OPINION OF THE COURT
HARDIMAN, Circuit Judge.
This case involves the regulation of electromagnetic spectrum by the Federal Communications Commission. In 1993, Congress amended the Communications Act of 1934, 47 U.S.C. §§ 151-622, to allow the FCC to grant spectrum licenses through a system of competitive bidding. 47 U.S.C. § 309(j)(1), (3). The Act requires the FCC to pursue certain objectives required by statute, including
47 U.S.C. § 309(j)(3)(B). In FCC parlance, these groups are known as designated entities (DEs). 47 C.F.R. § 1.2110(a). At this time, the FCC's "principal means of fulfilling the statutory objectives for DEs" is to confer bidding credits upon small and rural businesses that participate in FCC auctions. Updating Competitive Bidding Rules, 80 Fed. Reg. 56764, 56766 (September 18, 2015). Bidding credits operate as a discount on the spectrum DEs purchase, allowing them sometimes to outbid companies that make higher bids. Council Tree Comm'ns, Inc. v. FCC, 619 F.3d 235, 239 (3d Cir. 2010) (Council Tree III).
The question presented here is whether the FCC acted legally when it limited the bidding credits available to DEs. We hold that it did.
In 2014, the FCC began a rulemaking proceeding in advance of a special 2016-17 Incentive Auction of "scarce low-band radio spectrum." FCC Br. 13. According to its Notice of Proposed Rulemaking, the FCC thought it appropriate to
29 FCC Rcd. 12426 (2014), 2014 WL 5088195 at *7. The FCC later indicated in April 2015 that it was specifically considering a proposal to cap available DE credits within "any given auction" in order to "ensure that DEs cannot acquire spectrum in a manner that is wildly disproportionate to the concept of a small business." Request for Further Comment, 30 FCC Rcd. 4153, 4158 (2015) (citations omitted).
On July 21, 2015, the FCC concluded its rulemaking and released a final Report and Order entitled In the Matter of Updating Part I Competitive Bidding Rules, et al., 30 FCC Rcd. 7493 (2015) (hereinafter Order), published at 80 Fed. Reg. 56764 (Sept. 18, 2015). In the Order, the FCC issued a series of new rules, some designed to assist DEs and others intended to rein in perceived abuses of the DE program. New rules designed to assist DEs included: "greater flexibility" as to certain eligibility requirements, 30 FCC Rcd. at 7504, the creation of a "new bidding credit for eligible rural service providers," id. at 7521, and raising the revenue ceiling to qualify for DE credits. New measures intended to curb abuses included: revenue attribution rules designed to "restric[t] certain large carriers or companies from . . . exercising control over a DE," id. at 7511, limitations on joint bidding arrangements, and the rule at issue in this case: caps on bidding credits.
Though the FCC determined generally that small business credits should be capped in future auctions, it did not mandate a particular dollar value for those caps. Rather, it determined that any future cap would be at least $25 million per auction. In doing so, the FCC noted that most DEs in three recent auctions would not have qualified for more than $25 million in bidding credits. The FCC also determined that the $25 million minimum would "allow bona fide small businesses" to participate meaningfully in auctions, particularly "taking into account the changes we make today to increase a DE's flexibility in other respects." Id. at 7541. Looking to the special 2016-17 Incentive Auction, the FCC decided to cap bidding credits at $150 million (well over the $25 million minimum) because of the "significant difference in value between low-band and higher-band spectrum." Id. at 7545. Based on data from prior auctions,
On November 13, 2015, Appellant Council Tree—a DE which had opposed caps during the rulemaking—filed a petition in this Court for review of the Order, claiming violations of the Administrative Procedure Act and the Communications Act.
We have jurisdiction over Council Tree's petition under 47 U.S.C. § 402(a) and 28 U.S.C. § 2342(1). We review "agency action, findings, and conclusions" to determine whether they are "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." 5 U.S.C. § 706(2). Agency action is not arbitrary and capricious when the agency "examine[d] the relevant data and articulate[d] a satisfactory explanation for its action including a rational connection between the facts found and the choice made." Council Tree III, 619 F.3d at 250 (quoting Motor Vehicle Mfrs. Ass'n of U.S. v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43 (1983)).
Council Tree seeks review of the FCC's Order for three reasons. First, it claims the FCC's explanation of its new rule ignored its statutory obligation to promote competition and avoid excessive concentration of licenses. Second, it argues the imposition of any cap on bidding credits was arbitrary and capricious because the FCC lacked evidence that the DE designation was being abused. Third, it contends the FCC set both its general minimum cap and the specific Incentive Auction cap in an arbitrary and capricious manner because these particular caps are unsupported by the data. We examine each argument in turn.
In designing its competitive bidding process, the FCC is bound by statute to "promot[e] economic opportunity and competition . . . by avoiding excessive concentration of licenses." 47 U.S.C. § 309(j)(3)(B). Council Tree claims the FCC violated its obligation to promote these two objectives by not "considering the caps' anti-competitive effects" or "explaining the agency's reasoning" in how it balanced concern for competition against its other statutory objectives. Council Tree Br. 28.
In Council Tree III, we considered a similar challenge to the FCC's imposition of a rule restricting the ability of DEs to lease or resell their spectrum capacity, which also was aimed at addressing abuse of the DE designation. See Council Tree III, 619 F.3d at 251. As part of its challenge in that case, Council Tree claimed the FCC had violated its statutory duty to promote "economic opportunity and competition . . . by avoiding excessive concentration of licenses and by disseminating licenses among a wide variety of applicants"— since such a rule was likely to impede DEs' profits and growth. Id. at 249 n.7 (alteration in original) (quoting 47 U.S.C. § 309(j)(3)(B)).
We rejected that argument. We noted the FCC's new policy served other statutory objectives, including its obligation to "recover a portion of the value of the spectrum and prevent unjust enrichment." Id. (citing 47 U.S.C. § 309(j)(3)(C)). Citing both "the general agreement that the DE program can be abused, [and] the continuing participation by DEs in auctions held under the new rules," we could not "conclude that the FCC has failed to promote small-business participation at all." Id. We also recognized that the administrative "record reflect[ed] the FCC's cognizance of the capitalization issue, and that it engaged in a line-drawing exercise in an attempt to prevent unjust enrichment without unduly impairing DEs' capital access." Id. at 251-52.
The FCC's Order at issue in this appeal did more than take cognizance of the issue of DEs' competitiveness. It reviewed data from past auctions before concluding that the cap would not significantly impair the ability of DEs (in the aggregate) to compete in auctions. See, e.g., 30 FCC Rcd. at 7541 (explaining the cap floor with reference to prior DE auction participation); id. at 7545-46 (justifying the $150 million cap with reference to past auction results and pricing in different markets). The FCC also noted its bidding credit cap was paired with regulations that increased flexibility for DEs seeking to obtain capital, bolstering its conclusion that DEs would still be able to compete in the spectrum license market. See id. at 7541 (referring to the adjusted policy of when an investor's revenue will be attributed to a DE, detailed at 30 FCC Rcd. 7502-21). The FCC therefore not only set forth a policy that is likely to allow continued participation by DEs, but also rationally explained why it expected no significant loss of DE participation. As such, under the standard we articulated in Council Tree III, the FCC did not fail to consider DEs' ability to compete for licenses.
Council Tree responds that the statements just noted did not relate to the FCC's obligations to promote competition or avoid excessive concentration of licenses at all. Rather, they related only to the FCC's obligation to "promote economic opportunity" for DEs. Council Tree Br. 37. This strikes us as an artificial distinction. The relevant statutory language (47 U.S.C. § 309(j)(3)(B)) links these goals together because they are inextricable. When the FCC helps small businesses compete in the marketplace against large telecommunications providers, it necessarily increases competition for licenses and reduces license concentration.
What Council Tree seems to suggest is that an appropriate analysis by the FCC would not have assessed only whether DEs could continue to participate at near-current levels, but also would have considered whether their current level of competition is adequate to challenge our telecommunications quadropoly. See, e.g., Council Tree Br. 15 & n.31 (referring to the "Big Four" of AT&T, Verizon, Sprint, and T-Mobile); id. at 37 (explaining that "the caps' effect on the amount of spectrum," rather than the number of DEs affected by the cap, is "far more relevant" to assessing competition); id. at 38 (using auction data to show a mid-level DE capped at the minimum of $25 million in bidding credits would more than exhaust its credits to win a single license in most major markets). The problem with this approach is that it adds requirements to the statute not found in the text. The FCC's statutory obligation is to ensure that its bidding system promotes competition and the broad dissemination of licenses, while also abiding by its obligations to, inter alia, "recove[r] for the public a portion of the value" of the spectrum licenses sold, avoid "unjust enrichment," and ensure efficient use of spectrum. 47 U.S.C. § 309(j)(3)(B)-(D). As much as Council Tree would like the FCC to not merely promote, but maximize, competition against the Big Four, § 309 creates no such obligation. And insofar as Council Tree comes close to defining "competition" as "competition to [the market share of] incumbents in large urban markets, or on a regional or national scale," Council Tree Br. 37, we will not insert this limiting language into the statute. See Alabama v. North Carolina, 560 U.S. 330, 352 (2010) ("We do not—we cannot—add provisions to a federal statute.")
The FCC's Order: (1) preserved a significant bidding credit program; (2) reviewed data suggesting DE participation would continue despite the proposed caps; and (3) altered other rules to make DEs more competitive. We therefore "cannot conclude that the FCC has failed to promote small-business participation at all," Council Tree III, 619 F.3d at 249 n.7, or that the Order at issue in this case failed to promote the FCC's dual objectives of competition and reduced concentration of licenses, see 47 U.S.C. § 309(j)(3)(B).
Council Tree argues that even if the FCC considered its statutory objectives under 47 U.S.C. § 309(j), it violated 5 U.S.C. § 706 by arbitrarily and capriciously imposing caps on bidding credits. According to Council Tree, the FCC was chasing a phantom because its Order provides no evidence that DEs are "gaming the rules" or that DE investors are "unjustly enriched." Council Tree Br. 28-29.
It is true that the FCC's Order did little to assess the scope and substantiality of the harm posed by large companies abusing the DE process. See 30 FCC Rcd. at 7540-41 (discussing concern for "discourag[ing] entities that seek to game the Commission's rules at taxpayer expense" without estimating that expense). However, as we stated in Council Tree III, there is at least "general agreement that the DE program can be abused" by larger companies. 619 F.3d at 249 n.7. Council Tree itself acknowledges, though it minimizes, complaints made to the FCC about one telecommunications company—DISH—abusing the DE program in a recent auction by allegedly exercising control over DE bidders. And the FCC articulated a concern regarding increased incentives for abuse, noting "that, as the cost of spectrum continues to grow, the incentives for structuring transactions to obtain bidding discounts increase significantly." 30 FCC Rcd. at 7540.
In reviewing the FCC's explanation for imposing its caps, we do not ask whether it would persuade us to impose the same policy. We instead consider whether the agency "examine[d] the relevant data and articulate[d] a satisfactory explanation for its action including a rational connection between the facts found and the choice made." Council Tree III, 619 F.3d at 250 (quoting State Farm, 463 U.S. at 43).
Here, the FCC identified a threat (abuse of the DE designation) to one of its statutory objectives (preventing unjust enrichment), and adopted a prophylactic measure. We see no reason to bar such measures as a general matter. See Stillwell v. Office of Thrift Supervision, 569 F.3d 514, 519 (D.C. Cir. 2009) ("[A]gencies can, of course, adopt prophylactic rules to prevent potential problems before they arise.").
It would, of course, be preferable to have some estimate of the unjust enrichment that would have been anticipated had the FCC not capped bidding credits. And perhaps a prophylactic rule could be irrational if an agency (1) made no effort to measure its benefits, and (2) its impact on other statutory obligations was also unassessed (or negative). See, e.g., Prometheus Radio Project v. FCC, 652 F.3d 431, 469-72 (3d Cir. 2011). But as explained in Part III-A, supra, the FCC articulated a rational explanation, based on relevant data, as to why it believed its bidding credit caps would have only a minor impact on DEs' ability to compete in the spectrum license marketplace.
Though Council Tree objects that the FCC did not truly balance its obligations here, Council Tree III is again instructive. In that case we found the question of whether the rule restricting spectrum leasing was arbitrary and capricious "a close one" on first review, because "the FCC made few factual findings on the impact of the new rules on DE financing." Council Tree III, 619 F.3d at 251. But despite the lack of concrete findings, we found "enough consideration of [the negative impact on] DE capitalization to pass the arbitrary and capricious threshold." Id. at 253.
We reached this conclusion for two reasons. First, the FCC's notice of proposed rulemaking "reflect[ed] the FCC's cognizance of the capitalization issue" based on its stated concern for a "delicate balance" between avoiding abuse and allowing small businesses to access flexible sources of capital. Id. at 251-52 (citation omitted). Second, we noted that predictions about the future impact of rules were "inherently speculative." Id. at 252. Under our "necessarily deferential" review of "line-drawing determinations," id. at 250-51 (citation omitted), we found it unnecessary to demand more despite the FCC's consideration being "neither as clear nor as thorough as would be ideal," id. at 252-53.
Unlike the rulemaking in Council Tree III, the FCC in this matter made factual findings on the likely impact of this new rule on DE auction competitiveness. Whether the FCC is right or wrong that the impact will be minimal, we cannot say its prediction is irrational based on the relevant data of record. The FCC's balancing here was not arbitrary and capricious.
Finally, we consider Council Tree's challenge to the specific caps imposed. Council Tree raises three related arguments contesting the rationality of these caps, none of which we find persuasive.
First, Council Tree complains that focusing on the number of affected DEs, instead of on their ability to break into major markets, is a poor way to measure the impact of a cap on competition. As discussed, we review the FCC's reasoning to confirm that it used relevant data—not that it applied the most rigorous analysis available. Here, the FCC reviewed relevant auction data before evaluating whether the caps on bidding credits would reduce competition.
Second, Council Tree argues that DEs were so hampered by old rules in two of the auctions that those auctions should not have been used in estimates, and that even the third (ostensibly adequate) auction cannot be compared to the "once-in-a-generation" Incentive Auction. Council Tree Br. 58 (quoting Order ¶ 95). Here again, we review only for the use of relevant, not perfect, data. Council Tree III, 619 F.3d at 250. And Council Tree has not offered, in briefing or at argument, any alternative data that the FCC should have considered in its estimates.
Finally, Council Tree argues that the FCC's justifications in support of the $25 million minimum on bidding credit caps and its $150 million cap for the Incentive Auction would be as consistent with limits of "$200 million, $500 million, or more," making its explanation inadequate. Council Tree Br. 58. In support of this argument, Council Tree points to Bluewater Network v. EPA, where the Court of Appeals for the D.C. Circuit ordered the EPA to clarify the basis of its determination that only 70% of new snowmobiles in eight years could be fitted with emission reduction technology. 370 F.3d 1, 21 (D.C. Cir. 2004). The Bluewater court held the EPA's statements regarding cost concerns for manufacturers—that design takes time and manufacturers have finite resources—were too vague. See id. ("The Agency's explanation of its reasoning could just as well support standards corresponding to 30% or 100% application in that time frame."). By contrast, the FCC's analysis here would not be consistent with a bidding credit cap of any value. For example, the FCC rejected caps of $10 million and $25 million for the Incentive Auction as less consistent with historical bidding thresholds. While the FCC did not explicitly reject any upper boundary as too high, we note that it set the $150 million cap such that, while "nearly all of the small businesses that claimed bidding credits . . . would have fallen under" the cap, 30 FCC Rcd. at 7545, two successful DEs in one of the previous auctions would have exceeded it. Council Tree admits those two DEs, allegedly "financed in large part by investments from DISH," prompted the concern that led to this rulemaking, allowing us to infer grounds for an upper bound in the record. Council Tree Br. 20. Even excluding that inference, however, the sources underlying the decision here are much more concrete than those in Bluewater. Besides, setting a reasonable cap is a quintessential "line-drawing determination" calling for "necessarily deferential" review in this Court. Council Tree III, 619 F.3d at 250-51 (citation omitted).
In sum, the FCC made a "rational connection" between "relevant data" and the caps on bidding credits under review. State Farm, 463 U.S. at 43.
For the reasons stated, we hold that the FCC's Order dated July 21, 2015 and published on September 18, 2015 was not arbitrary, capricious, an abuse of discretion, or otherwise contrary to law. We will deny Council Tree's petition for review.