Safe Extensions

509 F.3d 593 (2007) | Cited 38 times | D.C. Circuit | December 11, 2007

Argued October 19, 2007

On Petition for Review of an Order of the Federal Aviation Administration.

Before: HENDERSON, TATEL, and KAVANAUGH, Circuit Judges.

In this case, a company argues that the Federal Aviation Administration arbitrarily and capriciously imposed a strict test on its product but not on other, similar products. The FAA responds with a laundry list of reasons why this court supposedly cannot hear this challenge. The FAA then argues that even if we may hear the case, substantial evidence supports its decision. Because the FAA's jurisdictional arguments are wholly meritless and because the agency offers nothing more to justify its decision than one employee's bare assertions unsupported by any actual evidence, we grant the petition for review.


Federal law authorizes the Federal Aviation Administration to "prescribe minimum safety standards for . . . operating an airport serving any passenger . . . aircraft designed for at least 31 passenger seats," 49 U.S.C. § 44701(b), and directs the agency to "promote safe flight of civil aircraft . . . by prescribing . . . regulations and minimum standards for . . . practices, methods, and procedure[s] the [agency] finds necessary for safety in air commerce," id. § 44701(a). Invoking this authority, the FAA issues "advisory circulars" that establish testing requirements and product specifications for a range of items airports use. Private labs test products to determine whether they comply with FAA standards, and the FAA then publishes a list of approved items. To obtain federal funding, airports must agree to buy only products on this list. Thus, airports receiving federal funding-"virtually every public use airport," Resp't's Br. 3-are prohibited from buying unapproved products.

Among the many products the FAA regulates are lights that line airport runways and the metal bases in which those lights sit. Light bases used in runways must be very strong because they experience powerful forces-such as airplanes landing on them and snow plows driving over them-and must retain their original alignment to illuminate the runway properly. Also, because runways are repaved frequently, the height of the bases must be adjustable so they can remain flush with the runway's surface.

Petitioner Safe Extensions, Inc., manufactures one type of light base used in airport runways. The FAA calls the light bases Safe Extensions produces "adjustable products." Other manufacturers make a competing technology that the FAA calls "fixed products." With both technologies, the base is placed in a hole in the runway and secured with concrete or special grout. The two technologies differ in the mechanisms used to adjust their height. Adjustable products-the ones Safe Extensions manufactures-have an extension piece at their top; the bottom of the extension piece is threaded, as is the top of the base. To adjust the height, the extension piece is twisted upwards. By contrast, fixed products are made taller by stacking linked extensions on top of the base.

The FAA has long required runway light bases to pass a torque test, which checks whether the base can withstand a strong force without rotating. From 1970 to 2005, this torque test applied to all light bases, both adjustable and fixed, and the test was conducted on light bases installed in the concrete or grout that secured them. In April 2005, however, the FAA issued Advisory Circular 42D, which specified that torque testing would only be required for "bases that utilize a method of height adjustment that is integral to the base or extension and are designed for field adjustment." Advisory Circular 150/5345-42D: Specification for Airport Light Bases, Transformer Housings, Junction Boxes, and Accessories ¶¶, 4.3.10 (Apr. 29, 2005). The FAA apparently intended this language to mean that only adjustable products, not fixed products, had to pass the torque test. Thus, though the torque test itself remained unchanged, only adjustable products now had to pass it.

Thirteen months later, the FAA issued Advisory Circular 42E (AC-42E), which made the required torque test far more stringent. Advisory Circular 150/5345-42E: Specification for Airport Light Bases, Transformer Housings, Junction Boxes, and Accessories (May 8, 2006). AC-42E specified that torque testing of adjustable products now had to be conducted on freestanding light bases, i.e., light bases not yet embedded in concrete or grout the way they would be in the field. Id. at note after ¶ 4.1.7, ¶ 4.3.10. According to Safe Extensions, this caused an outcry among companies that install adjustable products, with several complaining to the FAA that the new test "simply won't work" and "essentially eliminates the use of" adjustable products. Oral Arg. at 3:25. After talking to the FAA, some of these companies were left with the impression that the agency would address their concerns by issuing a revised advisory circular. Id. at 2:04, 3:34.

Just two months after issuing AC-42E, the FAA emailed a draft Advisory Circular 42F to a few companies that make or install runway light bases, though not Safe Extensions. Email from David Evans de Maria, FAA Airport Engineering Division, to David Edwards et al. (Aug. 2, 2006). The torque testing requirements in the draft circular, however, were identical to those in AC-42E; that is, the draft required freestanding torque tests for adjustable products but no torque test for fixed products. Draft Advisory Circular 150/5345-42F: Specification for Airport Light Bases, Transformer Housings, Junction Boxes, and Accessories ¶¶ 4.1.7, 4.3.10 (July 31, 2006).

Two companies offered comments criticizing the draft of AC-42F. Olson Industries, which installs both adjustable and fixed products, argued that adjustable products could never pass the freestanding torque test and that the circular was unfair because fixed products, if subjected to the test, couldn't pass it either. Email from Ted Olson, Jr., President, Olson Industries, to David Evans de Maria, FAA Airport Engineering Division (Aug. 16, 2006). The company also complained that the revised test was unjustified because it had never seen a problem with an installed adjustable product in its twenty years of experience. Id. Another installer of runway lights, Siemens Airfield Solutions, made exactly the same points. Email from Willis Trainor, Certification Test Engineer, Siemens Airfield Solutions, to David Evans de Maria, FAA Airport Engineering Division (Aug. 16, 2006).

At some point before the FAA issued the final version of AC-42F, an agency employee prepared a response to these comments. Because this response is the only justification the agency has provided for its decision, we quote it in full:

REJECTED -- Fixed [products] have anti-rotational devices inherent to the physical structure of the device itself that act in bearing against the surrounding grade. Adjustable [products] in known currently available designs have no physical features inherent to the device itself to prevent rotation, and rely on a chemical bond acting in shear. There are known examples where this bond has failed in actual application following exposure to real world loading and environmental conditions. Torque testing of specimens prepared in laboratory conditions may be able to pass torque tests simulating being mounted in surrounding grade, but experience has shown that the same care and attention is not possible to control in the field. Experience has also shown that required production testing of in situ torque tests designed to give a certain level of confidence are not performed. As adjustable height extensions have no inherent anti-rotational physical feature that bears against the surrounding grade it must demonstrate its anti-rotational capability free standing so that the uncontrollable quality of its installation is not relied upon for public safety.

Comment Resolutions for Draft Advisory Circular 150/5345-42F: Specification for Airport Light Bases, Transformer Housings, Junction Boxes, and Accessories 1-2 (Sept. 5, 2006). This response appears on a plain sheet of paper in the agency's appendix to its brief. The response says nothing about when or even if the FAA shared it with anyone. At oral argument, Safe Extensions's counsel told us that the FAA emailed the response to selected companies at approximately the same time the agency published the final version of AC-42F. Oral Arg. at 7:40, 8:15. According to Safe Extensions, it never received the response directly from the FAA and had no opportunity to address it before the final circular was published. Id. at 7:40. The FAA's counsel disputed none of this.

The FAA issued the final version of AC-42F on October 17, 2006, making no changes from the draft circular regarding the torque testing required for adjustable products used in runways. Advisory Circular 150/5345-42F: Specification for Airport Light Bases, Transformer Housings, Junction Boxes, and Accessories ¶ 4.3.10 (Oct. 17, 2006). Thus, adjustable products remained subject to a freestanding torque test fixed products did not have to pass. On December 15, Safe Extensions filed a petition for review of AC-42F pursuant to 49 U.S.C. § 46110(a), which allows any "person disclosing a substantial interest in an order issued by the . . . Federal Aviation Administration with respect to aviation duties and powers designated to be carried out by the [agency] . . . [to] apply for review of the order by filing a petition for review in the United States Court of Appeals for the District of Columbia Circuit." Safe Extensions argues that the FAA acted arbitrarily and capriciously by imposing a freestanding torque test on adjustable products but not fixed products. The FAA responds that this court lacks jurisdiction to review Safe Extensions's challenge and that, in any event, it acted reasonably when it decided to treat the two products differently. We address the jurisdictional issues first and then consider whether the FAA acted arbitrarily and capriciously.


The FAA argues that we lack jurisdiction to hear this case for four independent reasons: (1) AC-42F is not an "order" reviewable under 49 U.S.C. § 46110; (2) Safe Extensions lacks prudential standing; (3) AC-42F addresses issues committed by law to agency discretion; and (4) because the differential treatment Safe Extensions objects to originated in AC-42E, not AC-42F, the company's petition is untimely given that the company filed it more than sixty days after the FAA issued AC-42E. We address each argument in turn.

Not an Order

The Administrative Procedure Act, 5 U.S.C. § 551 et seq., defines an "order" as "the whole or a part of a final disposition . . . of an agency in a matter other than rulemaking." Id. § 551(6) (emphasis added). To be deemed "final" and thus reviewable as an order under 49 U.S.C. § 46110, an agency disposition "must mark the 'consummation' of the agency's decisionmaking process," and it "must determine 'rights or obligations' or give rise to 'legal consequences.'" City of Dania Beach v. FAA, 485 F.3d 1181, 1187 (D.C. Cir. 2007) (quoting Vill. of Bensenville v. FAA, 457 F.3d 52, 68 (D.C. Cir. 2006)). As a general principle, "the term 'order' in [section 46110] should be read expansively." Id.

The FAA first argues that AC-42F "neither imposes a legal obligation upon any person nor creates any legal rights." Resp't's Br. 15. This is absurd. As Safe Extensions points out, "if a manufacturer's equipment does not meet AC-42F's specifications, it cannot be" placed on the list of FAA-approved products airports can buy. Pet'r's Opening Br. 22. Thus, AC-42F effectively prohibits airports from buying light bases that fail the new torque test, and it bars manufacturers like Safe Extensions from selling their products to airports. These are clear legal consequences of enormous significance to Safe Extensions.

Nothing in Aerosource, Inc. v. Slater, 142 F.3d 572 (3d Cir. 1998), upon which the FAA relies, suggests a different result. There the FAA had simply issued a warning about mistakes Aerosource made in its repair work. See id. at 575-76; see also id. at 581 ("[T]his case concerns nothing more than the issuance of advisory warnings and the FAA's refusal to withdraw the warnings."). The warning neither barred anyone from contracting with the company nor barred the company from continuing to work. See id. at 581. Thus, while the warning created bad publicity for the company and harmed its finances, it imposed no legal obligations on it or anyone else. By contrast, AC-42F has the obvious legal consequences mentioned above.

Next, the FAA argues that to qualify as an order, an agency decision must not only be final, but also "be accompanied by a record sufficient to permit judicial review," Resp't's Br. 16-a record the FAA claims AC-42F lacks. This argument ignores our cases interpreting section 46110. In both Dania Beach and Bensenville we held that agency actions are reviewable as orders under section 46110 so long as they are final, i.e., so long as they mark the consummation of the agency's decisionmaking process and determine rights or obligations or give rise to legal consequences. 485 F.3d at 1187; 457 F.3d at 68.

Ignoring these recent and controlling cases, the FAA points to our decision over thirty years ago in Deutsche Lufthansa Aktiengesellshchaft v. Civil Aeronautics Board, 479 F.2d 912 (D.C. Cir. 1973), where, reviewing an earlier version of section 46110, we said: "It is the availability of a record for review and not the holding of a quasi[-]judicial hearing which is now the jurisdictional touchstone." Id. at 915. The FAA has taken this statement entirely out of context. In Lufthansa, we considered whether we were bound by United Gas Pipe Line Co. v. Federal Power Commission, 181 F.2d 796 (D.C. Cir. 1950), in which "this court held that it had no jurisdiction over direct appeals from the promulgation of agency regulations where there had not been an evidentiary record established in a quasi[-]judicial proceeding before the agency." Lufthansa, 479 F.2d at 915. The key question in Lufthansa, then, was whether we could review an agency's action despite the absence of a quasi-judicial proceeding. We held that we could, at least when an evidentiary record existed. Id. at 916. Thus, in Lufthansa we set forth a sufficient condition for review when no quasi-judicial proceeding had occurred, not a necessary condition for reviewing agency action. And now that we have rejected the United Gas rule entirely, see Inv. Co. Inst. v. Bd. of Governors of Fed. Reserve Sys., 551 F.2d 1270, 1276-77 (D.C. Cir. 1977), this sufficient condition is irrelevant because there is no longer any doubt about whether we may review agency actions when the agency held no hearing.

Several factors confirm this reading of Lufthansa. First, as noted above, our recent cases regarding whether agency actions qualify as orders never consider the adequacy of the record, instead asking only whether the action was final. See Dania Beach, 485 F.3d at 1187; Bensenville, 457 F.3d at 68. Second, the statement in Lufthansa that the FAA relies on is a dead letter: though we cited it a few times after Lufthansa, see, e.g., City of Rochester v. Bond, 603 F.2d 927, 933 n.26 (D.C. Cir. 1979), and a few other circuits have picked it up, see, e.g., Sierra Club v. Skinner, 885 F.2d 591, 593 (9th Cir. 1989), Sima Prods. Corp. v. McLucas, 612 F.2d 309, 313 (7th Cir. 1980), we haven't cited it for decades, most likely because the demise of the United Gas rule has rendered Lufthansa's reasoning irrelevant. Third, the FAA's proposed reading of Lufthansa is precluded by subsequent Supreme Court cases making clear that the lack of an adequate agency record to review does not eliminate a circuit court's jurisdiction, but rather requires the court to remand to the agency so the agency can provide a record for review. For example, in Florida Power & Light Co. v. Lorion, 470 U.S. 729 (1985), the Court rejected the argument that if "the reviewing [circuit] court [lacks] an adequate agency-compiled factual basis to evaluate the agency action" it should require the case to be brought in "a district court with factfinding powers [that] could make up that deficiency," id. at 743, explaining:

Such a [result] cannot . . . be squared with fundamental principles of judicial review of agency action. "[T]he focal point for judicial review should be the administrative record already in existence, not some new record made initially in the reviewing court." Camp v. Pitts, 411 U.S. 138, 142 (1973). The task of the reviewing court is to apply the appropriate APA standard of review to the agency decision based on the record the agency presents to the reviewing court. If the record before the agency does not support the agency action, if the agency has not considered all relevant factors, or if the reviewing court simply cannot evaluate the challenged agency action on the basis of the record before it, the proper course, except in rare circumstances, is to remand to the agency for additional investigation or explanation.

Id. at 743-44 (some citations omitted). Fourth, adopting the FAA's additional proposed requirement for an agency act to count as an order would create perverse incentives. If an agency action qualified as an order only when accompanied by a sufficient record to permit review, agencies could escape judicial review by simply refusing to create a record to support their decisions. This cannot be the law. Finally, directly contradicting its argument here, the FAA has itself contended elsewhere that section 46110 applies even when the agency fails to provide an adequate record for review. For example, in Dania Beach the FAA argued that for an act to qualify as an order under section 46110, it need only "mark the consummation of the agency's decisionmaking process" and have legal consequences. Resp't's Br. 16, Dania Beach, 485 F.3d 1181 (No. 05-1328) (quoting Bennett v. Spear, 520 U.S. 154, 178 (1997)). Similarly, in Gilmore v. Gonzales, 435 F.3d 1125 (9th Cir. 2006), the FAA flatly stated: "the presence of an administrative record is not required for § 46110 to apply." Defs.-Appellees' Br. 28, Gilmore, 435 F.3d 1125 (No. 04-15736).

Even if we accepted the FAA's argument that actions qualify as reviewable orders only when supported by an adequate factual record, however, the record here, though exceedingly thin, is sufficient to permit review. The FAA assures us that its appendix "includes all the documents upon which the agency relied." Resp't's Br. 17. Although it claims that the documents "cannot reflect the entirety of the information and experience that the agency brought to bear in issuing the Circular," id., we have never required that much information from agencies-we just need enough to determine whether the agency's decision was arbitrary. If the FAA's documents fail to demonstrate the reasonableness of its decision, it means that the agency either has chosen not to write down the reasons for its decision or is unable to do so. Neither possibility is acceptable under the Administrative Procedure Act.

Prudential Standing

"To establish prudential standing, a party's 'grievance must arguably fall within the zone of interests protected or regulated by the statutory provision . . . invoked in the suit.'" Nuclear Energy Inst., Inc. v. EPA, 373 F.3d 1251, 1266 (D.C. Cir. 2004) (quoting Bennett, 520 U.S. at 162). The basic question is: "who may and who may not invoke the power of the courts to enforce the terms of a statute[?]" Hazardous Waste Treatment Council v. Thomas, 885 F.2d 918, 921-22 (D.C. Cir. 1989). In this case, Congress has already answered this question in section 46110(a): "a person disclosing a substantial interest in an order issued by the . . . Federal Aviation Administration with respect to aviation duties and powers designated to be carried out by the Administrator . . . may apply for review of the order." See also Int'l Bhd. of Teamsters v. Transp. Sec. Admin., 429 F.3d 1130, 1134 n.3 (D.C. Cir. 2005) (equating prudential standing under section 46110(a) with a party's substantial interest in the order). Here Safe Extensions unquestionably has a substantial interest in the order: if left in place, AC-42F will destroy the market for the company's product.

The FAA argues that Safe Extensions nevertheless falls outside the zone of interests protected by section 46110(a) because the statute technically allows the FAA to regulate airports, not manufacturers. Again, this argument is absurd. AC-42F effectively regulates Safe Extensions because it prevents the company from selling its product to airports. Moreover, to have prudential standing, Safe Extensions must show only that its "interest is 'arguably' one regulated or protected by 'the statutory provision at issue.'" PDK Labs., Inc. v. DEA, 362 F.3d 786, 791 (D.C. Cir. 2004) (quoting Nat'l Credit Union Admin. v. First Nat'l Bank, 522 U.S. 479, 492 (1998)). Because Safe Extensions has easily met that minimal standard, the company has prudential standing.

Agency Discretion

For its third jurisdictional argument, the FAA claims that AC-42F is unreviewable because the issues it addresses are committed by law to agency discretion. Specifically, the FAA claims that there is no meaningful standard for this court to apply, that advisory circulars are within the FAA's "managerial discretion," and that judicial review of advisory circulars would greatly tax the agency's resources. All three arguments lack merit.

As to the first point, the FAA is certainly correct that "[i]f no 'judicially manageable standard' exists by which to judge the agency's action, meaningful judicial review is impossible and the courts are without jurisdiction to review that action." Steenholdt v. FAA, 314 F.3d 633, 638 (D.C. Cir. 2003) (quoting Heckler v. Chaney, 470 U.S. 821, 830 (1984)). But that is hardly the case here. The FAA issued AC-42F pursuant to 49 U.S.C. § 44701, which authorizes the agency to "(b) . . . prescribe minimum safety standards for . . . (2) operating an airport serving any passenger . . . aircraft designed for at least 31 passenger seats," and directs the agency to "(a) . . . promote safe flight of civil aircraft . . . by prescribing . . . (5) regulations and minimum standards for . . . practices, methods, and procedure[s] the [agency] finds necessary for safety in air commerce." This provides a perfectly workable standard to guide the court, namely whether the FAA's actions promote air safety.

We have previously found far more ambiguous statutory directives reviewable. For example, in Dickson v. Secretary of Defense, 68 F.3d 1396 (D.C. Cir. 1995), we held that we could review the Army Board for Correction of Military Records's refusal to waive a statute of limitations period under 10 U.S.C. § 1552(b), which allows the Board to "excuse a failure to file within three years . . . if it finds it to be in the interest of justice."

If "in the interest of justice" provides a judicially manageable standard, then "necessary for safety" certainly does as well. See, e.g., Union of Concerned Scientists v. Nuclear Regulatory Comm'n, 824 F.2d 108 (D.C. Cir. 1987) (reviewing whether regulations "provide[d] adequate protection to the health and safety of the public" under 42 U.S.C. § 2232(a)).

This conclusion finds support in the APA's "strong presumption of reviewability." Steenholdt, 314 F.3d at 638. As the Supreme Court has declared: "judicial review of a final agency action by an aggrieved person will not be cut off unless there is persuasive reason to believe that such was the purpose of Congress." Abbott Labs. v. Gardner, 387 U.S. 136, 140 (1967). "Because of this presumption favoring judicial review, we require 'clear and convincing evidence of a legislative intention' to bar such review." James Madison Ltd. by Hecht v. Ludwig, 82 F.3d 1085, 1092 (D.C. Cir. 1996) (quoting Ball, Ball & Brosamer, Inc. v. Reich, 24 F.3d 1447, 1450 (D.C. Cir. 1994)). Nothing in the Federal Aviation Act even hints that Congress intended decisions like the one at issue here to be unreviewable. Indeed, demonstrating just the opposite, 49 U.S.C. § 46110(a) gives this court jurisdiction to review FAA orders challenged by any "person disclosing a substantial interest in [them]."

The FAA next argues that AC-42F "reflects the FAA's managerial choice, one not susceptible to judicial review." Resp't's Br. 28. Not only does this argument amount to an attack on the very principles underlying the APA, including its "basic presumption of judicial review [for] one 'suffering legal wrong because of agency action,'" Abbott Labs., 387 U.S. at 140 (quoting 5 U.S.C. § 702)), but it also ignores that advisory circulars are no more a "managerial choice" than are the regulations this court routinely reviews from countless other agencies. See, e.g., ASG Indus., Inc. v. Consumer Prod. Safety Comm'n, 593 F.2d 1323 (D.C. Cir. 1979) (reviewing the Consumer Product Safety Commission's regulations governing architectural glazing materials). The FAA cites not a single case to support this baseless contention.

Finally, the FAA argues that "allowing judicial review of issuances like this Circular would be very disruptive to the agency's operations." Resp't's Br. 28. If review is allowed here, the agency complains, disgruntled manufacturers will sue the FAA whenever it issues new advisory circulars. Given that this is, as best we can tell, the first time in the FAA's decades of issuing advisory circulars that a manufacturer has ever petitioned for review of a circular regulating its products, we highly doubt the factual premise underlying this argument. More fundamentally, the FAA's argument ignores the APA's very purpose: to subject agency decisions to judicial scrutiny. No one pretends that judicial review of agency action is a pleasant day at the beach for agencies, and although escaping judicial review would of course be less "disruptive to the [FAA's] operations," id., it would also leave regulated entities, as well as the flying public-which depends for its safety on solid, well supported FAA decisionmaking-unprotected from arbitrary and capricious agency action.

Furthermore, the cases the FAA cites to support its argument have nothing to do with the issue before us. Both National Federation of Federal Employees v. United States, 905 F.2d 400 (D.C. Cir. 1990), and Curran v. Laird, 420 F.2d 122 (D.C. Cir. 1969), involved military decisions about national security issues this court felt were beyond its purview. See Nat'l Fed'n, 905 F.2d at 405-06; Curran, 420 F.2d at 131-32. And in Southern Railway Co. v. Seaboard Allied Milling Corp., 442 U.S. 444 (1979), the Supreme Court concluded that it could find no law to apply and only then said the disruptive consequences of allowing review confirmed this position. See id. at 455-59.

Moreover, a ruling for the petitioner in Southern Railway would have required the Interstate Commerce Commission to provide detailed justifications for millions of individual rates it approved each year, a disruption several orders of magnitude greater than the FAA would suffer in this case. See id. at 457. In short, these cases provide no support for the ludicrous proposition that courts may not review FAA advisory circulars because judicial review would create more work for the agency. Like virtually every other agency, the FAA must defend its decisions in court.


Although the statute granting us jurisdiction in this case generally requires petitions for review of FAA orders to be filed within sixty days of the challenged order, it provides that we "may allow the petition to be filed after the 60th day . . . if there are reasonable grounds for not filing by the 60th day." 49 U.S.C. § 46110(a). Neither party discussed timeliness in their briefs, but in preparing for this case we discovered that the differential treatment Safe Extensions challenges stems from AC-42E, not AC-42F, and the company filed its petition more than sixty days after the FAA issued AC-42E. Accordingly, we directed the parties to be prepared to discuss this issue at oral argument.

At oral argument, Safe Extensions's counsel asserted that there were "reasonable grounds" for the company's failure to file by the 60th day after AC-42E's issuance. He said that after the FAA issued AC-42E, the agency "told the industry . . . : 'Ignore Advisory Circular 42E.'" Oral Arg. at 2:04. The reason for the FAA's directive, according to the company, was that AC-42E produced "a significant uproar in the industry," id. at 3:06, causing numerous companies to meet with the FAA and tell the agency the advisory circular was unworkable. "As a result," counsel said, "the FAA told industry virtually immediately:

'You're right. . . . We're going to issue a new draft, a draft F." Id. at 3:34. Based on these representations, and hoping to avoid litigation, the company decided to "wait and see if the FAA [would] address[] the issues [Safe Extensions] had with 42E" in 42F. Id. at 4:25.

The FAA's counsel responded that now that the agency had thought about the issue, it believed the petition was untimely. Disputing the version of events described by Safe Extensions, the agency's counsel said "I cannot believe that there was any dialogue in which the FAA told the industry that it need not comply with 42E," id. at 17:51, and "I've talked to the responsible FAA officials on this, and if there had been some informal undertaking to re-examine this issue, I am confident that they would have told me about it," id. at 21:31.

Because of this factual dispute we asked both parties to submit affidavits or documents to support their claims. Safe Extensions's affidavits strongly support the company's description of what happened. One affidavit, from a former FAA employee who now works for a company that writes specifications for the agency, declares that FAA officials told him "to basically ignore AC-42E because [it] would be eliminated and replaced with [AC-42F]," and that he passed this information along to Safe Extensions. Oswald Aff. 1-2. In another affidavit, the president of a company that manufactures light bases declares that an FAA employee told him to "forget about AC-42E because the FAA was currently revising AC-42E and that it would become [AC-42F]." Tappe Aff. 1-2. Moreover, the president of Safe Extensions tells us that when he "expressed [his] concerns to [the FAA] about how AC-42E dealt with [adjustable products]," the FAA responded "that [he] should wait until AC-42F comes out because the FAA was currently revising AC-42E." Reinert Aff. 1-2. Bolstering these affidavits, Safe Extensions's petition for review alleged that the FAA had "failed to provide any rational basis for refusing to revise the Advisory Circular," Pet. for Review 2, Safe Extensions, Inc. v. FAA, No. 06-1412 (D.C. Cir. Dec. 15, 2006) (emphasis added), suggesting the company had reason to believe the agency was planning to address its concerns with AC-42E in AC-42F.

For their part, the FAA's affidavits make only the very narrow claim that the agency never told anyone that it "would reconsider the torque testing standards for [runway] light bases in connection with a revision of Circular 42-E," Smith Aff. 4; de Maria Aff. 4, leaving open the possibility that FAA employees made more general or ambiguous statements about AC-42E that could have confused petitioner and others. Moreover, the FAA's failure to raise the timeliness issue itself-despite presenting nearly every other jurisdictional argument imaginable-leads us to question how strongly the agency really believes that Safe Extensions should have filed earlier.

Based on this evidence, we conclude that Safe Extensions had reasonable grounds for filing more than sixty days after the FAA issued AC-42E. This conclusion finds support in our precedent. In Paralyzed Veterans of America v. Civil Aeronautics Board, 752 F.2d 694 (D.C. Cir. 1985), rev'd on other grounds, Dep't of Transp. v. Paralyzed Veterans of America, 477 U.S. 597 (1986), we held that an organization had reasonable grounds for waiting more than sixty days to file a challenge when the group was "[a]ware that the rule might be undergoing modification[] and unable to predict how extensive any modification would be," and therefore "elected to wait until the regulation was in final form before seeking review." Id. at 705 n.82. This aptly describes what occurred here. As we said in Paralyzed Veterans: "Any delay simply served properly to exhaust petitioners' administrative remedies, and to conserve the resources of both the litigants and this court." Id.


Having disposed of the FAA's jurisdictional arguments, we turn to the merits. Our standard of review is both familiar and deferential: we review the FAA's actions under the APA to determine whether they were "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law." 5 U.S.C. § 706(2)(A); Cmtys. Against Runway Expansion, Inc. v. FAA, 355 F.3d 678, 685 (D.C. Cir. 2004). "Under this standard, we 'may reverse only if the agency's decision is not supported by substantial evidence, or the agency has made a clear error in judgment.'" J.A. Jones Mgmt. Servs. v. FAA, 225 F.3d 761, 764 (D.C. Cir. 2000) (quoting Kisser v. Cisneros, 14 F.3d 615, 619 (D.C. Cir. 1994)); see also 49 U.S.C. § 46110(c) (FAA "[f]indings of fact . . . , if supported by substantial evidence, are conclusive.").

To be sure, no statute requires the FAA to engage in the notice and comment process or hold proceedings on the record when issuing advisory circulars. Instead, advisory circulars fall into the vast category of "informal adjudications" in which agencies routinely engage. Nonetheless, as we explained in Association of Data Processing Service Organizations, Inc. v. Board of Governors of Federal Reserve System, 745 F.2d 677 (D.C. Cir. 1984), the agency's decision still must be supported by substantial evidence-otherwise it would be arbitrary and capricious. For "it is impossible to conceive of a 'non-arbitrary' factual judgment supported only by evidence that is not substantial in the APA sense." Id. at 684. The difference between this informal adjudication and a formal adjudication on the record is thus not the amount of evidence the agency must provide to support its decision, but where the evidence may be found. In an adjudication on the record "substantial evidence [must] be found within the record of closed-record proceedings," while in an informal adjudication the agency can provide the court with any evidence it had before it when it made its decision. Id.

In its brief, the FAA offers three rationales for requiring freestanding torque testing of adjustable products but not fixed products. First, it claims that it has had "extensive experience with [f]ixed [p]roducts," while "[a]djustable [p]roducts are relatively new," justifying the "FAA's determination to subject the newer product to more rigorous laboratory tests." Resp't's Br. 33. If true, this rationale might support the FAA's differential treatment of the two products. But the FAA has provided absolutely no evidence to back it up, and as we made clear in McDonnell Douglas Corp. v. Department of the Air Force, 375 F.3d 1182 (D.C. Cir. 2004), an agency's "declaration of fact that is 'capable of exact proof' but is unsupported by any evidence" is insufficient to make the agency's decision non-arbitrary. Id. at 1191 n.4.

Second, the FAA claims that fixed products "are configured with anti-rotational devices as part of their design," while adjustable products "lack this feature." Resp't's Br. 31. For this reason, adjustable products, but not fixed products, "depend on chemical bonding" with the concrete or grout surrounding them "to maintain their stability." Id. at 33. Since it cannot possibly test every adjustable product after installation to ensure the chemical bond works properly, the agency claims "it was appropriate . . . to guard against the heightened risk posed by improper installation . . . by requiring more rigorous laboratory testing." Id. For its part, Safe Extensions claims in its reply brief and an attached appendix that its products actually do have anti-rotational devices and that many fixed products lack such devices. The FAA asks us to ignore this argument and Safe Extensions's appendix because the company never submitted evidence to support its argument to the FAA. The FAA may well be right about this. But we have no need to examine the reply brief and appendix because the FAA has failed to provide substantial evidence for its argument that fixed products, unlike adjustable products, have anti-rotational devices.

The only evidence regarding the presence of anti-rotational devices in light bases appears in the agency's response to industry comments on AC-42F. There, an FAA employee wrote: "Fixed . . . bases and extensions have anti-rotational devices inherent to the physical structure of the device itself . . . . Adjustable height extensions in known currently available designs have no physical features inherent to the device itself to prevent rotation . . . ." Comment Resolutions for Draft Advisory Circular 150/5345-42F, at 1. The FAA, however, has provided no documents, drawings, or affidavits to support this claim, and the agency admits that the record it has provided us "includes all the documents upon which the agency relied." Resp't's Br. 17. As we have said many times before, "[a]n agency's unsupported assertion does not amount to substantial evidence." Algonquin Gas Transmission Co. v. FERC, 948 F.2d 1305, 1313 (D.C. Cir. 1991).

As its final rationale for treating fixed and adjustable products differently, the FAA tells us that "it received a number [sic] reports of problems in the field with respect to [a]djustable [p]roducts." Resp't's Br. 34. Again, the only evidence the FAA has provided to support this rationale appears in its response to industry comments:

Adjustable [products] in known currently available designs have no physical features inherent to the device itself to prevent rotation, and rely on a chemical bond acting in shear. There are known examples where this bond has failed in actual application following exposure to real world loading and environmental conditions.

Comment Resolutions for Draft Advisory Circular 150/5345-42F, at 1. This rationale is even weaker than the previous two. While the first rationale entirely lacked support and the second had insufficient support, this one is actually contradicted by evidence the FAA had before it. In their comments on the draft of AC-42F, two companies that have installed thousands of adjustable products said they were "not aware of any field related issues with" adjustable products. Email from Ted Olson, Jr.; Email from Willis Trainor. By contrast, the FAA offered no evidence whatsoever of problems in the field with adjustable products, thus failing to provide substantial evidence to support this rationale.

In sum, because the agency's decision to treat fixed and adjustable products differently finds no support in the evidence the agency considered, we find it arbitrary and capricious. See Data Processing Serv. Orgs.,745 F.2d at 683-84 (explaining that an agency decision unsupported by substantial evidence is arbitrary and capricious).


Because the FAA's approach in this case flouts fundamental principles underlying the Administrative Procedure Act, we grant the petition for review.

So ordered.

Back to top