates for transit services fell outside the range of reasonable TELRIC rates. I therefore cannot find that DPUC's decision with regard to either of the two challenged factors—much less its entire determination—was arbitrary and capricious.
CONCLUSION
For the foregoing reasons, the Court AFFIRMS the decision of the DPUC with regard to the challenges presented in Counts Four and Five of plaintiff's complaint. The Clerk shall close this case.
It is so ordered.
Dated at Bridgeport this 31st day of March 2015.
/s/ Jeffrey Alker Meyer
Jeffrey Alker Meyer
United States District Judge