Qwest Corp. v. Federal Communications Commission

HOLMES, Circuit Judge. Petitioner Qwest Corporation (“Qwest”) seeks our review of an order of the Federal Communications Commission (“Commission”) denying Qwest’s petition for regulatory forbearance pursuant to 47 U.S.C. § 160(a). Qwest filed a petition with the Commission in March 2009 seeking relief from certain regulations pertaining to telecommunications services that it provides in the Phoenix, Arizona, metropolitan statistical area (“MSA”). The Commission denied the petition, citing insufficient evidence of sufficiently robust competition that would preclude Qwest from raising prices, unreasonably discriminating, and harming consumers. Qwest challenges the Commission’s decision only as it pertains to Qwest’s mass-market retail services in the Phoenix MSA. For the reasons set forth below, we deny Qwest’s petition. I In the Telecommunications Act of 1996 (“1996 Act”), Congress upended the existing telecommunications regulatory regime and imposed on the monopolistic local phone companies (called “local exchange carriers” or “LECs”) several new requirements designed to enhance competition in the market for local telephone service. See Qwest Corp. v. Colo. Pub. Utils. Comm’n, 656 F.3d 1093, 1096 (10th Cir.2011). Foremost among these is the requirement that incumbent carriers share their networks with competitors. Because new market entrants would find it prohibitively costly to replicate the infrastructure necessary to provide local service, the 1996 Act requires incumbent carriers to provide competitors with access to existing network elements on an unbundled basis at *1217 “just” and “reasonable” rates. See 47 U.S.C. § 251(c)(3). The Commission is responsible for determining which unbundled network elements (“UNEs”) an incumbent LEC must make available to competitors. See id. § 251(d)(2)(B). A second critical feature of the 1996 Act is section 10, codified at 47 U.S.C. § 160. Because newly competitive conditions could make the heavy-handed regulation of incumbent carriers obsolete, section 10 provides that the Commission “shall forbear” from applying certain statutory or regulatory requirements to an incumbent carrier if it determines that those requirements are (1) not necessary to ensure just, reasonable, and nondiscriminatory terms of service, (2) not necessary to protect consumers, and (3) consistent with the public interest. See 47 U.S.C. § 160(a). A carrier can petition the Commission for forbearance. Id. § 160(c). The petition “shall be deemed granted if the Commission does not deny the petition for failure to meet the requirements for forbearance under subsection (a) ... within one year.” Id. (emphasis added). The Commission may extend that one-year deadline by an additional ninety days. See id. The dispute here arises out of a June 2010 order of the Commission denying Qwest’s petition for forbearance from unbundling obligations and dominant-carrier regulations pertaining to Qwest’s provision of mass-market services in the Phoenix MSA. See In the Matter of Petition of Qwest Corporation for Forbearance Pursuant to 47 U.S.C. § 160(c) in the Phoenix, Arizona Metropolitan Statistical Area, 25 FCC Rcd. 8622, 8677, ¶ 109, 2010 WL 2526677 (2010) [hereinafter Phoenix Order]. Before reviewing that order, however, some additional background is necessary. In particular, we summarize three prior orders of the Commission that provide a central backdrop to the issues in this appeal. After providing key background information, we describe the Phoenix Order and the Commission’s reasons for denying forbearance to Qwest. A In 2004, Qwest filed a petition for forbearance from . unbundling requirements and dominant-carrier regulations in the MSA of Omaha, Nebraska, where it competes extensively with Cox Communications, a cable provider. In response to the petition, the Commission found that “sufficient facilities-based competition ... exists in certain of Qwest’s Omaha MSA wire center service areas to justify forbearance.” In the Matter of Petition of Qwest Corporation for Forbearance Pursuant to 47 U.S.C. § 160(c) in the Omaha Metropolitan Statistical Area, 20 FCC Rcd. 19,415, 19,447, ¶ 64, 2005 WL 3287482 (2005) [hereinafter Omaha Order). To reach this conclusion, the Commission essentially applied a two-prong test. Under the first prong (the “market-share test”), the Commission assessed the level of retail competition in the Omaha market and found it “compelling” that Qwest’s market share for retail mass-market customers was “less than” a specified percentage. 1 Id. at 19,430, ¶ 28; see also id. at 19,448, ¶ 66. Under the second prong (the “coverage test”), *1218 the Commission considered the geographic reach of Cox’s cable network and found that, in certain locations called “wire centers,” Cox had deployed facilities capable of reaching a very significant percentage of end-users. See id. at 19,446, ¶ 62. Concluding that Cox could successfully compete with Qwest, the Commission granted forbearance to Qwest with respect to those wire centers. See id. at 19,447, ¶ 64. Then, in 2006, Verizon Telephone Companies (“Verizon”) sought regulatory forbearance in six MSAs where it provided services as an incumbent carrier. The Commission denied Verizon’s petition, finding insufficient evidence of facilities-based competition. See In the Matter of Petitions of the Verizon Telephone Companies for Forbearance Pursuant to 47 U.S.C. § 160(c) in the Boston, New York, Philadelphia, Pittsburgh, Providence and Virginia Beach Metropolitan Statistical Areas, 22 FCC Rcd. 21,293, 21,313, ¶ 37, 2007 WL 4270630 (2007) [hereinafter Verizon Six-MSA Order]. Specifically, the Commission stated, “Overall, in all of the 6 MSAs, it appears that cable operators are presently making some competitive gains against Verizon by providing voice service to consumers in the residential markets, however competition from cable operators does not yet present a sufficient basis for relief.” Id. at 21,314 n. 116. In assessing Verizon’s market share in a given MSA, the Commission looked to the number of Verizon customers as a percentage of both the total number of customers who subscribed to landline (“wireline”) service in the MSA and the total number of customers who subscribed exclusively to mobile wireless service — viz., customers who had “cut the cord.” See id. at 21,323, App. B; see also id. at 21,308 n. 89, ¶ 27 (“[BJased on the record here, and consistent with recent precedent, we include [in market-share computations] cut-the-cord wireless substitution.”). Estimates for the latter group of customers were not available on an MSA-specific basis, so the Commission utilized a national estimate and “assume[d]” that 12.8% of households in a given MSA had cut the cord. Id. at 21,323, App. B. In 2007, Qwest sought regulatory forbearance in four MSAs, including Phoenix. The Commission acknowledged that the coverage test articulated in the Omaha Order was met because Cox, Qwest’s primary competitor in Phoenix, was capable of reaching at least seventy-five percent of end-users with its own facilities. See In the Matter of Petitions of Qwest Corporation for Forbearance Pursuant to 47 U.S.C. § 160(c) in the Denver, Minneapolis-St. Paul, Phoenix, and Seattle Metropolitan Statistical Areas, 23 FCC Rcd. 11,729, 11,754, ¶ 35 & n. 127, 2008 WL 2884078 (2008) [hereinafter Qwest Four-MSA Order]. Focusing on the market-share test, the Commission opined: In calculating market shares, we believe it is appropriate to include wireless-only households (ie., residential telephone customers who have “cut the cord”). In particular, we find that mobile wireless service should be included in the local services product market to the extent that it is used as a complete substitute for all of a consumer’s voice communications needs. Over the past several years, as wireless substitution rates have continued to rise, the Commission has begun including such intermodal substitution in its competitive analyses of the local services market. Id. at 11,742, ¶ 19 (footnote omitted). The Commission nevertheless found that Qwest did not meet the market-share test. The Commission noted that it had previously sought to determine whether carriers have “a retail market share of less than 50 percent.” Id. at 11750, ¶28. In particular, the Commission found fault with the geographic scope and the reliability of Qwest’s market-share data. In this *1219 regard, the Commission declined to rely on “national wireless-only household data published by the Center for Disease Control (CDC) or the more localized information,” id. at 11,743, ¶ 21, which the Commission considered to be uninformative and unreliable, see id. at 11,744-45, ¶21. In sum, the Commission stated, “Qwest seeks regulatory relief for particular MSAs based on the specific competitive conditions in those markets, but the CDC estimates and the record generally do not contain reliable data of this type.” Id. at 11,744, ¶21. More specifically, the Commission concluded: Qwest has not sufficiently supported its case for forbearance on the basis of reliable, geographically-specific data regarding the measure of wireless substitution in the four MSAs.... We emphasize that petitioners relying on mobile wireless substitution to support forbearance relief should submit complete and reliable data that is geographically specific to the areas for which forbearance is sought. Id. at 11,745, ¶ 22. Because reliable Phoenix-specific data was not available in the record, the Commission calculated Qwest’s market share without accounting for cut-the-cord customers and denied the request for forbearance. See id. Verizon, and then Qwest, appealed the Commission’s orders to the D.C. Circuit. That court stayed Qwest’s appeal and heard Verizon’s challenge first. Verizon argued that the Commission’s exclusive reliance on existing market share to deny forbearance was arbitrary and capricious, and the D.C. Circuit agreed. See Verizon Tel. Cos. v. FCC, 570 F.3d 294, 302 (D.C.Cir.2009). The court noted that the Commission had “zeroed in on Verizon’s market share as the dispositive factor in its ... analysis.” Id. at 301. However, in previous orders, including the Omaha Order, the Commission had considered both actual competition (as represented by existing market share) and the potential for competition from other carriers. See id. at 303-04. The court concluded that the “per se market share test” applied in the Verizon Six-MSA Order marked an “unexplained departure from [the Commission’s] precedent.” Id. at 296. The court remanded the order, directing the Commission either to consider factors other than a market-share benchmark or to “justify its departure from ... precedent.” Id. at 305. Because the Qwest Four-MSA Order employed a similar analytical approach, the Commission sought and was granted a voluntary remand of that order. Order, Qwest Corp. v. FCC, No. 08-1257 (D.C.Cir. Aug. 5, 2009). B Following remand of the two orders, in March 2009, Qwest filed a petition for forbearance specific to the Phoenix MSA. In particular, it sought to explain its low and dwindling retail market share in the Phoenix MSA by providing the Commission with Phoenix-specific cut-the-cord data. See Joint App. at 261-65 (Petition of Qwest Corp. for Forbearance Pursuant to 47 U.S.C. § 160(c), dated Mar. 24, 2009). Qwest acknowledged that, in the Qwest Four-MSA Order, the Commission had rejected its forbearance request in part because of its “lack of Phoenix-specific wireless substitution data that the Commission deemed reliable.” Id. at 245. “To address concerns raised by the Commission as to the reliability and generality of its wireless substitution study,” Qwest retained a consulting firm “to determine the' penetration of wireless-only households in the Phoenix MSA.” Id. at 261. Qwest said that its data indicated that Qwest’s market share, taking cord-cutting into account, was “substantially less than the 50% market share figure the Commission has previously relied upon.” Id. at 246; see id. at 262 (noting that “25 percent *1220 of surveyed Phoenix households reported that they are relying upon wireless services for all of their communication needs”). Highlighting that it previously had demonstrated in the Qwest Four-MSA Order proceeding that Cox’s facilities-based coverage in certain areas of Phoenix surpassed seventy-five percent, Qwest noted “publicly available information” indicated that “Cox currently appears to offer voice services even more widely in Phoenix” than Qwest had shown earlier. Id. at 246-47. Thus, Qwest believed it qualified for forbearance. Significantly, Qwest expressly noted that “[t]he analysis of competition in [its] Petition is designed to be consistent with the analytical framework applied by the Commission in Qwest’s earlier request for forbearance in the Omaha MSA, as well as the analysis applied in the subsequent Verizon Six MSA and Qwest k MSA proceedings.” Id. at 245-46 (footnotes omitted). However, at that time, the Commission was reconsidering its analytical framework for forbearance petitions. In August 2009, it sought comments on the analytical approach that it should employ in resolving the forbearance issues presented in the remanded orders. In particular, it sought input regarding whether it should “depart from its recent precedent regarding marketplace analysis in forbearance petitions,” and concerning the kinds of evidence beyond “market share for a particular product market” that are relevant in determining “whether forbearance from unbundling regulations is warranted.” Wireline Competition Bureau Seeks Comment on Remands of Verizon 6 MSA Forbearance Order and Qwest I MSA Forbearance Or der, 24 FCC Red. 10,881, 10,883, 2009 WL 2595760 (2009) [hereinafter Remand Comment Request ]. Most significantly, in April 2010, the Commission also issued a public notice requesting additional comments and data specifically related to Qwest’s request for forbearance in the Phoenix MSA. See Joint App. at 1198 (Request for Additional Comment and Data Related to Qwest Corporation’s Petition for Forbearance, dated Apr. 15, 2010). The notice stated that a number of commenters had urged the Commission to analyze Qwest’s petition in light of “basic principles of competition policy and the [Federal Trade Commission-Department of Justice] Horizontal Merger Guidelines” and to “examine Qwest’s market power in discrete product and geographic markets under the standards of the FTC-DOJ Horizontal Merger Guidelines and the similar market power analysis that the Commission has conducted in recent merger decisions.” Id. (quoting Opposition of Integra Telecom, Inc., WC Docket No. 09-135, at 2 (filed Sep. 21, 2009)) (internal quotation marks omitted). The Commission specifically noted its interest in determining the proper analytical approach toward the cut-the-cord phenomenon, seeking comment on “whether, for mass market consumers, mobile wireless service is in the same relevant product market as wireline telephone service.” Id. at 1199. The Commission observed, “A key element of the [FTC-DOJ Horizontal Merger] Guidelines, as they relate to product market definition, includes an inquiry into whether a hypothetical monopoly provider of a service [i.e., wireline voice service] profitably could impose a ‘small but significant and nontransitory’ increase in the price of such service.” Id. and 1199 n.4. (quoting U.S. Dep’t of Justice & Fed. Trade Comm’n, Horizontal Merger Guidelines §§ 1.11, 1.12 (Apr. 2, 1992, revised Apr. 8, 1997), available at http://www. justice.gov/atr/public/guidelines/hmg.htm). “We ask Qwest and other interested persons,” the notice concluded, “to cite specific evidence in the record or provide new data as needed to support their pleadings in response to these issues.” Id. (emphasis added). *1221 Two months later, in June 2010, the Commission denied Qwest’s forbearance request. See Phoenix Order, 25 FCC Red. at 8677, ¶ 109. The Commission began by repudiating the test it had articulated in the Omaha Order, finding that test “problematic” and “not adequately justified as a matter of economics.” Id. at 8633-34, ¶¶24, 26. It explained that the first prong, the market-share test, was unduly narrow in focus and inadequate for assessing a carrier’s market power — i.e., its ability to maintain prices above competitive levels. See id. at 8635, ¶ 28. The second prong of the test — the extent of facilities-based coverage by a single competitor such as Cox — “inappropriately assumed that a duopoly always constitutes effective competition.” 2 Id. at 8635-36, ¶ 29. The Commission explained that duopolies pose “significant risks of collusion and supra-competitive pricing” that are inconsistent with consumer welfare. Id. at 8636, ¶ 29. The Commission also found that real-world developments in the Omaha market following its grant there of forbearance were not consistent with predictions it had made in the Omaha Order. See id. at 8639-41, ¶34. For example, the Omaha Order predicted that Qwest and other providers in the area would continue to offer wholesale services at competitive prices despite elimination of the requirement that Qwest share some of its UNEs. See id. at 8640, ¶ 34. However, the record showed that competition in Omaha had actually decreased since the order was issued: McLeod USA, “the only other competitor of significant size” in the Omaha area, had significantly curtailed its operations there, and Integra, which had been contemplating entry into the Omaha market, had abandoned its plans to do so after issuance of the order. Id. at 8640-41, ¶ 34. “With the benefit of hindsight and upon further consideration,” id. at 8633, ¶ 24, the Commission decided to adopt “a more comprehensive analytical framework” to evaluate the state of competition in telecommunications markets, id. at 8642, ¶ 37. As the April 2010 public notice had hinted, the Commission “return[ed] to a traditional market power framework,” an analytical approach employed in earlier proceedings and embodied in the FTC-DOJ Horizontal Merger Guidelines. Id. As the Commission explained it, a market-power analysis begins by delineating the relevant product and geographic markets and identifying market participants, then examines market-share data, and finally considers whether the potential for competitive market entry is sufficient to constrain an incumbent carrier’s ability to maintain prices above competitive levels. See id. at 8646-47, ¶ 42. The Commission found that this analysis was consistent with approaches used by other U.S. and foreign regulators; was “better suited” for analyzing competitive conditions in a given market; was more consistent with the letter and spirit of section 10; and adequately accounted for both actual and potential competition. Id. at 8642-44, ¶¶ 37-38. With its new framework in place, the Commission proceeded to analyze Qwest’s petition. As relevant to this appeal, 3 the Commission began by defining the relevant product market for Qwest’s mass-market retail services. Id. at 8649, ¶ 51. This included traditional wireline service, as well as facilities-based (but not over-the-top) VoIP services. 4 Id. at 8650, ¶¶ 53-54. *1222 The Commission then discussed whether to include mobile wireless services in the product market. Id. at 8651, ¶ 55. Under the new market-power framework, inclusion depended on whether the Phoenix residents’ increasing use of wireless services “materially constrained] the price of residential wireline voice service[s].” Id. The Commission elaborated on this point: These two services should be in the same relevant market only if the prospect of buyer substitution to mobile wireless access constrains the price of wireline access. The first question before us then is whether a hypothetical profit-maximizing firm that was the only present and future seller of wireline local access services could profitably impose a small but significant and nontransitory increase in price (S SNIP). In other words, we consider whether there are a sufficient number of wireline service customers who, in response to a price increase in wireline local access service, would stop subscribing to their wireline service and instead rely exclusively on mobile wireless service, so as to render the price increase unprofitable. Id. at 8651-52, ¶ 56 (footnote omitted). The Commission averred that the issue was “complicated” and “one that is evolving over time.” Id. at 8651, ¶ 55. On the one hand, the Commission observed that “[t]he increasing percentage of residential customers that rely solely on mobile wireless voice service suggests that an increasing percentage of voice customers view wireless and wireline services as close substitutes,” thus “increasing the likelihood” that wireless services “may materially constrain the price of residential wireline voice service” and belong in the same product market as wireline services. Id. On the other hand, it noted that “the Commission, the DOJ, and foreign regulators have previously found that mobile wireless service does not constrain the price of wireline service.” Id. at 8652, ¶ 57. Ultimately, the Commission determined that the record before it did not allow it to make a finding “for purposes of Qwest’s forbearance request” that wireless voice services have a material price-constraining effect with respect to wireline voice services. Id. at 8651, ¶ 55. The Commission elaborated on the point and, significantly, identified certain deficiencies in Qwest’s evidentiary showing: [N]either Qwest nor any other commenter has submitted evidence that would support a conclusion that mobile wireless service constrains the price of wire-line service. For example, Qwest has produced no econometric analyses that estimate the cross-elasticity of demand between mobile wireless and wireline access services. [ 5 ] Nor has it produced *1223 any evidence that it has reduced prices for its wireline services or otherwise adjusted its marketing for wireline service in response to changes in the price of mobile wireless service. Nor has it produced any marketing studies that show the extent to which consumers view wireless and wireline access services as close substitutes. Id. at 8653-54, ¶ 58 (footnotes omitted). The Commission noted that, although Qwest had provided evidence concerning the percentage of households in the Phoenix area that depend exclusively on mobile wireless services, that “cannot alone establish whether mobile -wireless services should be included in the same relevant product market as residential wireline voice service.” Id. at 8654, ¶ 59. Notably, the Commission observed, “Knowing the percentage of households that rely exclusively upon mobile wireless is insufficient to determine whether mobile wireless services have a price-constraining effect on wireline access services.” Id. In this regard, it stated that “while ... the number of customers that rely solely on mobile wireless service has been growing steadily, we find that other reasons may explain the growth in the number of wireless-only customers, besides an increasing eross-elasticity of demand between mobile wireless and wireline services.” Id. Thus, the Commission excluded mobile wireless services from its analysis. See id. at 8655, ¶ 60. It noted, however, that it was “mak[ing] no affirmative finding that mobile wireless services do not currently, or may not soon, belong in the same product market as residential wireline voice services.” Id. It also “acknowledge^ ... that more consumers may view mobile wireless as a closer substitute for wireline voice service than in the past.” Id. However, in the Commission’s view, there was “insufficient data in the record to make such a determination here.” Id. Having excluded mobile wireless services, the Commission identified the market participants in the Phoenix area as Qwest, Cox, and various smaller competitors that relied predominantly, if not exclusively, on Qwest facilities. Id. at 8657, ¶¶ 64-67. It found that retail mass-market services in the region were “highly concentrated with two dominant providers, Qwest and Cox.” Id. at 8664, ¶ 80. That duopolistic structure, with the potential for tacit price coordination, necessitated an inquiry into whether any other competitors in the Phoenix MSA had deployed or could deploy their own facilities to any significant degree and also into the potential for de *1224 novo entry by new competitors. See id. at 8665-67, ¶¶ 82-84. On both fronts, the Commission found competition in the area insufficiently robust to put downward pressure on Qwest’s prices. See id. at 8667-68, ¶ 86. That finding formed the basis for the Commission’s conclusion that regulatory requirements, particularly unbundling, remained necessary for continued assurance of “just, reasonable, and non-discriminatory” terms of service. See id. at 8673, ¶ 98. The Commission also held that these requirements “remain[ed] necessary to protect consumers,” id. at 8674, ¶ 101, and that forbearance was not in the public interest because it would not “promote competitive market conditions,” id. at 8675, ¶ 104 (quoting 47 U.S.C. § 160(b)) (internal quotation marks omitted). The Commission denied forbearance. Qwest now challenges that decision in its petition for review. II A “We have jurisdiction to review final orders of the [Commission] under 28 U.S.C. § 2342(1).” Qwest Corp. v. FCC, 258 F.3d 1191, 1198 (10th Cir.2001). We ordinarily defer under Chevron, U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 104 S.Ct. 2778, 81 L.Ed.2d 694 (1984), to the Commission’s formal interpretations of the 1996 Act made in, for instance, notice-and-comment rulemaking. See Qwest Commc’ns Int’l, Inc. v. FCC, 398 F.3d 1222, 1229-30 (10th Cir.2005); see also United States v. Mead Corp., 533 U.S. 218, 230, 121 S.Ct. 2164, 150 L.Ed.2d 292 (2001) (“[T]he overwhelming number of our eases applying Chevron deference have reviewed the fruits of notice-and-comment rulemaking or formal adjudication.”); Mead Corp., 533 U.S. at 230, 121 S.Ct. 2164 (“It is fair to assume generally that Congress contemplates administrative action with the effect of law when it provides for a relatively formal administrative procedure tending to foster the fairness and deliberation that should underlie a pronouncement of such force.”). If “Congress has not directly addressed the precise question at issue,” we ask whether the Commission’s formal interpretation is a “permissible construction of the statute,” and we will not substitute our own views for a “reasonable interpretation” by the Commission. Chevron, 467 U.S. at 843-44, 104 S.Ct. 2778. In reviewing the Commission’s decisionmaking process, we ask whether the Commission’s action was “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with the law.” Sorenson Commc’ns, Inc. v. FCC, 659 F.3d 1035, 1045 (10th Cir.2011) (“Sorenson II ”) (quoting 5 U.S.C. § 706(2)(A)) (internal quotation marks omitted). This is a “narrow” standard of review, and we require only that the Commission “examine the relevant data and articulate a satisfactory explanation for its action.” FCC v. Fox Television Stations, Inc., 556 U.S. 502, 513, 129 S.Ct. 1800, 173 L.Ed.2d 738 (2009) (quoting Motor Vehicle Mfrs. Ass’n of U.S., Inc. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43, 103 S.Ct. 2856, 77 L.Ed.2d 443 (1983)) (internal quotation marks omitted). Generally, when a decision by the Commission represents a change in policy, our review is no more searching; in other words, no heightened level of scrutiny attends a policy change. See Rivera-Barrientos v. Holder, 666 F.3d 641, 645 (10th Cir.2012) (“This standard of review is not more searching where the agency’s decision is a change from prior policy.” (citing Fox Television, 556 U.S. at 513-14, 129 S.Ct. 1800)); see also Sorenson Commc’ns, Inc. v. FCC, 567 F.3d 1215, 1221 (10th Cir.2009) (“Sorenson I ”) (“The same stan *1225 dard of review applies to both initial policy decisions and subsequent changes in policy.”). “[I]t suffices that the new policy is permissible under the statute, [and] that there are good reasons for it....” Fox Television, 556 U.S. at 515, 129 S.Ct. 1800. However, some additional justification is required if the Commission’s new policy “rests upon factual findings that contradict those which underlay its prior policy” or the prior policy “has engendered serious reliance interests that must be taken into account.” Id. Brushing aside such matters would be arbitrary and capricious, and thus we would require the Commission to offer a “reasoned explanation ... for disregarding facts and circumstances that underlay or were engendered by the prior policy.” Id. at 516, 129 S.Ct. 1800. B Qwest challenges the Commission’s Phoenix Order on two grounds. First, it argues that section 10 requires the Commission to make affirmative findings on the substantive prerequisites for granting forbearance; that the Commission here made no such affirmative findings; and, therefore, that the request for forbearance should be “deemed granted” by operation of law. Second, Qwest argues that the Commission’s decision was irrational. We reject both contentions. We conclude that the Phoenix Order was a reasoned and reasonable decision. — 1 — With respect to a petition for forbearance, section 10(c) of the 1996 Act provides as follows: Any such petition shall be deemed granted if the Commission does not deny the petition for failure to meet the requirements for forbearance under subsection (a) of this section within one year after the Commission receives it.... The Commission may grant or deny a petition in whole or in part and shall explain its decision in writing. 47 U.S.C. § 160(c). Qwest reads this provision as a “default rule” — a “statutory thumb on the scale in favor of forbearance” — and argues that the Phoenix Order was tantamount to a default triggering a “deemed grant[ ]” of forbearance because the Commission did not affirmatively deny Qwest’s petition on the merits. Qwest Opening Br. at 29-30. In Qwest’s view, it “met all preexisting standards for forbearance,” and the Commission denied its petition, not because it failed to meet those standards, but because the Commission “wished to suspend judgment on whether those standards are in fact the right ones.” Id. at 28. Qwest argues that “the legal consequence of such bureaucratic indecision is to ‘deem[]’ the petition ‘granted’ by operation of law.” Id. at 29 (quoting 47 U.S.C. § 160(c)). The Commission counters that Qwest is improperly attempting to reverse the burden of proof in forbearance proceedings. See Aplee. Br. at 28-29. In the Commission’s view, section 10(c) does not make forbearance the “default,” and it does not put the onus on the Commission to prove that the prerequisites for forbearance are not satisfied. Rather, section 10(c) “simply means” that the Commission must attend promptly to forbearance petitions, and the burden is on the petitioner to show that a regulatory obligation is not needed and that forbearance is consonant with the public interest. Id. at 29-30 (citing 47 U.S.C. § 160(a)(l)-(3)). The Commission says that Qwest failed to meet that burden here and the Phoenix Order so concluded. See id. at 32-33. We first address the issue of burden of proof. Although the statute says nothing about it, the Commission has determined through a notice-and-comment proceeding that the burden of proof — en *1226 compassing the burdens of both production and persuasion — is on the petitioner. See In the Matter of Petition to Establish Procedural Requirements to Govern Proceedings for Forbearance Under Section 10 of the Communications Act of 1934, As Amended, 24 FCC Rcd. 9543, 9554-55, ¶ 20 (2009) [hereinafter Forbearance Procedures Order], That construction of section 10(c) is entitled to our deference. See Mead Corp., 533 U.S. at 230, 121 S.Ct. 2164 (holding that agency interpretations promulgated by notice-and-comment rule-making warrant Chevron deference). And we believe the construction to be a reasonable one. See Schaffer ex rel. Schaffer v. Weast, 546 U.S. 49, 57, 126 S.Ct. 528, 163 L.Ed.2d 387 (2005) (concluding that, where a statute was silent on the burden of persuasion and “[a]bsent some reason to believe that Congress intended otherwise,” the burden “lies where it usually falls, upon the party seeking relief’); see also Zhang v. Slattery, 55 F.3d 732, 756 (2d Cir.1995) (holding that in the absence of an express statutory allocation of the burden of proof, the Board of Immigration Appeals’s placement of the burden on the alien was reasonable under Chevron), superseded by statute on other grounds, 8 U.S.C. § 1101(a)(42); Panhandle Producers & Royalty Owners Ass’n v. Econ. Regulatory Admin., 847 F.2d 1168, 1175-76 (5th Cir.1988) (holding that an agency’s policy shift, reallocating the burden of proof from natural gas importers to parties opposing importation, “was based on a reasonable interpretation of the Natural Gas Act”). Even so, a plain reading of section 10 makes clear that the Commission has obligations of its own when a party petitions for forbearance. To avoid a “deemed grant[]” of forbearance, the Commission must: (1) deny a petition “for failure to meet the requirements for forbearance” under section 10(a), (2) explain its decision in writing, and (3) do so within the statutorily prescribed period (i.e., one year, with a possible extension of up to ninety days). 47 U.S.C. § 160(c) (emphasis added). That is, the Commission must, within the statutorily prescribed period, affirmatively find that at least one of the substantive prerequisites for forbearance is not satisfied — for example, that enforcement remains necessary to ensure just, reasonable, and nondiscriminatory terms of service, id. § 160(a)(1); that enforcement remains necessary to protect consumers, id. § 160(a)(2); or that forbearance would not be consistent with the public interest, id. § 160(a)(3) — and offer a written explanation to that effect. The Commission need only find that one of the prerequisites for forbearance is not met to justify denying the petition. See Cellular Telecomms. & Internet Ass’n v. FCC, 330 F.3d 502, 509 (D.C.Cir.2003) (“The three prongs of § 10(a) are conjunctive. The Commission could properly deny a petition for forbearance if it finds that any one of the three prongs is unsatisfied.”). But at least one such finding is necessary for a proper denial and to avoid a “deemed grant[ ].” Qwest argues that the Commission dropped the ball here; that the Phoenix Order “fail[ed] to make the requisite statutory findings” under section 10(a); and that we should therefore “deem[ ]” Qwest’s petition “granted.” Qwest Opening Br. at 30. We are not persuaded. In the Phoenix Order, the Commission explicitly found that maintaining Qwest’s unbundling obligations was necessary both for continued assurance of just, reasonable, and nondiscriminatory terms of service and for protection of Phoenix consumers, see 25 FCC Red. at 8673-74, ¶¶98, 101 (citing 47 U.S.C. § 160(a)(1), (a)(2)), and that forbearance would not be consistent with the public interest, see id. at 8675, ¶ 105 (citing 47 U.S.C. § 160(a)(3)). Moreover, as we explain below, the Commission adequately justified those conclusions, both on the *1227 merits and in light of the conscious policy-shift it was making. Qwest attempts to frame the Phoenix Order as doing nothing more than expressing the Commission’s “uncertainty” about its prior forbearance standards and “thank[ing]” Qwest for its submission. Qwest Opening Br. at 29-30. We need not opine on whether that kind of order would trigger a “deemed grant[ ]” under section 10(c) because we disagree with Qwest’s characterization. In our view, the Commission rendered a reasoned decision on the merits of Qwest’s petition in writing and within the statutorily prescribed time frame. To be sure, as we explicate further below, the Phoenix Order altered precisely how the merits of forbearance petitions are assessed, and it might well be true (though we need not say) that Qwest “met all preexisting standards for forbearance.” Id. at 28 (emphasis added). But we decline to hold that a forbearance petition is “deemed granted” under section 10(c) whenever the Commission embarks upon a policy shift — at least when, as here, we cannot conclude that the Commission’s policy shift was arbitrary or capricious. We hold that the Commission met its procedural obligations under section 10 in denying Qwest’s forbearance petition. In other words, the petition was not “deemed granted” through some procedural omission of the Commission. Accordingly, we proceed to Qwest’s merits-based challenges to the Phoenix Order. — 2 — Qwest attacks the Phoenix Order as irrational, arguing that the Commission “ignored” the cut-the-cord phenomenon, contrary to its own precedent, and that the Commission’s assessment of competitive conditions in the Phoenix market was unreasonable. We reject both arguments and conclude that the Commission’s decision was not “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with the law.” Sorenson II, 659 F.3d at 1045 (quoting 5 U.S.C. § 706(2)(A)) (internal quotation marks omitted). a We begin with Qwest’s arguments concerning the Commission’s treatment of the Phoenix-specific cut-the-cord data supplied by Qwest. We acknowledge that the Commission engaged in some goalpost-moving on this issue. We nevertheless hold that the exclusion of the cut-the-cord data in the Phoenix Order was not arbitrary and capricious based on the unique circumstances of this case and the Commission’s adequate explanation for its shift to a market-power framework. In its Verizon Six-MSA Order, the Commission used a national cut-the-cord estimate to calculate Verizon’s market share in various MSAs. See 22 FCC Rcd. at 21,323, App. B. It found that Verizon did not satisfy the market-share benchmark even including the cut-the-cord data. See id. at 21,307-08, ¶ 27. Subsequently, in the Qwest Four-MSA Order, the Commission found that it was “appropriate to include wireless-only households” in calculating a carrier’s market share and observed that “[o]ver the past several years, as wireless substitution rates have continued to rise, the Commission has begun including such intermodal substitution in its competitive analyses of the local services market.” 23 FCC Red. at 11,742, ¶ 19 (citing several past orders). The Commission, however, faulted Qwest for relying on national estimates of wireless-only households — despite the Commission’s prior reliance on such national data in the Verizon Six-MSA Order. It insisted that Qwest provide reliable cut-the-cord data related to the Phoenix MSA and noted that the more localized data in the *1228 record was neither informative nor reliable. The Commission noted that “Qwest might have qualified for some forbearance upon a better evidentiary showing.” Id. at 11,745, ¶ 22. But it went on to “emphasize that petitioners relying on mobile wireless substitution to support forbearance relief should submit complete and reliable data that is geographically specific to the areas for which forbearance is sought.” Id. In the Phoenix Order, Qwest delivered precisely the sort of reliable data that was lacking in the Qwest Four-MSA Order concerning the percentage of wireless-only voice customers in the Phoenix MSA. And the Commission seemed to acknowledge as much. See Phoenix Order, 25 FCC Red. at 8,654, ¶ 59 & n. 178 (noting that “Qwest submitted studies that estimate the percentage of households that exclusively rely upon mobile wireless services in the Phoenix area” and “acknowledging] that the Commission in the Qwest ¿ MSA Forbearance Order suggested that geographically disaggregated evidence of the percentage of voice subscribers that rely on mobile wireless only might demonstrate that Qwest was entitled to forbearance in the Phoenix MSA.”). However, the Commission moved the goalpost. It criticized Qwest’s evidentiary showing again, this time because Qwest failed to present evidence that consumers consider wireline voice services and wireless voice services to be substitutes, such that the latter materially constrains the price of the former. Specifically, the Commission observed, “Knowing the percentage of households that rely exclusively upon mobile wireless is insufficient to determine whether mobile wireless services have a price-constraining effect on wireline access services.” Id. at 8654, ¶ 59. This kind of goalpost-moving does not reflect an optimal mode of administrative decisionmaking. And we do not foreclose the possibility that under some circumstances an agency’s shifting of the policy goalpost (e.g., the evidentiary requirements for satisfying a particular statutory or regulatory standard) may lead us to conclude that the agency has acted arbitrarily or capriciously. See Smiley v. Citibank (South Dakota), N.A., 517 U.S. 735, 742, 116 S.Ct. 1730, 135 L.Ed.2d 25 (1996) (noting that “[s]udden and unexplained change [in an agency’s position], or change that does not take account of legitimate reliance on prior interpretation, may be arbitrary, capricious [or] an abuse of discretion” (alteration in original) (citations omitted) (quoting 5 U.S.C. § 706(2)(A)) (internal quotation marks omitted)); Hatch v. Fed. Energy Regulatory Comm’n, 654 F.2d 825, 834-35 (D.C.Cir.1981) (holding that an agency’s sudden shift in the nature of proof required of the regulated party was not sufficiently explained and necessitated remand); Pub. Serv. Co. of Ind., Inc. v. Fed. Energy Regulatory Comm’n, 584 F.2d 1084, 1087-88 (D.C.Cir.1978) (holding that an agency’s sudden, unexplained shift in the kind of data that a regulated party was required to submit was arbitrary); Verizon Tel., 570 F.3d at 304 (“[I]t is arbitrary and capricious for the FCC to apply such new approaches without providing a satisfactory explanation when it has not followed such approaches in the past.”); see also Fed. Energy Regulatory Comm’n v. Triton Oil & Gas Corp., 750 F.2d 113, 116 (D.C.Cir.1984) (“The Commission may not abuse its discretion by arbitrarily choosing to disregard its own established rules and procedures in a single, specific case. Agencies must implement their rules and regulations in a consistent, evenhanded manner.”); cf. Fox Television, 556 U.S. at 518, 129 S.Ct. 1800 (“[T]he agency’s decision not to impose any forfeiture or other sanction precludes any argument that it is arbitrarily punishing parties without notice of the potential consequences of their action.”). *1229 However, this is not a case where a conclusion of arbitrary or capricious agency action is warranted. Two factors provide the central pillars supporting our view: first, the unique factual circumstances surrounding the issuance of the Phoenix Order, which should have provided Qwest with at least some (albeit limited) notice that a policy shift to a market-power analytical framework might be in the offing; and second, the Commission did not ignore the cut-the-cord phenomenon, but rather offered a reasonable explanation for why under the circumstances of this proceeding wireless voice services were excluded from the product market. The factual circumstances surrounding the issuance of the Phoenix Order are somewhat unusual. Qwest filed the earlier petition associated with the Qwest Four-MSA Order while Verizon’s appeal of the Verizon Six-MSA Order was pending. Upon review of the Verizon order, the D.C. Circuit disapproved the Commission’s unexplained policy shift toward an exclusive reliance upon a carrier’s market share, and it remanded the order. See Verizon Tel., 570 F.3d at 305. In the meantime, the Commission had issued its decision in the Qwest FourMSA Order. Because that order embraced an analytical approach similar to the Verizon order, the D.C. Circuit’s decision applied equally to it. That order, too, was remanded (at the Commission’s request). Both remands were an opportunity for the Commission to reconsider its reasoning in those orders and, more generally, its approach to forbearance petitions. The Commission sought comment on the analytical approach that it should adopt for the remanded orders. Among other things, the Commission sought input regarding whether it should “depart from its recent precedent regarding marketplace analysis in forbearance petitions,” and concerning the kinds of evidence beyond “market share for a particular product market” that are relevant in determining “whether forbearance from unbundling regulations is warranted.” Remand Comment Request, 24 FCC Rcd. at 10,883; see also Phoenix Order, 25 FCC Rcd. at 8632, ¶ 20 (discussing the regulatory history concerning forbearance). At the time, Qwest’s Phoenix-MSA petition was pending before the Commission. Although the request for comments in connection with the remanded orders was quite general, the Commission was more specific when it requested comments in connection with Qwest’s Phoenix-MSA petition in April 2010. There, the Commission indicated that it was reconsidering its analytical approach to forbearance petitions; that, specifically, it was pondering adopting a market-power approach as reflected in the FTC-DOJ Horizontal Merger Guidelines; that “[a] key element of the Guidelines, as they relate to product market definition, includes an inquiry into whether a hypothetical monopoly provider of a service [e.g., wireline voice service] profitably could impose a ‘small but significant and nontransitory’ increase in the price of such service,” Joint App. at 1199 n.4 (quoting FTC-DOJ Horizontal Merger Guidelines §§ 1. 11, 1.12); and the Commission invited Qwest to “cite specific evidence in the record or provide new data ... in response to these issues,” id. at 1199. The Commission, to be sure, did not commit to a definite policy shift in the April 2010 notice. And this more specific signal that a policy change might be coming came pretty late in the day. The notice was issued only a little more than two months before the Commission actually issued the Phoenix Order. However, the foregoing factual circumstances suggest to us that the Commission did not act whimsically or rashly in altering its for *1230 bearance policy — the D.C. Circuit remands gave the Commission a concrete reason to step back and assess its current policy direction — and that the Commission did provide Qwest with some notice (albeit of a temporally and substantively limited sort) that a policy move to a market-power approach might be in the offing, thus significantly diminishing the reliance that Qwest reasonably could have placed on the Commission’s previous policy stance. Therefore, we conclude that these unique circumstances counsel against a determination that the Commission acted arbitrarily or capriciously in moving the policy goalpost. 6 Furthermore, the Commission offered a reasonable explanation for its movement to a market-power analytical framework, which necessitated the production of qualitatively different evidence to warrant regulatory forbearance. As the Commission indicated, the market-power framework necessitated a more rigorous inquiry than the Commission had undertaken in prior orders (such as the Omaha Order), and part of the inquiry entailed a specific delineation of the relevant product market for evaluating competitive conditions in the Phoenix MSA. See Phoenix Order, 25 FCC Rcd. at 8646, ¶¶ 41-42. Further, the product-market analysis required more than a facile reliance on cut-the-cord percentages. See id. at 8654-55, ¶ 59. Rather, under the market-power rubric, the relevant question was the degree to which the cut-the-cord phenomenon materially constrained the prices that a wireline carrier like Qwest could charge its existing customers. Id. at 8651-52, ¶ 56. That in turn required an inquiry, not simply into the number of wireless-only customers, but into the cross-elasticity of demand between wireline and wireless services — that is, roughly speaking, the likelihood that customers can and will switch between the two forms of service in response to price changes. As to that inquiry, the Commission found that Qwest had offered no evidence. In sum, the Commission offered an extensive discussion of its reasons for abandoning the two-part test in the Omaha Order and for adopting the market-power approach — an approach with some basis in the Commission’s precedent and, in the Commission’s view, better in keeping with the underlying purposes of section 10. 7 *1231 See id. at 8633-45, ¶¶ 24-40. The Commission, therefore, was conscious of the change it was making, believed it to be better, explained why it was necessary, and offered a sound basis for repudiating its prior decisions. See Fox Television, 556 U.S. at 515-16, 129 S.Ct. 1800; cf. Verizon Tel., 570 F.3d at 304 (stating that “[t]he flaw” in the Commission’s policy shift was “not in the th[e] change, but rather in the [Commission’s] failure to explain it”). No doubt, the Commission did move the goalpost here, but it did so under somewhat unique circumstances and it “articulate[d] a satisfactory explanation” for doing so. Fox Television, 556 U.S. at 513, 129 S.Ct. 1800 (quoting State Farm, 463 U.S. at 43, 103 S.Ct. 2856) (internal quotation marks omitted). We conclude, therefore, that its decision was not arbitrary or capricious. 8 b We turn to Qwest’s second major challenge to the Phoenix Order. Qwest assails as unreasonable the Commission’s assessment of competitive conditions in the retail mass market in Phoenix, arguing that the Commission unreasonably declined to assess competition from wireless companies and unreasonably viewed the Phoenix market as an anti-competitive duopoly. We reject Qwest’s arguments. In the Phoenix Order, after delineating the product market and excluding mobile *1232 wireless services, the Commission went on to identify the participants in the retail mass market in Phoenix. See Phoenix Order, 25 FCC Rcd. at 8657, ¶ 67. It found that Qwest faced competition from Cox and a smattering of “fringe” competitors that were “able to compete only by relying extensively on UNEs and other Qwest wholesale services.” 9 Id. at 8664, ¶ 80 (internal quotation marks omitted). Qwest faults the Commission for ignoring competition from national wireless providers such as AT & T and Verizon Wireless. But the market participants that the Commission identified flowed logically from its definition of the relevant product market. Because it had determined that the product market excluded mobile wireless services — and because that exclusion was not arbitrary in this case — it followed that the pool of service providers in the Phoenix MSA did not include the national wireless companies. Furthermore, given the market dominance of Qwest and Cox in wireline services, it was proper for the Commission to treat the Phoenix MSA as effectively a duopoly. See id. (observing that “Cox is Qwest’s only competitor that now provides or is soon likely to provide retail service to mass market customers over its own last-mile network to any significant extent in the Phoenix MSA”). While the Commission acknowledged that “under certain conditions duopoly will yield a competitive outcome,” id. at 8637, ¶ 30, it observed that “[e]conomists, courts, and the Commission have long recognized that duopolies may present significant risks of collusion and supracompetitive pricing,” id. at 8636, ¶ 29 (footnotes omitted). Qwest counters that the Commission justified this conclusion only by reference to antitrust literature concerning “ordinary duopoly markets” and ignored the unique characteristics of the telecommunications industry: high fixed costs and low marginal costs. Qwest Reply Br. at 24; see Qwest Opening Br. at 51. Qwest points out that the Commission has found in previous orders that even a single new entrant who has successfully established extensive facilities coverage, such as Cox, has “demonstrated a deep commitment to compete vigorously for customers,” thus “lessening the need for regulatory intervention.” In the Matter of the Petition of ACS of Anchorage, Inc. Pursuant to Section 10 of the Communications Act of 1934, As Amended, for Forbearance from Sections 251(c)(3) and 252(d)(1) in the Anchorage Study Area, 22 FCC Rcd. 1958, 1977, ¶ 31, 2007 WL 256302 (2007); see also Omaha Order, 20 FCC Rcd. at 19,432, ¶ 33 (“The Commission has repeatedly found that residential customers are highly demand-elastic, and willing to switch to or from their provider to obtain price reductions and desired features.”). The Commission had good reasons, however, to call this prior finding into question, or at least to question the breadth of its applicability. In explaining why the test articulated in the Omaha Order was inadequate, the Commission noted that post-forbearance developments in the Omaha market gave it pause. See Phoenix Order, 25 FCC Rcd. at 8639-42, ¶¶ 33-36. Among other things, competition for retail mass-market services in Omaha declined following the Commission’s grant of forbearance to Qwest, even though the Omaha Order predicted the contrary outcome. In the Omaha Order, the Commission believed that forbearance would not *1233 cause Qwest to curtail competitor access to network elements. See 20 FCC Rcd. at 19,455, ¶ 81. In the Commission’s view, Cox’s aggressive competition would motivate Qwest to offer retail and wholesale services at competitive rates in order to minimize customer defections to Cox’s service. See id. That prediction, however, did not pan out. As the Commission noted in the Phoenix Order. [T]he Commission has long recognized that a vertically integrated firm with market power in one market — here upstream wholesale markets where, as discussed below, Qwest remains dominant — may have the incentive and ability to discriminate against rivals in downstream retail markets or raise rivals’ costs.... [T]here is little evidence, either in the record or of which we otherwise are aware, that ... incumbent LECs have voluntarily offered wholesale services at competitive prices once regulatory requirements governing wholesale prices were eliminated. For example, other than Cox, McLeod USA was the only other competitor of significant size cited by the Commission in the Qwest Omaha Forbearance Order. The record indicates that subsequent to the Qwest Omaha Forbearance Order, Qwest, with one exception, was not spurred to offer McLeodUSA any wholesale alternatives to UNEs that were not already offered prior to the grant of forbearance. Moreover, the record indicates that McLeodUSA has removed most of its employees from the Omaha marketplace, has limited its operations primarily to serving its existing customer base, and has ceased sales of residential and nearly all business services in Omaha. This suggests that McLeodUSA likewise no longer should be considered a significant competitor in the Omaha marketplace. We also note record evidence that Integra, which had been contemplating entry into the Omaha market, abandoned its plans to do so after the Commission issued the Qwest Omaha Forbearance Order. 25 FCC Rcd. at 8640-41, ¶ 34 (footnotes omitted). The Commission acknowledged that “multiple factors” could explain this decline and that it was “beyond the scope of this proceeding to estimate the extent of competition in Omaha today.” Id. at 8640 n. 105, 8641, ¶ 34. However, the Commission was entitled to consider this new data as probative of the soundness (or lack thereof) of its prior conclusions. See George E. Warren Corp. v. U.S. EPA, 159 F.3d 616, 626-27 (D.C.Cir.1998) (holding that it was not arbitrary and capricious for an agency to change course when “one of its basic assumptions in [earlier] rulemaking was contradicted by new data”), amended on other grounds by 164 F.3d 676 (D.C.Cir.1999). Finally, the Commission evaluated whether potential competition in the Phoenix MSA could ameliorate the risks of duopolistie price coordination. See Phoenix Order, 25 FCC Rcd. at 8666, ¶ 83. It observed that there were no competitors for mass-market services in Phoenix that had deployed or could deploy their own facilities to any meaningful degree, and that the potential for de novo entry by new competitors was limited. See id. at 8665-67, ¶¶ 82-84. Taking these factors together — specifically, the well-documented anticompetitive risks of duopoly, the subsequent developments in Omaha, and the lack of effective competition in the Phoenix market — the Commission could rationally call into question its earlier predictions, perceive the need for a different approach, and proceed cautiously regarding the possibility of granting forbearance in the Phoenix MSA, given the real-world understanding that doing so might result in a Qwest^Cox duo *1234 poly similar to the one in Omaha. Based upon the foregoing reasoning, we reject Qwest’s contention that the Phoenix Order is unreasonable. Ill Our task here is a “narrow” one. Fox Television, 556 U.S. at 513, 129 S.Ct. 1800 (quoting State Farm, 463 U.S. at 43, 103 S.Ct. 2856) (internal quotation marks omitted). We are not a “panel of referees on a professional economics journal,” but a “panel of generalist judges obliged to defer to a reasonable judgment by an agency acting pursuant to congressionally delegated authority.” City of L.A. v. U.S. Dep’t of Transp., 165 F.3d 972, 977 (D.C.Cir.1999). For the reasons articulated above, we find that the Phoenix Order was not “arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law.” Qwest Commc’ns, 398 F.3d at 1229 (quoting 5 U.S.C. § 706(2)(A)) (internal quotation marks omitted). Accordingly, we DENY Qwest’s petition for review. Footnotes [1] Some documents relevant to our disposition of this case contain competitively sensitive information (e.g., certain market-share data) that is the subject of FCC protective orders and not available in publicly accessible materials. In the publicly accessible versions of these documents, this sensitive information has been redacted. Although we have been furnished access to this sensitive information by the parties in sealed filings, we cite in this opinion to materials that do not contain this information — having no need to do otherwise, and to facilitate public access. Consequently, in some instances we do not supply precise figures, as they are redacted in the publicly accessible documents. [2] A “duopoly'' has been defined as "[a] market in which there are only two sellers of a product.” Black's Law Dictionary 577 (9th ed. 2009). [3] Qwest has challenged only that portion of the Phoenix Order pertaining to mass-market retail services. Qwest Opening Br. at 4 n.l. [4] Facilities-based voice-over-internet-protocol ("VoIP”) services are voice services offered by *1222 cable providers over their own networks. By contrast, over-the-top VoIP providers such as Vonage do not operate their own networks and instead require customers to obtain access from an incumbent carrier such as Qwest. See Phoenix Order, 25 FCC Rcd. at 8650 n. 163, ¶ 54. The Commission has traditionally viewed facilities-based VoIP services as "sufficiently close substitutes for local service to include them in the relevant product market.” Id. at 8650, ¶ 54. It has traditionally not viewed over-the-top VoIP the same way. See id. at 8650, ¶ 54 & n. 163. [5] The economic concept of elasticity of demand — more specifically, cross-elasticity of demand — is relevant in determining the appropriate scope of a product market. See, e.g., Brown Shoe Co. v. United States, 370 U.S. 294, 325, 82 S.Ct. 1502, 8 L.Ed.2d 510 (1962) ("The outer boundaries of a product market are determined by the reasonable interchangeability of use or the cross-elasticity of demand between the product itself and substitutes for it.”); United States v. Archer-Daniels-Midland Co., 866 F.2d 242, 246 (8th Cir.1988) ("The boundaries of the product market of a particular product can be determined by the reasonable interchangeability or cross-elasticity of demand between itself and possible substitutes for it.”). We have had occasion to discuss the concept in a helpful man *1223 ner: "The basic relevant product market test is 'reasonable interchangeability.’ Interchangeability may be measured by, and is substantially synonymous with, cross-elasticity. A market is elastic if demand goes down as price goes up. A market is cross-elastic if rising prices for product A cause consumers to switch to product B.” Telecor Commc’ns, Inc. v. Sw. Bell Tel. Co., 305 F.3d 1124, 1131 (10th Cir.2002) (citations omitted) (quoting SCFC ILC, Inc. v. Visa USA, Inc., 36 F.3d 958, 966 (10th Cir.1994)); see Westman Comm’n Co. v. Hobart Int'l, Inc., 796 F.2d 1216, 1221 (10th Cir.1986) (“Defining the relevant market first requires a determination of the product market. This inquiry necessitates an examination of which commodities are 'reasonably interchangeable by consumers for the same purposes.’ ” (quoting United States v. E.I. du Pont de Nemours & Co., 351 U.S. 377, 395, 76 S.Ct. 994, 100 L.Ed. 1264 (1956))). See generally A Dictionary of Economics 100 (John Black et al. eds., 3d ed. 2009) (defining "cross-price elasticity of demand” as the "[t]he ratio between the proportional change in demand for a good and the proportional change in the price of a different good” (emphases added)). Generally speaking, as relevant here, the Commission was inquiring into the extent to which the demand for access to wireless voice services would increase in response to an increase in the price of wireline voice services. [6] Indeed, in the Phoenix Order, the Commission expressly referred to this unique set of circumstances in justifying its policy shift: We acknowledge that the Commission in the Qwest 4 MSA Forbearance Order suggested that geographically disaggregated evidence of the percentage of voice subscribers that rely on mobile wireless only might demonstrate that Qwest was entitled to forbearance in the Phoenix MSA. See Qwest 4 MSA Forbearance Order, 23 FCC Red at 11745, para. 22. Since those statements were made, however, the D.C. Circuit remanded the Verizon 6 MSA Forbearance Order and instructed the Commission to provide a more complete economic analysis of its decision to deny forbearance. The D.C. Circuit, at the Commission’s request, also remanded the Qwest 4 MSA Forbearance Order. After the remands, the Wireline Competition Bureau issued a Pub-lie Notice in the present proceeding seeking new record data that might show that, under a traditional market power analysis, mobile wireless service is in the same relevant product market as mass market wire-line telephone service. Phoenix Order, 25 FCC Red. at 8654 n. 178, II 59. [7] It also bears mention that Qwest has not challenged the market-power approach as inconsistent per se with section 10. Furthermore, because "Congress did not prescribe a 'particular mode of market analysis' or otherwise dictate how the [Commission] must make predictive judgments 'within [its] field of discretion and expertise,' such as those required under § 10,” Verizon Tel., 570 F.3d at 304 (second alteration in original) (quoting EarthLink, Inc. v. FCC, 462 F.3d 1, 8, 12 (D.C.Cir.2006)), the Commission's construe *1231 tion of section 10 is entitled to deference, see Chevron, 467 U.S. at 843-44, 104 S.Ct. 2778 ("If Congress has explicitly left a gap for the agency to fill, ... [the agency’s] regulations [filling in that gap] are given controlling weight unless they are arbitrary, capricious, or manifestly contrary to the statute.”). [8] Qwest vigorously objects that the Commission misapplied antitrust concepts by relying on the fact that Qwest had not decreased its prices in the Phoenix area in response to the cut-the-cord phenomenon. See Phoenix Order, 25 FCC Rcd. at 8653, ¶ 58 (noting that Qwest did not "produce! ] any evidence that it has reduced prices for its wireline services ... in response to changes in the price of mobile wireless service.”). Qwest argues that the Commission's reliance on this observation is unreasonable because Qwest is a regulated entity that cannot decrease prices for some customers while maintaining them at current levels for others. Qwest suggests that, insofar as the Commission relies on proof of a price reduction in making its product-market determination, it erects an unreasonable and virtually insurmountable evidentiary hurdle for regulated entities like Qwest. However, we do not believe that the Commission has acted unreasonably. As noted in text supra, the Commission specifically identified other types of proof — besides evidence indicating that Qwest lowered its prices — that Qwest could have relied upon to support its position that wireline and wireless voice services should be in the same product market: (1) econometric studies that "estimate the cross-elasticity of demand between mobile wireless and wireline access services”; (2) documentation revealing that Qwest has "adjusted its marketing for wireline service in response to changes in the price of mobile wireless service”; or (3) "marketing studies that show the extent to which consumers view wireless and wireline access services as close substitutes.” Phoenix Order, 25 FCC Rcd. at 8653-54, ¶ 58. Thus, the Commission explicitly left open the door for Qwest to demonstrate, through various forms of evidence, that wireless voice services belong in the same product market as traditional wireline voice services. We would expect, therefore, that with persuasive market surveys or econometric studies on substitutability, Qwest’s failure to lower its prices would not, of itself, be an insurmountable hurdle — viz., standing alone, it would not preclude the Commission from concluding that wireless voice services should be in the same product market as Qwest’s wireline voice services. Finally, we cannot conclude that it is necessarily unreasonable for the Commission to consider Qwest’s pricing behavior. See id. at 8653 n. 175 (citing a Department of Justice study reasoning that an incumbent carrier would lower prices in response to customer cord-cutting "if the loss of revenue from cord-cutting is expected to be greater than the loss of revenue from reducing the fees paid by customers who would not switch”). [9] As the Commission explained, "[a] fringe competitor is a small firm operating in a market that is dominated by a single firm or a few firms. The fringe competitors take the price set by the dominant firm(s) as given and maximize their profits given this price.” Phoenix Order, 25 FCC Rcd. at 8664 n. 241, ¶ 80.