Opinion
No. 20-5304
02-09-2021
HAMILTON COUNTY EMERGENCY COMMUNICATIONS DISTRICT, et al., Plaintiffs-Appellants, v. LEVEL 3 COMMUNICATIONS, LLC, Defendant-Appellee.
NOT RECOMMENDED FOR PUBLICATION
File Name: 21a0084n.06 ON APPEAL FROM THE THE UNITED STATES DISTRICT COURT FOR THE EASTERN DISTRICT OF TENNESSEE
OPINION
BEFORE: DAUGHTREY, NALBANDIAN, and MURPHY, Circuit Judges.
NALBANDIAN, Circuit Judge. Like every state, Tennessee uses 911 as the phone number people call when they need emergency help. And for a long time in the Volunteer State, the municipal corporations that ran 911 call centers got their funding from charges that certain telephone companies billed certain customers. The plaintiffs here, several Emergency Communications Districts, are some of those corporations. They sued Level 3, a telecommunications company that does business in Tennessee. They claim Level 3 withheld millions of dollars in fees that the state obligated Level 3 to charge its customers and then send to the Districts.
The district court didn't buy the Districts' theory. It viewed their experts as unreliable and found that they otherwise presented no evidence that Level 3 underbilled 911 charges. So the court granted Level 3 summary judgment. We spot no errors in the district court's judgment, so we AFFIRM.
I.
A.
This diversity case is factually complex. It has its genesis in 1984, when Tennessee passed the Emergency Communications District Law (ECD Law). Tenn. Code Ann. §§ 7-86-101-151. The ECD Law formally established 911 as the main emergency telephone number for Tennessee residents. And the law also established emergency communications districts—the plaintiffs here. These Districts are municipal corporations that operate 911 call centers in various counties across Tennessee. They receive and then route emergency calls.
In 2014, Tennessee enacted an overhaul of the ECD Law. See 2014 Tenn. Pub. Acts 795. "This Act repealed and replaced the 911 law at issue here, such that many of the statutory provisions cited herein have been deleted or relocated within the Tennessee Code." Hamilton Cnty. Emergency Commc'ns Dist. v. BellSouth Telecomms. LLC, 852 F.3d 521, 524-25 n.1 (6th Cir. 2017). For this litigation, the pre-2014 regime is the relevant law, so all citations to the Tennessee Code refer to that regime.
Of course, the Districts need money to operate. So the ECD Law authorized the Districts to impose a charge on those telephone users' phone lines that can call 911. Hamilton Cnty. Emergency Commc'ns Dist. v. BellSouth Telecomms., LLC, 852 F.3d 521, 525 (6th Cir. 2017). Toward that end, the ECD Law required telephone "service suppliers" to bill and collect the 911 charge from their "service users." Tenn. Code Ann. §§ 7-86-108, 7-86-110. By way of background, a service supplier is "any person, corporation or entity providing exchange telephone service to any service user." Id. § 7-86-103(15). And a service user is "any person, corporation or entity that is provided 911 service." Id. § 7-86-103(16). So a provider of exchange telephone service would charge its customers who received 911-capable phone service an extra amount on top of its normal charges. And after charging the user, the supplier would then send the money to the Districts. Id. § 7-86-110(a).
How does Level 3 fit within this scheme? Level 3 has been around since 1998 as a wholesale provider of Internet Protocol (IP) data, transport, and communications services. (R. 117-19, Robles Decl. ¶ 3, PageID # 3298.) Level 3 provides these services to telecoms carriers and internet service providers. And it does so over an internet fiber network it built around two decades ago.
But Level 3 differs from traditional telephone service. It uses what's called "packet-switched technology," which breaks down voice or data into digital packages that are reassembled at their destination using the same technology that sends data over the internet. (R. 117-19, Robles Decl. at ¶ 4, PageID # 3299.) Contrast this with traditional exchange telephone service, "which provides a dedicated transmission path or 'circuit' for the duration of the call." (Id.) Traditional exchange service uses different facilities and equipment than IP-enabled services, so Level 3 cannot provide traditional exchange phone service.
After building its IP fiber network, Level 3 began to provide wholesale Internet access and other data IP-based services, then later started to offer Voice over Internet Protocol (VoIP). Still, before 2008, Level 3 was known as "a carrier's carrier" because it only provided services to other telecoms, VoIP, and internet service providers. (Id. at ¶ 8, PageID # 3300.) These companies used Level 3's services to support their own network facilities or combine them with their own services to create a finished product that they then resold, either to other providers or to end users. (Id.) Level 3 shifted from this exclusively wholesale model in 2008, when it began to offer "enterprise services" to businesses. Still, even now, most of Level 3's services cannot call 911. And it offers most of its voice-capable services to facilities-based wholesale providers, who configure or combine Level 3's services and sell them either to other telecoms companies or to end users.
B.
This case arises out of the billing scheme the ECD Law created. Over a dozen Districts sued Level 3 in federal court. They claimed Level 3 "intentionally failed to fulfill its obligation to bill, collect, report, and remit to the Districts" the 911 charge. (R. 18, First Am. Compl. ¶ 5, PageID # 174.) And they alleged several causes of action. First, they said Level 3 violated the Tennessee False Claims Act by filing false remittance reports with the Districts understating the number of lines Level 3 needed to charge. They also sued under an implied cause of action in the ECD Law itself, seeking compensation for the fees Level 3 should have but didn't charge and remit. See BellSouth, 852 F.3d at 531 (holding that a private right of action exists under the ECD Law). And along with these statutory claims, the Districts also brought some state-law claims, including breach of fiduciary duty, fraudulent misrepresentation, fraudulent concealment, and negligent misrepresentation.
Where did the Districts come up with their theory of Level 3's serial underbilling? A brief detour can help explain. The Districts first suspected Level 3 of underbilling when they compared WARs—Wireline Activity Reports—that Level 3 submitted to the state against the remittance reports Level 3 filed with the Districts. These remittance reports detailed the number of 911 charges that Level 3 billed its customers and sent to the Districts. And the WARs are voluntary reports that the Tennessee Regulatory Authority asks telecoms companies like Level 3 to send it so that it can monitor competition in the telecoms market.
The Districts' theory is that the WARs accurately represent the number of 911-billable customers a telecoms company like Level 3 has. So, in theory, the number of reported lines in a WAR should equal the number of charges that a telecoms company like Level 3 sends to the Districts. But that wasn't the case here. Instead, Level 3 reported far more lines in its WARs than it did in its remittance reports to the Districts. So, according to the Districts, if a telecoms company like Level 3 is following the law, then A (the number of lines Level 3 reported in its WARs) minus B (the number of lines Level 3 charged in its remittance reports) should be zero. But here, A minus B spits out a positive number. That means the number of lines Level 3 reported in its WARs is higher than the number of lines it charged the 911 fee. Thus, in the Districts' eyes, Level 3 underbilled—the WARs show that Level 3 serves many more lines than it billed 911 fees.
Using this theory, the Districts sued Level 3. After the Districts filed a first amended complaint, Level 3 moved to dismiss it. The district court obliged, but only in part. The court allowed the claims under the False Claims Act and ECD Law to proceed. But it dismissed all the state common law claims because "[t]he Districts' causes of action under the 911 Law are exclusive of any common-law actions to seek vindication of their rights under the 911 Law." The Districts have not appealed the claims that the trial court dismissed initially.
So all that survived Level 3's motion to dismiss were the Districts' two statutory claims. And extensive discovery followed the partial dismissal. It included interrogatories, depositions, and expert disclosures and declarations. For their part, the Districts offered two expert witnesses. The first, Joseph Gillan, opined that the WARs are an accurate representation of the number of 911-billable lines that Level 3 serves. The second, Randall Hebert, opined about damages that the Districts suffered because of Level 3's underbilling.
Level 3 moved to exclude both experts. It sought to exclude Gillan on two main grounds. First, Gillan's opinion was premised on, and included, testimony that Level 3 is in fact a "service supplier" under the ECD Law. But because Gillan reached this conclusion through statutory interpretation—and he's a telecoms expert rather than a lawyer—Level 3 argued it was an inappropriate legal opinion. Second, Level 3 argued that Gillan's opinion—that WARs accurately approximate the number of lines a service supplier should charge 911 fees—is not reliable because it was not based on enough investigation, and other, undisputed record evidence contradicts it.
This second basis for excluding Gillan also did much of the work in Level 3's effort to exclude Hebert. After all, Hebert's calculation of damages presupposed that the WARs reflected the number of 911 charges Level 3 should have billed. So because this assumption was ill-founded, Hebert's testimony was unreliable, as well.
Before the district court ruled on Level 3's motion to exclude, Level 3 moved for summary judgment. It argued mainly that the ECD Law didn't require it to remit charges to the Districts because it isn't a "service supplier." This motion remained pending while the district court decided Level 3's motion to exclude. But after a Daubert hearing, the district court granted Level 3's motion to exclude Gillan's opinion on the WARs and Hebert's opinion on damages. See Daubert v. Merrell Dow Pharms., Inc., 509 U.S. 579 (1993). As to Gillan, the court reasoned that "there is no evidence to support Mr. Gillan's opinion that the Wireline Activity Reports accurately show the number of voice lines [Level 3] had in service that could make voice calls to 911 in the relevant months, and . . . the available evidence shows the opposite." And Hebert's opinions "are entirely dependent on the proposition that the Wireline Activity Reports are a valid source of data on how many of [Level 3's] lines were capable of making calls to 911." So without Gillan's testimony coming in, "there is currently no evidentiary support for Mr. Hebert's assumption that the Wireline Activity Reports accurately count the number of lines on which 911 charges should have been made."
After the district court excluded the opinions of Hebert and Gillan, it ordered the Districts to brief whether a genuine dispute of material fact remained. Both parties briefed the issue. And just a couple of days later, the Districts moved to exclude the testimony of Michael Robles, Level 3's corporate representative. They argued that his declaration contained expert testimony, but Level 3 never disclosed him as an expert.
Following this briefing and motion, the district court granted Level 3's motion for summary judgment. The court noted that after it excluded Gillan and Hebert's expert testimony, the only evidence the Districts had of liability was Level 3's response to an interrogatory along with certain portions of a declaration and deposition by Michael Robles. But neither piece of evidence showed Level 3 failed to bill lines the ECD Law required it to bill. And other undisputed evidence in the record established Level 3's entitlement to summary judgment. For instance, undisputed deposition testimony showed that Level 3 included all sorts of lines in its WARs that the ECD Law didn't require it to charge a 911 fee. And no evidence in the record suggested that WARs "are an accurate basis from which to determine how many 911 charges Level 3 should have collected." (R. 139, Mem. at 17, PageID # 6131.) Finding no evidence of Level 3's liability, the court granted Level 3's motion for summary judgment and then denied as moot the Districts' pending motion to strike the Robles declaration.
That takes us to this appeal. The Districts ask us to review several of the district court's decisions. They start with the larger claim that the district court should not have granted summary judgment. As part of this claim, the Districts say the district court should have admitted the WARs as secondary evidence of the lines Level 3 should have been charging the 911 fee. And they say the district court did not make the correct inferences from existing facts before it granted summary judgment. Next, the Districts say the district court erred when it held that the Districts could not rely on Level 3's interrogatory responses to establish liability. And the Districts also charge the district court with error for excluding their expert witnesses but not Level 3's corporate representative, Michael Robles.
We find none of these arguments persuasive. So we affirm the district court.
II.
The Districts' various legal challenges call for different standards of review. Start with summary judgment, which is governed by a familiar standard. We review a grant of summary judgment de novo. Griffith v. Franklin County, 975 F.3d 554, 566 (6th Cir. 2020). Summary judgment is appropriate when "no genuine dispute as to any material fact" exists, and the moving party "is entitled to judgment as a matter of law." Fed. R. Civ. P. 56(a).
"A genuine dispute of material fact exists if the evidence is such that a reasonable jury could return a verdict for the nonmoving party." Peffer v. Stephens, 880 F.3d 256, 262 (6th Cir. 2018) (quotations omitted). We review the evidence in favor of the nonmoving party and draw reasonable inferences in its favor. Burgess v. Fischer, 735 F.3d 462, 471 (6th Cir. 2013). But "we need not make assumptions that strain credulity." Gecewicz v. Henry Ford Macomb Hosp. Corp., 683 F.3d 316, 323 (6th Cir. 2012). So "[w]hen opposing parties tell two different stories, one of which is blatantly contradicted by the record, so that no reasonable jury could believe it, a court should not adopt that version of the facts for purposes of ruling on a motion for summary judgment." Scott v. Harris, 550 U.S. 372, 380 (2007).
Next, the evidentiary rulings. This includes the refusal to strike the Robles Declaration, the decision to strike the Districts' two experts, and the refusal to admit the WARs under Rule 1004. We review the district court's decisions under the highly deferential abuse-of-discretion standard. See, e.g., Baker Hughes Inc. v. S&S Chem., LLC, 836 F.3d 554, 560 (6th Cir. 2016); United States v. Jones, 107 F.3d 1147, 1150-51 (6th Cir. 1997). That's true even when excluding expert testimony leads to summary judgment, as it did here. Pluck v. BP Oil Pipeline Co., 640 F.3d 671, 676 (6th Cir. 2011). We avoid substituting our judgment for the district court's, and we reverse only if we are "left with a definite and firm conviction that the district court committed a clear error of judgment." Id. (alterations and quotations omitted).
The Districts urge us to review the district court's Daubert ruling de novo. But de novo review "is appropriate only where the district court excludes expert testimony on the basis of an incomplete record, or after conducting nothing more than a perfunctory review." Ky. Speedway, LLC v. Nat'l Ass'n of Stock Car Auto Racing, Inc., 588 F.3d 908, 915 (6th Cir. 2009). Here, the district court held a Daubert hearing, the parties briefed the issue, and the district court issued a memorandum detailing its reasons for excluding the Districts' witnesses. So de novo review isn't appropriate. Cf. Jahn v. Equine Servs., PSC, 233 F.3d 382, 393 (6th Cir. 2000) (reversing under a de novo standard of review where the district court "prematurely" excluded experts' reports without conducting a Daubert hearing or allowing the parties to submit briefs).
A.
1.
We start by reviewing the evidence the district court excluded. The Districts claim that the district court erred in excluding the expert opinions of Randall Hebert and Joseph Gillan. Gillan's expert testimony was that lines reported in the WARs are lines able to call 911 and thus lines subject to the 911 charge. In other words, Gillan would testify about Level 3's liability. And Hebert's expert testimony was on damages. He assumed the WARs accurately represented the number of lines Level 3 ought to have billed 911, then compared that number of lines to the data in Level 3's remittance reports.
The district court excluded the testimony of both witnesses (but only part of Gillan's opinion). In excluding the two, the court reasoned that none of the available data showed that the WARs accurately reflect the number of lines for which Level 3 should have charged 911 fees. So Gillan's expert opinion was not based on sufficient facts or data, and there was too great an analytical gap between the data and Gillan's opinion. And because Hebert's testimony depended on assumptions from Gillan's testimony, he was unreliable, as well.
Before continuing, an overview of the WARs is helpful. WARs—Wireline Activity Reports—are reports that telecoms companies voluntarily file with the Tennessee Regulatory Authority. The Authority asks companies to file these reports so that it can monitor competition in local telecoms markets. But the Authority asks companies to report their "local communications" services. And this embraces several services—"more than just the retail local exchange voice services that would be subject to 911 charges." (R. 92-6, Expert Report of Eddie Roberson, PageID # 1863.) In other words, WARs ask telecoms companies to include in their reports more services than just those able to call 911. That's because the state "wanted to know not just what services were actually in use, but instead what physical capacity the CLECs were installing within the State that had the potential to provide telecom service." (Id. at PageID # 1864.) So the WARs instruct the filing companies to include digital and analog, data and voice, single-line and multi-line, switched and non-switched, and local and long-distance services.
a.
With that in mind, we turn to the expert opinions. Because the admissibility of Gillan's opinion determines the same for Hebert, we address Gillan first. Under the Federal Rules of Evidence, "the trial judge must ensure that any and all scientific testimony or evidence admitted is not only relevant, but reliable." Daubert, 509 U.S. at 589; see Kumho Tire Co., Ltd. v. Carmichael, 526 U.S. 137, 147-48 (1999) (extending Daubert to all forms of expert knowledge, not just scientific knowledge). This has its roots in Rule 702. The Rule allows a witness "qualified as an expert by knowledge, skill, experience, training, or education" to testify in the form of an opinion if four conditions are met. Fed. R. Evid. 702. First, "the expert's scientific, technical, or other specialized knowledge will help the trier of fact to understand the evidence or to determine a fact in issue." Id. Second, "the testimony is based on sufficient facts or data." Id. Third, "the testimony is the product of reliable principles and methods." Id. And finally, "the expert has reliably applied the principles and methods to the facts of the case." Id.
Daubert articulated a few non-exhaustive factors that a district court considers in deciding whether to admit expert testimony. 509 U.S. at 593-94. These include whether the theory can be tested; whether it has been subject to peer review and publication; what the technique's error rate is; and whether the theory or technique has been generally accepted by the relevant expert community. Id. Also relevant is whether an analytical gap exists between the data the expert relies on and the conclusion he draws. Gen. Elec. Co. v. Joiner, 522 U.S. 136, 146-47 (1997).
Here, several of the Daubert factors do not apply. Testing, peer review, and error rates are not appropriate factors to evaluate Gillan's opinion that the WARs accurately capture 911-billable lines. But there are still a couple of factors to consider in asking if the district court abused its discretion by excluding Gillan's expert testimony. First, we can look at whether there is any evidence that the telecoms industry has generally accepted Gillan's theory about Level 3's WARs. See Daubert, 509 U.S. at 594 ("[A] known technique which has been able to attract only minimal support within the community may properly be viewed with skepticism." (quotations and citation omitted)). And second, we can look at whether there is any factual basis to support the conclusion Gillan reaches. Joiner, 522 U.S. at 146-47 ("[B]ecause it was within the District Court's discretion to conclude that the studies upon which the experts relied were not sufficient, whether individually or in combination, to support their conclusions that Joiner's exposure to PCB's contributed to his cancer, the District Court did not abuse its discretion in excluding their testimony.").
It turns out that Gillan's opinion is not based on sufficient facts or data. Indeed, it is based on no facts or data, at least none in the record. There is no connection between the WARs and the number of 911 charges that Level 3 should have been charging. And the Districts have presented no evidence to rebut Level 3's witnesses who testified that the WARs have no relation to 911 charges.
Level 3 presented expert testimony from Eddie Roberson, who testified that "Wireline Activity Reports had nothing whatsoever to do with 911 charges." (R. 92-6, Expert Report of Eddie Roberson, PageID # 1860.) And Tennessee used the WARs to ask "providers to report facilities even if they were carrying services that are not subject to 911 charges." (Id. at PageID # 1861.) So the WARs Level 3 submitted "may not be used to provide an accurate or reliable estimate of the number of 911 charges providers should have remitted to the local districts." (Id.) That was because the WARs asked for "information on more than just the retail local exchange voice services that would be subject to 911 charges." (Id. at PageID # 1863.) In fact, Roberson had never heard of anyone using WARs to estimate the 911 charges a company should have been collecting. So the WARs "are not a reasonable proxy, nor even a reasonable starting point, to estimate the number of 911 charges that should have been collected or remitted by any provider." (Id.) And Gillan himself acknowledged that he had never heard of any court or regulator using WAR lines as a proxy for 911-billable lines.
Level 3 also presented evidence that it included in its WARs services that are not subject to the 911 charge. In its interrogatory responses, Level 3 noted that it "include[d] numerous types of lines and products that were not subject to the 911 taxes imposed by Plaintiffs' Districts." (R. 117-20, Level 3 Suppl. Resp. to Plaintiffs' First Set of Common Interrogatories, PageID # 3322.) This included "wireless lines, IP-enabled services, federal government lines, lines in excess of the 100-charge cap, lines provided on a wholesale basis to resellers, data-only, incoming-only or outgoing-only lines, lines not capable of connecting to the 911 PSAP, etc." (Id.; see id. at PageID # 3331-33.) And finally, Level 3 also presented evidence that it included in its WARs services that its subsidiaries provided. And no one claims Level 3 would have to charge and remit the 911 fee on those services.
Given all this, why would Gillan think that the WARs accurately represent the number of 911-billable lines that Level 3 provides? Level 3's counsel asked him at his deposition if it is his opinion that "it's appropriate to use the [WARs] as an indication of the 911 charges that should have been billed simply because the [WARs] are the best data that have been made available" to him—and he said yes, even acknowledging that it is possible the WARs might include lines that cannot call 911. (R. 92-4, Gillan Dep. 156:6-21, PageID # 1631.) Gillan also admitted in his deposition that he was not aware of the types of services Level 3 provides, nor did he review the billing data that Level 3 provided during discovery. And neither did he try to determine whether Level 3's WARs included information from subsidiaries.
Even when counsel asked Gillan at his deposition if anything in the WAR instructions directed carriers "to only report switched access lines that are capable of carrying an outbound voice call," Gillan responded "no." (Id. 53:4-11, PageID # 1618.) He also said that "[t]here's no perfect equivalency" between the types of lines the WARs ask telecoms companies to include and outbound voice channels that can call 911. (Id.) And he acknowledged that it is possible a carrier could correctly bill no 911 charges yet still have lines to report in its WARs.
The Districts presented no evidence to contradict Level 3's evidence. They point to only one piece of record evidence in their attempt to link the WARs to 911 charges. They claim Level 3 admitted that the WARs captured 911-billable lines in its response to one of the Districts' interrogatories. In Interrogatory 4, the Districts asked Level 3 to identify the policy it used to prepare WARs. Level 3 responded that it included all types of lines and products in its WARs that were not subject to any 911 charge. But it also said that it generated its WARs "based upon telephone numbers" before December 2011 and that it switched to using a line-count table that "included all types of lines, not just local exchange access lines" in December 2011. (R. 117-20, Level 3 Suppl. Resp. to Plaintiffs' First Set of Common Interrogatories, PageID # 3322.) The Districts seize on this single line as suggesting that before December 2011, when Level 3 first started using that table as the source of its WAR data, it only included "local exchange access lines" in its WARs, and those lines are subject to the 911 charge. So, the Districts contend, the pre-2011 WARs are an accurate count of the lines Level 3 should have been billing the 911 charge.
But the district court rightly pointed out that this interrogatory response says nothing about the types of lines Level 3 reported in its WARs before December 2011. And the inference that the Districts draw makes no sense given the entire interrogatory response. See Gecewicz, 683 F.3d at 323 ("Though we must draw all reasonable inferences in favor of Gecewicz, the nonmoving party, we need not make assumptions that strain credulity."). That's because Level 3 says throughout its response that the information it provided in the WARs "include[s] numerous types of lines and products that were not subject to the 911 taxes imposed by" the Districts. (R. 117-20, Level 3 Suppl. Resp. to Plaintiffs' First Set of Common Interrogatories, PageID # 3322.) And "lines and products . . . not subject to the 911 taxes" is just another way to refer to things other than "local exchange access lines" because the latter is subject to the 911 charge. In other words, Level 3's reference to "not just local exchange access lines" repeats what it already said: that its WARs "include numerous types of lines and products that were not subject to the 911 taxes imposed by" the Districts. And it is unreasonable for the Districts to infer an admission that prior to December 2011, Level 3 was only including "local exchange access lines" in its WARs given that Level 3 repeatedly noted throughout its interrogatory responses that it includes in its WARs services not subject to 911 charges.
There is another problem with Gillan's expert opinion. His entire opinion is premised on an assumption that Level 3 is a "service supplier" subject to the 911 fee. But we have serious doubts that it is. Recall that the Districts received their funding from fees that "service suppliers" charged "service users." Tenn. Code Ann. §§ 7-86-103, 108, 110. And under the ECD Law, a service supplier is "any person, corporation or entity providing exchange telephone service to any service user," Id. § 7-86-103(14), while a service user is "any person, corporation or entity that is provided 911 service," Id. § 7-86-103(15). Once a service supplier charged the user the 911 fee, it sent the fee to the Districts. Id. § 7-86-110(a).
The ECD Law did not define "exchange telephone service." But the Tennessee legislature defined "basic local exchange telephone services" in another statute. See In re Kaliyah S., 455 S.W.3d 533, 552 (Tenn. 2015) ("Statutes that relate to the same subject matter or have a common purpose must be read in pari materia so as to give the intended effect to both."). In that statute, basic local exchange telephone service means:
[T]elecommunications services which are comprised of an access line, dial tone, touch-tone and usage provided to the premises for the provision of two-way switched voice or data transmission over voice grade facilities of residential customers or business customers within a local calling area, Lifeline, Link-Up Tennessee, 911 Emergency Services and educational discounts existing on June 6, 1995, or other services required by state or federal statute.Tenn. Code Ann. § 65-5-108(a)(1). And Tennessee regulations mirror this definition. One defines basic local exchange service as "telecommunications services comprised of an access line, dial tone, touch tone, and usage provided to the premises of residential customers or business customers for the provision of high quality, two-way switched voice, or data transmission over voice grade facilities." Tenn. Comp. Rules & Regs. § 1220-04-08-.01(1)(c).
So a service supplier is the entity that provides an "access line, dial tone, touch tone, and usage" to the user, who can then use the service to call 911. Tenn. Code Ann. § 65-5-108(a)(1); Tenn. Comp. Rules & Regs. § 1220-04-08-.01(1)(c). And the language of the ECD Law suggests there is only a single service supplier for each service user. Tenn. Code Ann. § 7-86-110(a) states: "The service supplier shall remit the funds collected as the service charge to the district every two (2) months." Similarly, the 911 charge "may be stated separately in the billing by the service supplier. . . . The service charge shall be collected at regular billing intervals in according with the regular billing practice of the service supplier." Tenn. Code Ann. § 7-86-108(c). Again, the statute speaks of "[t]he service charge collected by the service supplier." Id.
So the language and structure of the ECD Law, and additional Tennessee authority, give us a fairly complete picture of a "service supplier." It is the single entity that has the direct service and billing relationship with the user, who can then use the provided services to call 911. In other words, the service supplier is the company that directly supplies 911 capability over a line to the service user. See Hamilton Cnty. Emergency Commc'ns Dist. v. BellSouth Telecomms., LLC, 154 F. Supp. 3d 666, 683-85 (E.D. Tenn. 2016), reversed on other grounds, 852 F.3d 521 (6th Cir. 2017).
But nothing in the record suggests this is the type of service Level 3 provides. For one thing, Level 3's 30(b)(6) representative Michael Robles testified that Level 3 never provided local basic exchange services—it provided only Internet Protocol, like VoIP. (Appellant's App. 69A, Robles Dep. 82:2-4 ("What I'm trying to clarify here is we were not providing local basic exchange services. We are providing IP-based services."); Id. 138A, 361:14-20 ("Q. Okay. Did Level 3 Communications, LLC provide a basic exchange local telephone service? A. We did not. Our network was not technically capable of providing that service because we had no circuit switching infrastructure to board.").) And for another, Level 3 sold most of its services wholesale. That means its customers repackaged what Level 3 sold them and then sold it either to other telecoms companies or directly to retail users. But if Level 3 had to charge the 911 fee and remit it to the Districts, even as a wholesaler, many customers would be subject to a double fee—the user's immediate supplier would also have to charge the fee. But as we have said, the text of the ECD Law does not support this reading, and the Districts acknowledged as much at oral argument. See BellSouth, 154 F. Supp. 3d at 684 ("The text of the statute contemplates only one service supplier per service user.").
Even more, the ECD Law did not provide for the type of technology Level 3 offers until it amended the law in 2006 to add non-wireline services—which includes IP-enabled technology—into its fold. It did so in Tenn. Code Ann. § 7-86-108(a)(1)(B)(vi). That provision extended the application of a separate 911 charge, one owed to the Tennessee Emergency Communications Board (TECB), to "all subscribers and users of non-wireline service." And "non-wireline service" includes IP-enabled services. Tenn. Code Ann. § 7-86-103(11). So the Tennessee legislature thought it was necessary to amend the ECD Law in 2006 to include the types of services that Level 3 provides. And even then, it subjected IP-enabled services to a different regulatory regime, one that sends 911 charges to the TECB rather than the Districts. Tenn. Code Ann. §§ 7-86-108(a)(1)(B), 7-86-303(c).
Two telecoms experts corroborate our reading of the ECD Law. For instance, Eddie Roberson says in his expert report that "'exchange telephone service' referred to traditional wireline exchange services." (R. 92-6, Expert Report of Eddie Roberson ¶ 45, PageID # 1867.) So exchange telephone service "did not include IP-enabled services, such as VoIP." (Id.) Of course, it is the latter that Level 3 provides. And Lynn Questell's expert opinion echoes Roberson's. Questell is a former executive director of the TECB. She says that before the 2006 amendment to the ECD Law, "the old 911 law did not include a 911 charge for IP-enabled services (including VoIP services), or for other non-landline, non-CMRS services." (R. 92-5, Expert Report of Lynn Questell, ¶ 16, PageID # 1642 (emphasis added).)
In short, Gillan's testimony suffers multiple fatal defects. So it is not based on sufficient facts or data. Gillan admitted that he did not know what services Level 3 provided or how Level 3 filled out its WARs. And he admitted that he was unaware of any court or regulatory authority that used WARs to gauge 911-billable lines. His belief that Level 3's WARs approximate 911-billable lines appears to depend on the lack of any other information. But one expert's ipse dixit, unsupported by any facts in the record and contradicted by multiple industry experts and even his own deposition testimony, does not satisfy Daubert or Rule 702. The district court did not abuse its discretion in excluding Gillan's expert opinion on the WARs.
b.
And for the same reason, the district court did not abuse its discretion in excluding Hebert's report. That report purported to calculate the Districts' alleged damages. But it did so by comparing two values: the number of reported lines in the WARs and the number of lines for which Level 3 remitted a 911 charge. The difference is the amount that Level 3 allegedly underbilled. (Id.) Then, to calculate damages, Hebert multiplied that difference by the applicable fee.
But this calculation assumes that lines Level 3 reported in WARs are a good proxy for lines Level 3 should have billed 911 charges. Otherwise, it makes no sense to say that the gap between WAR lines and remittance lines shows the amount that Level 3 underbilled 911 charges. So the district court did not abuse its discretion in excluding Hebert's calculation of damages because it depended on the same ill-founded assumption that led to the exclusion of Gillan's testimony.
2.
Without their expert reports, the Districts present little evidence to establish that Level 3 underbilled 911 charges. They try to avoid summary judgment by arguing that the district court erred in not allowing the WARs as "secondary evidence" under Federal Rule of Evidence 1004. They say the WARs are secondary evidence of an "original" dataset Level 3 refused to produce that shows the lines Level 3 should have but did not charge. This is an odd argument. The district court did not exclude the WARs as evidence. Instead, it concluded that they do not create a dispute of fact sufficient to overcome summary judgment because they say nothing about the number of lines Level 3 should have billed the 911 charge.
So the Districts' argument is just another attempt to establish a link between the WARs and Level 3's 911-billable lines. And we have already explained why the record establishes that such a link does not exist. Level 3's WARs are simply not a reliable or accurate proxy for lines on which it owed a 911 charge. And so we question whether the WARs are even relevant, let alone admissible secondary evidence. After all, if they do not accurately represent the number of Level 3's 911-billable lines, then they do not make any fact of consequence more or less probable. Fed. R. Evid. 401. They represent a figure that is immaterial to this litigation.
Even putting all that to the side, the Districts' Rule 1004 argument is still a curious one. Level 3 said in its interrogatory response that it does not keep reports listing the number of lines on which it does not charge 911 fees. But Rule 1004 allows admission of secondary evidence only if the original is unavailable. See Fed. R. Evid. 1004 (allowing other evidence of an original if the original is lost or destroyed or can't be obtained by judicial process). The Rule thus presupposes that the original existed at some point. Otherwise, it makes no sense for it to allow secondary evidence only if the original is "lost" or "destroyed" or cannot otherwise be obtained. Fed. R. Evid. 1004. But here, it appears the "original"—a record detailing the lines Level 3 should have but did not bill—never existed in the first place, so there is no original to replace with a secondary.
The Districts' argument here also fails even if we assume the WARs are secondary evidence of Level 3's billing data more generally. And for an obvious reason: Level 3 produced all its billing data, and the Districts never requested more discovery. So there is no missing original to substitute with the WARs.
And that's not all—the Districts run into yet another snag here. "Where the secondary evidence is in tangible form, such as a writing, the proponent must prove two facts: the original was itself authentic and the writing is an accurate reproduction of the contents of the original." Wright & Miller, Fed. Practice & Proc: Evidence § 8013. And the proponent must preliminarily satisfy the court by a reasonable-juror standard that "that the secondary evidence correctly reflects the contents of the original." United States v. Gerhart, 538 F.2d 807, 809 (8th Cir. 1976). But the Districts do not show that the WARs accurately reflect the contents of Level 3's nonexistent billing dataset showing the lines Level 3 should have but did not bill the 911 charge. Indeed, as discussed, the WAR's are potentially over-inclusive—by a lot—of the number of lines a telecoms company must bill the 911 charge. (Appellant's App. 140A, Robles Dep. 366:10-21 (noting that the billing spreadsheets included services that weren't eligible for the 911 charge).) So the WARs do not "accurately reflect[] the contents of the original," if such an original ever even existed. Gerhart, 538 F.2d at 810. Rule 1004 is no help to the Districts.
The Districts make one last ditch effort to link the WARs to 911 charges, this time apart from Rule 1004. They claim Michael Robles admitted in his deposition that the billing reports Level 3 produced do not show the lines Level 3 should have billed but did not. So, they argue, Level 3's refusal to produce complete billing data sets justifies using the WARs as a proxy for Level 3's unbilled lines. But like they did with the interrogatories, the Districts take Robles's comment out of context. This line of questioning was limited to one service Level 3 provides: Enhanced Local Service (ELS). And the testimony related only to a subset of ELS customers whom Level 3 charged the 911 fee. But ELS is one of only a couple of services Level 3 offers that can connect a user to 911. And even then, Level 3 sells 99 percent of its 911-capable services wholesale—so it is not responsible for collecting 911 charges on those services. All in all, then, only 1 percent of its 911-capable services are eligible for the 911 charge. Given that such a small percentage of Level 3's services are 911 billable, but Level 3 still reported many of those services on its WARs, the WARs are not an appropriate proxy for the lines Level should have but did not bill the 911 charge.
The district court did not abuse its discretion in failing to admit the WARs as secondary evidence under Rule 1004 or otherwise.
3.
The Districts also challenge the district court's denial (as moot) of their motion to strike "expert" testimony in the Robles Declaration. They do this in multiple iterations throughout their appellate brief. And they point to several instances in the Robles Declaration that show Robles relying on technical opinions, arguing Level 3 never disclosed Robles as an expert.
But the district court denied the motion to strike as moot because it granted summary judgment without relying on any of the content from the Robles Declaration. In fact, the district court cited the Declaration in only one paragraph, and even then, it was to provide context to a statement from the Declaration that the Districts were citing. The Districts have to take the bitter with the sweet: they cannot cite the Robles Declaration then cry foul when the district court cites it back to fill in the context they left out. At any rate, the district court did not rely on the Robles Declaration before finding that no genuine dispute of material fact remained. So the district court properly dismissed the motion to strike as moot.
But the Districts still want the Robles Declaration excluded. They say there are several inconsistencies between Robles's deposition testimony and his Declaration. For instance, they say Robles testified at his deposition that he had not done the required "analysis" to say whether Level 3's customers were facilities-based carriers, but in his Declaration, he said that all of Level 3's wholesale provider customers were facilities-based. The Districts, though, take this out of context. In his deposition, Robles said he would need to do analysis to determine which customers specifically were facilities-based resellers. The Districts' counsel asked Robles for a list of Level 3's customers that were facilities-based resellers. But Robles could only name one specific company. He never said, however, that he would need to analyze whether Level 3's customers generally were facilities-based; he just could not name specific companies off the top of his head.
The Districts also claim Robles contradicted himself in his Declaration when he said the billing spreadsheets contain all the information necessary to determine which lines Level 3 did not bill 911 charges. They say Robles said at his deposition that "spreadsheets produced by Level 3 do not contain information needed to determine the lines for which Level 3 should have billed 911 Charges but did not." (Appellant Br. at 41.) But that is not what Robles said. The Districts point to a portion of Robles's testimony where he was talking about just one type of service Level 3 provides: Enhanced Local Service. And although ELS can connect a user to 911, Level 3 provides most of its 911-capable services wholesale rather than directly to end users. So the Districts try to give Robles's deposition testimony broader import than its context allows.
Besides, the Districts spend a lot less time talking about specific instances of expert opinion in the Declaration than they do about its alleged inconsistencies. But it is the former, not the latter, that forms the basis for their motion to strike. And they claim to have identified fifty-seven instances of expert testimony in the Robles Declaration. But they identify only one in their appellate brief. And though the language in the cited instance is "technical," it is a factual account of the types of services Level 3 provides, information Robles gathered by reviewing corporate records. And as Level 3's 30(b)(6) representative, he was tasked with testifying "about information known or reasonably available to the organization." Fed. R. Civ. P. 30(b)(6).
All told, the district court did not abuse its discretion in denying the motion to strike the Robles Declaration as moot.
4.
The Districts also claim the district court erred in holding that they could not rely on Level 3's interrogatory responses because Level 3 made the response only on information and belief, so the Districts could not introduce it in admissible form at trial. But this ruling was only "an additional basis for concluding that Plaintiffs' arguments based on Interrogatory 4 cannot supply Plaintiffs' missing evidence of liability." (R. 139, Mem. at 14, PageID # 6128.) So whatever error the district court made, if any, it did not affect the outcome.
B.
After all this, the Districts' last remaining argument is that the district court decided disputed facts and did not draw appropriate inferences in their favor. The Districts divide their proposed disputed facts and inferences into two periods: pre- and post-December 2011. That is when Level 3 changed the way it filled out its WARs, moving from telephone numbers to a line-count table. The Districts' pre-December 2011 argument is that Level 3 provided ELS and Session Initiation Protocol, which can call 911; that Level 3 used telephone numbers to bill customers and charge the 911 fee; and it also used telephone numbers to fill out WARs. Thus, linked by the common element of telephone numbers, the WARs are an accurate proxy for Level 3's 911 billable lines.
But the Districts' proposed facts do not create a genuine dispute of material fact. Start with the same fact the Districts do. "Level 3 sold Enhanced Local Service (ELS) and SIP lines that could access the telephone network and call 911." (Appellant Br. at 16.) That does not tell the whole story. The undisputed facts in the record show that while ELS and SIP can indeed call 911, those services are a minority of Level 3's business. And Level 3 sells most of its SIP and ELS wholesale. Those services are not subject to the 911 charge, and certainly not one owed to the Districts.
Next, the Districts try to link Level 3's WARs, customer billing, and 911 obligation through telephone numbers. It says Level 3 used these numbers for both the 911 charge and the WARs. But the sets of numbers are not equivalent. The district court pointed out why this argument fails:
Level 3's decision, whether mistaken or not, to use telephone numbers to complete a voluntary report, for a different entity, calling for a measurement of its facilities in voice-grade equivalents, is not a concession that those telephone numbers had a one-to-one correspondence to exchange access lines that could be used to make outgoing voice calls to 911.(R. 139, Mem., PageID # 6128.) And since the facts the Districts use to draw their inferences are inaccurate, the inferences they draw are, too. Indeed, many of the "inferences" are just iterations of the same conclusion that Level 3 underbilled 911 charges. But nothing in the record suggests that is true.
So the Districts' pre-December 2011 facts and inferences fail. So too do their post-December 2011 facts and inferences. The Districts point out that counsel for the Hamilton County ECD wrote Level 3 a letter notifying Level 3 that the ECD was beginning to investigate Level 3's remittance reports. Then, Level 3 did not file any WARs for the next two months. And after it began to file WARs again, suddenly its reported line count plummeted. The Districts infer from these facts that Level 3 tried to hide evidence of wrongdoing. But that is hardly a reasonable inference. Indeed, as both Level 3 and the district court pointed out, no one before this litigation used WARs as a proxy for 911-billable lines. And the letter said nothing about Level 3's WARs. Level 3 thus would have had to correctly guess that the Districts were comparing two things that had never been compared before. That is not a reasonable inference. So the Districts' post-2011 facts and inferences do not create a genuine dispute of material fact.
That leaves the Districts with two pieces of evidence: their reading of Level 3's interrogatory response, and the information from the Robles Declaration and deposition. And we have already discussed why the Districts' reading of Level 3's response to Interrogatory 4 does not create a genuine dispute of fact. So all that is left to address is the Robles testimony.
The Districts' last argument is that Michael Robles, Level 3's corporate representative, said that 40 percent of the lines Level 3 sold could call 911 and thus should have been billed. But again, they take that comment out of context. Robles followed this line up by noting that of that 40 percent, Level 3 sold only 1 percent of the lines retail and only beginning in 2009—the other 99 percent Level 3 sold wholesale to other facilities-based telecoms and VoIP providers. The Districts' misreading of the Robles Declaration is not enough to create a genuine dispute of material fact.
III.
We AFFIRM the district court's grant of summary judgment.