Summary
In Aman, the petitioning pilots argued that any undetected aging decrements or increased risks of incapacitation that accompanied pilots sixty and over were offset by the added experience that attends to pilots of that age.
Summary of this case from Yetman v. GarveyOpinion
No. 87-2598.
Argued April 15, 1988.
Decided September 12, 1988. As Amended October 19, 1988. Rehearing Denied November 29, 1988.
Raymond C. Fay, Bell, Boyd Lloyd, Washington, D.C., for petitioners.
John Craig Weller, Office of Chief Counsel, Washington, D.C., for respondents.
Petition for review from Federal Aviation Administration.
Before WOOD, Jr. and CUDAHY, Circuit Judges, and FAIRCHILD, Senior Circuit Judge.
Petitioners in this case, pilots employed or formerly employed as captains for major airlines, seek review under 49 U.S.C.App. Section 1486(a) of an order of the Federal Aviation Administration ("FAA"). The challenged order denied a Petition for Exemptions from section 121.383(c) of the FAA regulations (the "age sixty rule"). Because we find that the FAA's findings and explanations with respect to an important aspect of the petitioners' argument for exemptions were inadequate, we vacate and remand.
I.
The age sixty rule prohibits flights covered by Part 121 of the FAA regulations, including commercial flights of airplanes seating more than thirty passengers, from taking off under the command of a pilot sixty years old or older. See 14 C.F.R. §§ 121.1, 121.383(c) (1988). The age sixty rule does not impose mandatory retirement; but to remain in the cockpit air carrier pilots must continue to meet medical certificate requirements and restrict themselves to flights outside the coverage of Part 121 or accept demotions to positions of lesser authority, such as those of flight instructor or flight engineer. The FAA put the age sixty rule into effect in 1960. In defending the rule then, the FAA cited concerns about sudden incapacitation due to strokes and heart attacks and about the ability of the most senior pilots to operate the largest and fastest planes, see Air Lines Pilots Ass'n, Int'l v. Quesada, 276 F.2d 892, 898 (2d Cir. 1960), cert. denied, 366 U.S. 962, 81 S.Ct. 1923, 6 L.Ed.2d 1254 (1961). The agency, however, held out the hope that "medical science may at some future time develop accurate, validly selective tests which would safely allow selected pilots to fly in air carrier operations after age 60." Certification and Operation Rules: Maximum Age Limitations for Pilots, 24 Fed.Reg. 9772, 9772 (1959). The rule was upheld against a variety of procedural and substantive challenges in Quesada and in O'Donnell v. Shaffer, 491 F.2d 59 (D.C. Cir. 1974).
Since its enactment, the age sixty rule has been the subject of continuous controversy and intermittent reconsideration by the FAA. A study undertaken in the early 1960's to determine the feasibility of individualized assessments of pilots over sixty was abandoned before any results were formally announced. In 1969, an outside report was commissioned and completed but never released. In 1979, the FAA reviewed a Navy study on the long-term health histories of 1,000 aviators but concluded that the study failed to provide an adequate basis for revising the age sixty rule. In that same year, Congress expressed its interest in the issue, enacting legislation that directed the National Institutes of Health to assess whether the rule should be retained. Pub.L. No. 96-171, 93 Stat. 1285, 49 U.S.C.App. § 1421 Note (1982). A panel of the National Institute of Aging was convened to fulfill the congressional mandate. The panel expressed doubts about the need for all pilots to step aside at age sixty, but recommended, in light of gaps in available information, that the rule be retained as a general policy while a study involving a selected group of older pilots exempted from the rule was conducted to assess the feasibility of a more individualized approach. After reviewing the panel's recommendation, the FAA published a notice soliciting comments on two proposals: an expansion of the scope of the age sixty rule to cover flight engineers as well as pilots and a test program that would allow selected pilots to continue serving as captains on commercial flights until their sixty-second birthdays to generate data for reconsideration of the rule. Advance Notice of Proposed Rule-making, 47 Fed.Reg. 29,782, 29,782-84 (1982). However, two years later the FAA withdrew its test program proposal (together with the proposal to extend the rule), stating that the comments had failed to reveal "any means by which a scientifically valid study can be done to produce information which would support . . . determination[s]" as to which pilots "could safely serve . . . after reaching age 60." Withdrawal of Advance Notice of Proposed Rulemaking, 49 Fed.Reg. 14,692, 14,692 (1984). The FAA found inadequacies in suggested methods for assessing performance, especially under conditions of stress and fatigue, and for assessing the risk of incapacitation, especially incapacitation due to stroke. Id. at 14,693.
A number of pilots have sought to circumvent the age sixty rule through the exemption process. To date, their efforts have been uniformly unsuccessful both before the FAA and before the courts of appeals. See Keating v. FAA, 610 F.2d 611, 613 (9th Cir. 1979); Gray v. FAA, 594 F.2d 793, 795 (10th Cir. 1979); Rombough v. FAA, 594 F.2d 893, 899-900 (2d Cir. 1979); Starr v. FAA, 589 F.2d 307, 311-14 (7th Cir. 1978). Both this court and the Tenth Circuit, however, have warned that later denials of exemptions could be overturned if petitioners were able to show that the FAA failed to consider the effect of advances in medical technology on the necessity of maintaining a uniform age sixty cutoff. Starr, 554 F.2d at 314; Gray, 594 F.2d at 795. Our impression is that in the ten years that have passed since Starr, when the FAA last appeared before this court defending the denial of an exemption to the age sixty rule (before a panel, incidentally, that included two members of the present panel), the agency's progress in developing an understanding of the relationship between aging and flight performance has been disappointing.
The petitioners before this court are twenty-eight pilots employed as captains or flight engineers by major airlines. (Eleven other pilots joined in the original petition to the FAA, but declined to join in the request for review.) When the petition for exemption was filed with the FAA on June 3, 1986, twenty-five of the twenty-eight had passed the age of sixty and were working (with one exception) as flight engineers; the remaining three, who were approaching sixty when the petition was filed, were working as captains. The petition relied heavily on the recommendations of a six-member "Age 60 Exemption Panel," comprising five physicians and a psychologist with impressive qualifications in the fields of cardiology, aerospace medicine and neuropsychology. The panel devised an extensive battery of physiological and psychological tests as a basic protocol for assessing the fitness of pilots over the age of sixty. In the petition to the FAA, the panel stated that this protocol, if properly administered and supplemented, where appropriate, by additional medical tests and by the existing operational tests required by the FAA and the airlines (such as flight simulator testing), provided an adequate basis for exempting some older pilots from the age sixty rule. Statement of Age 60 Exemption Panel (Apr. 1986), reprinted in Joint Appendix 1-3.
The FAA excluded 5 of the original 39 petitioners from consideration because they had not yet reached their sixtieth birthdays when the petitions were filed. Although petitioners contest this refusal, we believe that the FAA is entitled to conserve its resources by declining to consider petitioners whose career plans or health could change before the age sixty rule applied to them. We note, however, that section 11. 25(b)(1) of the FAA regulations requires that petitions for exemption be filed within 120 days of the requested effective date of the exemption. 14 C.F.R. § 11.25(b)(1) (1988). We see no justification for the FAA to decline to consider a request for exemption submitted by a pilot within 120 days of his or her sixtieth birthday.
On July 24, 1986, the FAA published a summary of the Petition for Exemption and invited public comment. In response to a request from one interested party, the FAA reopened the docket for thirty days after the normal twenty-day comment period had ended. The final docket contained over 180 submissions from physicians, scientists, Congressmen, pilots, professional organizations, companies and the FAA itself, together with the petitioners' two supplements to the original petition. On September 8, 1987, the FAA denied the requested exemptions based on its findings that granting individualized exemptions under the petitioners' standards would not ensure the level of safety achieved by uniform enforcement of the age sixty rule. Petitioners then brought this action for review.
Raising a procedural objection that appears nowhere in the order, the FAA urges us to dismiss the petitioners' claim as an improperly framed request for a change in the rule rather than for exemptions to the rule. The FAA may be correct in asserting that if the petitioners were to establish that they are entitled to exemptions under the relevant statutory standard they would call into question the underpinnings of the rule itself. However, the FAA has presented no authority or logical argument for the proposition that an exemption must be denied if approval would cast doubt on the basis for a rule.
II.
The parties disagree first as to whether this court should review the FAA's findings of fact under a "substantial evidence" or an "arbitrary and capricious" standard. Section 1006(e) of the Federal Aviation Act, 49 U.S.C.App. § 1486(e) (1982), provides that the FAA's "findings of fact . . . , if supported by substantial evidence, shall be conclusive." Several courts of appeals, however, have held that under some circumstances the arguably more deferential arbitrary and capricious standard applies, section 1006(e) notwithstanding.
A number of this circuit's decisions have suggested that in view of the requirement for some evidentiary support under the arbitrary and capricious standard the difference between the "substantial evidence" and "arbitrary and capricious" standards may be "primarily a semantic distinction." See, e.g., Central States Enters., Inc. v. ICC, 780 F.2d 664, 674 n. 10 (7th Cir. 1985); see Indiana Sugars, Inc. v. ICC, 694 F.2d 1098, 1099 (7th Cir. 1982); People of the State of Illinois v. United States, 666 F.2d 1066, 1072 n. 6 (7th Cir. 1981); see also Black v. ICC, 737 F.2d 643, 650 (7th Cir. 1984) (in some circumstances "these two standards of review are the same for all practical purposes"). To be sure, an agency determination must have some evidentiary basis to avoid being held "arbitrary and capricious." The difference between "some" and "substantial" probably cannot be precisely stated except in the context of particular cases. See, e.g., Aqua Slide `N' Dive Corp. v. Consumer Product Safety Comm'n, 569 F.2d 831, 841 (5th Cir. 1978). The difference may go more to atmospherics than to the patently demonstrable although, as indicated, certain cases may show the difference more palpably than others.
In Tiger International, Inc. v. Civil Aeronautics Board, 554 F.2d 926 (9th Cir.), cert. denied, 434 U.S. 975, 98 S.Ct. 532, 54 L.Ed.2d 467 (1977), the Ninth Circuit reviewed a quasi-adjudicatory determination by the Board, a decision to impose conditions on transactions among affiliates to protect the financial well-being of an air carrier. The court held that although the determination might appear to have been covered by section 1006(e), it should be reviewed under an arbitrary and capricious standard because it had been reached in a case "where no hearing was required or requested" and where the record, as a consequence, consisted primarily of the petitioner's submissions. 554 F.2d at 936. The court noted that the Supreme Court has applied differing standards of review to factual findings made under the Administrative Procedure Act based on permissible differences in fact-finding processes. See Camp v. Pitts, 411 U.S. 138, 140-42, 93 S.Ct. 1241, 1243-44, 36 L.Ed.2d 106 (1973), cited in Tiger Int'l, 554 F.2d at 935. The Ninth Circuit observed that uniform application of the substantial evidence standard would require the administrative decision maker to convene a hearing or introduce evidence contrary to a petitioner's position in some other manner in every proceeding, even where Congress has plainly indicated that the FAA should be permitted to reach a decision without holding a hearing. Tiger Int'l, 554 F.2d at 936-37 n. 20.
Since Tiger International was decided, four circuits, including this one, have reviewed denials of exemptions to the age sixty rule under an arbitrary and capricious standard. Keating, 610 F.2d at 612; Gray, 594 F.2d at 795; Rombough, 594 F.2d at 896-97; Starr, 589 F.2d at 310-11. Starr approvingly cited Tiger International's discussion of the impossibility of reviewing one-sided records under a substantial evidence standard and added its own observation that if FAA decisions on exemption requests were reviewed under that standard, the "agency would be forced to defend its standard rule in every exemption proceeding." Starr, 589 F.2d at 311, see also Association of Bank Travel Bureaus, Inc. v. Board of Governors, 568 F.2d 549, 552 (7th Cir. 1978) (reviewing findings in informal rulemaking under arbitrary and capricious standard instead of statutory substantial evidence standard).
An earlier decision of this circuit, Coppenbarger v. FAA, 558 F.2d 836 (7th Cir. 1977), considered a challenge to the constitutionality of the FAA's exemption procedures. That decision, however, did not address the standard of review question.
Other recent decisions, reviewing FAA decisions outside the exemption context, have expressed reluctance to extend the scope of Tiger International. In Aircraft Owners Pilots Association v. FAA, 600 F.2d 965 (D.C. Cir. 1979), the District of Columbia Circuit considered a challenge to an FAA determination that the construction of a television antenna would not pose a hazard to air navigation. The court observed that Tiger International was forced to deal with the "inherent problems" of applying a substantial evidence test "to an informal adjudicatory decision made absent the creation of an adequate record."
600 F.2d at 970. That practical difficulty was not encountered in Aircraft Owners, however. The "record before the FAA in [that] action was not sparse"; it "embraced all facts necessary to make a hazard/no hazard determination." Id. at 971. Thus, the court applied the statutory substantial evidence standard, while expressly reserving the question of what standard would apply if a one-sided record made the substantial evidence test infeasible. See also State of South Dakota v. Civil Aeronautics Bd., 740 F.2d 619, 621 (8th Cir. 1984) (substantial evidence review of decision permitting decrease in service where statutory standard "can be applied to the record before [the court]").
This court relied on Aircraft Owners to justify substantial evidence review of FAA decisions approving the expansion of Chicago's O'Hare International Airport. Suburban O'Hare Comm'n v. Dole, 787 F.2d 186, 194-95 (7th Cir.), cert. denied, 479 U.S. 847, 107 S.Ct. 169, 93 L.Ed.2d 106 (1986). That decision criticized Tiger International for according too much weight to case law interpreting the Administrative Procedure Act and too little weight to the specific review provisions of section 1006(e). The court stopped short, however, of overturning Starr, distinguishing it instead on the grounds that Starr raised special problems associated with applying a substantial evidence test to facts found in connection with agency consideration of a petition for exemption. Suburban O'Hare, 787 F.2d at 195.
In some cases, the outcome might turn on whether Starr or Suburban O'Hare set the standard of review for the factual predicates of exemption determinations in which the FAA, in the exercise of its considerable discretion over procedural matters, develops a full record. We think that here, because of the fullness of the record, the Suburban O'Hare standard is quite workable and we will apply it, although if we employed the arbitrary and capricious standard we would reach the same result. Here, as in Suburban O'Hare, "lengthy and elaborate decision making procedures" preceded the FAA's determination. 787 F.2d at 195. The FAA's docket contained not only a lengthy petition from the pilots (augmented by two supplemental submissions), but also 183 separate comments filed by parties on both sides of the issue during a fifty-day comment period. Although the FAA is not required to develop a record conducive to substantial evidence review, see 14 C.F.R. § 11.27(a), (j)(3) (1988), once it elects to build such a record there is a strong argument that its decision should be reviewed under the standard set forth in section 1006(e). There has been speculation that the approach set out in Suburban O'Hare and adopted here could tend to encourage the FAA to restrict its fact-finding efforts in informal adjudications in order to preclude the application of the more penetrating standard of review. This problem, however, stems not from Suburban O'Hare, but from the statute's establishment of a substantial evidence standard of review while conferring discretion to utilize procedures that so limit the record as to preclude meaningful application of this standard. Abandoning Suburban O'Hare to defuse this problem would inappropriately allow concerns about the FAA's procedural choices, concerns resting at present on speculation alone, to broaden the exception that Starr recognized to the application of the substantial evidence standard of review called for by section 1006(e).
It is conceivable, moreover, that the FAA's limitation of the scope of its inquiry could contribute to a finding that it abused its discretion. Cf. Starr, 589 F.2d at 314 (suggesting that the FAA could abuse its discretion "by blindly adhering to an outdated rule").
Our application of the substantial evidence standard of section 1006(e) is limited, both by the terms of the statute and by common sense, to the FAA's findings of fact. In other respects, including most notably the application of legal standards to the facts as found, the FAA's decision to grant or deny an exemption may be set aside if "arbitrary, capricious, an abuse of discretion, or otherwise not in accordance with law" — the generally applicable standard of review for administrative actions under 5 U.S.C. § 706(2)(A). See City of Pompano Beach v. FAA, 774 F.2d 1529, 1540 (11th Cir. 1985). The authoritative explication of this standard was set forth in Motor Vehicle Manufacturers Association v. State Farm Mutual Insurance Co., in which the Supreme Court stated:
The scope of review under the "arbitrary and capricious" standard is narrow and a court is not to substitute its judgment for that of the agency. Nevertheless, the agency must examine the relevant data and articulate a satisfactory explanation for its action including a "rational connection between the facts found and the choice made."
463 U.S. 29, 103 S.Ct. 2856, 2866, 77 L.Ed.2d 443 (1983) (quoting Burlington Truck Lines, Inc. v. United States, 371 U.S. 156, 168, 83 S.Ct. 239, 245, 9 L.Ed.2d 207 (1962)); see Central States Enters., Inc. v. ICC, 780 F.2d 664, 674 (7th Cir. 1985). We will assess the FAA's application of the Federal Aviation Act to its factual findings under this standard, bearing in mind that the construction of the Act by the FAA — the agency charged with administering the Act — is entitled to substantial deference. United States v. Rutherford, 442 U.S. 544, 553, 99 S.Ct. 2470, 2475, 61 L.Ed.2d 68 (1979).
III.
Petitioners' argument may be divided, for purposes of this discussion, into two alternative claims. There is, first, a claim that pilots over the age of sixty who meet the psychological and medical standards of the petitioners' protocol, as well as FAA and airline operational testing standards, are no more likely to cause accidents due to sudden incapacitation or undetected deterioration of piloting skills than are pilots below the age of sixty. This claim requires the petitioners to make a demanding factual argument, an argument that the FAA lacked substantial evidence for a finding that strict enforcement of the age sixty rule reduces age-related risks of incapacitation and undetected deterioration of piloting skills. But the petitioners may rest this claim on an uncontroversial legal argument, that to justify its position under the Act's public interest standard for exemptions the FAA must at least identify some risk to air carrier safety that is reduced by strict enforcement of the age sixty rule.
The petitioners' alternative claim asserts that an exemption must be granted because older pilots who satisfy the protocol and existing operational tests are safer than the average pilot because performance improves with experience. According to the petitioners, the experience factor at least offsets, and may even outweigh, any increased risk of incapacitation or skill deterioration. From the petitioners' perspective, this claim rests on a more promising argument about the facts, that the FAA lacked substantial evidence for its conclusions that allowing selected pilots to remain in command past the age of sixty would cause a net decrease in safety. But petitioners must make a more difficult legal argument under their alternative claim, that the public interest standard for exemptions requires the FAA to justify the denial of exemptions in terms of net risks. We consider these two claims separately.
A.
The FAA devoted the bulk of its order to a defense of its conclusion that the petitioners' health protocol, even when coupled with existing operational tests, cannot entirely screen out the increased risks of incapacitation and error due to the deterioration of skills that come with age. The FAA's most concise formulation of its position on this point states that "[b]ecause the likelihood of sudden death, disability, or incapacitation due to previously undetected disease increases at an accelerating rate with each additional year of chronological age, granting exemptions would compromise, by some amount, the current level of safety." Denial of Exemption, Docket No. 25008 at 13 (Sept. 8, 1987) (hereinafter "FAA Order"). The FAA supports this position with citations to a number of sources. On the increased incidence of sudden incapacitating illness with age, the order cites a docket submission by Dr. Lawrence Marinelli, Chairman of the Medical Panel of the Air Transport Association, who refers to the increased rate of denials of first class medical certificates among older pilots and expresses concern about the heightened risk among older pilots of significant physical deterioration during the six-month intervals between examinations. See Submission to Docket 25008 (Oct. 16, 1986), reprinted in Joint Appendix at 285-86.
The order cites additional evidence for the proposition that neither the "neuropsychological" components of the protocol nor the operational tests conducted by the FAA and the airlines (principally flight simulator testing monitored and evaluated by flight testers, see 14 C.F.R. §§ 121.407-.413, .419, .424 App. H (1988)), are capable of identifying, much less predicting, many potentially significant losses in flying skills. A submission to the docket by the Airline Pilots Association contained the assertion, distilled from the observations of the Association's members, "that a pilot's skill deteriorates with age." Submission to Docket No. 25008 (Oct. 28, 1986), reprinted in Joint Appendix at 297-A. The order also quotes from the statements of several scientists holding similar views, including a 1985 statement by a NASA scientist specializing in research on human performance in flight, who found " no evidence that objective, reliable tests of cognitive function in a rich and rapidly changing environment exist, let alone that they have been applied to or validated in a pilot population of any age." FAA Order at 14-15 (quoting statement of Dr. Charles E. Billings); see also id. at 14 (quoting statement of Dr. Don Flinn, Professor of Psychiatry at Texas Tech Health Science Center School of Medicine); id. at 14-15 (quoting letter from William Aker, a British consultant and psychologist, to House Select Committee on Aging (Dec. 16, 1986), reprinted in Joint Appendix at 416-17).
Petitioners protest the FAA's reliance on the statements of Flinn, Billings and Acker on the grounds that these comments were not submitted to the docket. (The status of Acker's letter is not entirely clear, however. The petitioners' brief does not refer to Acker's letter in one of the two passages where this objection is raised, see Petitioners' Brief at 43 n. 24, and Acker's letter is reproduced in the Joint Appendix.) This problem, which the FAA could easily have avoided merely by entering these comments into the docket along with the other documents that it submitted, raises questions about the characterization of the FAA exemption determination, whether it was an adjudication or a rulemaking, and about the characterization of the factual claims in the comments, whether they are adjudicative or legislative facts. Petitioners tacitly assume that the statements should be assessed as statements of legislative fact in a rulemaking proceeding, referring us to Aqua Slide, 569 F.2d at 842-43, which addressed the evidentiary weight that a reviewing court should accord a report submitted after the close of the public comment period in a rulemaking proceeding. This may not be an unreasonable characterization inasmuch as the FAA's procedural regulations accord very similar treatment to petitions for rulemaking and petitions for exemption. See 14 C.F.R. §§ 11.25-.27 (1988). Following Aqua Slide, we would assign diminished weight to the statements in question in view of the absence of an opportunity for public comment on these statements, though the adjustment would be smaller in this case since interested parties were generally aware that some experts held these positions due to the FAA's reliance on this viewpoint when it revoked its proposal to create limited exemptions to the age sixty rule in 1984. Even with such an adjustment, we believe that the FAA had substantial evidence for its position that some of the risks associated with age cannot be screened out completely by the protocol and operational tests. Moreover, even if we were to assume, contrary to the implication of the petitioners' argument, that the statements in question were statements of adjudicative fact considered in the context of an informal adjudication (that is, an adjudication to which the requirement of decision making on the record does not pertain), there is no indication, given petitioners' awareness of the basis for the FAA's 1984 action, that limited reliance on these statements deprived the petitioners of what one commentator has called "fair informal procedure." See 2 K. Davis, Administrative Law Treatise § 13:2 (2d ed. 1979) (discussing principle of Goss v. Lopez, 419 U.S. 565, 581-82, 95 S.Ct. 729, 739-40, 42 L.Ed.2d 725 (1975), that in some circumstances parties to informal adjudication must have an opportunity to respond to a summary of adverse evidence); see also United States v. Pierce Auto Freight Lines, 327 U.S. 515, 530, 66 S.Ct. 687, 695, 90 L.Ed. 821 (1946) ("mere fact that the determining body has looked beyond the record proper does not invalidate its action unless substantial prejudice is shown to result").
Petitioners offered a wealth of support for the contrary view that modern testing methods can eliminate risks associated with incapacitating illness and the loss of essential skills. They rely both on the opinions of the Age 60 Exemption Panel and on docket submissions from doctors, pilots, companies and organizations who support the exemptions. Petitioners also criticize the methods, analyses and, in some case, motives of the sources upon which the FAA relied. A detailed recitation of the petitioners' evidence would serve no useful purpose, however. We think it is clear, with respect to the issue whether the requested exemptions would add any risk of incapacitation or error due to the deterioration of piloting skills, that substantial evidence supported the FAA's conclusions. Were we to conduct a de novo review of the docket, we might conceivably side with the petitioners. But this is not our job. When the question is whether the petitioners' protocol eliminates all of the incremental risk associated with sudden incapacitation or undetected deterioration of skills among pilots over sixty, a substantial body of medical opinion continues "to doubt the feasibility" of the project. See Starr, 589 F.2d at 314.
Petitioners' submissions to this court contain thinly veiled allegations that the FAA's position with respect to the fitness of pilots over sixty has been shaped by improper influences. They suggest that the original rule may have been tainted by ex parte communications between the FAA Administrator and the president of a major airline, that the FAA's loss of the original docket may have been intentional, that the FAA's abandonment of the 1982 proposal to study the age sixty rule through selective exemptions was influenced by ex parte communications with the Air Line Pilots Association and that the FAA's strict adherence to the age sixty rule since 1960 reflects the FAA's hidden agenda — a concern for airline costs rather than for passenger safety. Petitioners proffer no factual support for these allegations, however, beyond some handwritten notes by Frank Austin, written during his tenure as Federal Air Surgeon, which characterized the rule as "operational" or "economic" rather than medical. See Joint Appendix at 203-06. Moreover, the petitioners have not argued that any ex parte communications, which are statutorily proscribed only where rulemaking is required to be "on the record," constituted the kind of "extremely rare" procedural transgression that could "justify a court in overturning agency action because of a failure to employ procedures beyond those required by the statute." Vermont Yankee Nuclear Power Corp. v. National Resources Defense Council, 435 U.S. 519, 524, 98 S.Ct. 1197, 1202, 55 L.Ed.2d 460 (1977). Compare National Small Shipments Traffic Conference v. ICC, 590 F.2d 345, 350 (D.C. Cir. 1978). Nor have petitioners attempted to sustain the heavy burden of establishing that decisions concerning the age sixty rule have been made by officials who should have been disqualified on account of their bias against older pilots. See FTC v. Cement Institute, 333 U.S. 683, 700-03, 68 S.Ct. 793, 803-05, 92 L.Ed. 1010 (1948); Starr, 589 F.2d at 315.
Petitioners sharply criticize passages of the FAA order that we have omitted from our summary of the FAA's strongest factual support. Some of these criticisms appear to have merit. For example, the FAA asserts that "population variability with respect to disability, death and deterioration increases with age" and that this leads to "a progressive decrease in [the] discriminatory power" of predictive indices. FAA Order at 12. The logic of the FAA's position on this point is elusive. Increases in the range of observations would seem to have no bearing whatever on the discriminatory power of the protocol in the cases that matter — those at the margin between fitness and unfitness — and increased dispersion could well decrease the number of marginal cases. In a similar vein, the FAA asserts that differences between the testing protocol developed by the Age 60 Exemption Panel and a series of tests that was devised for the same purpose at the request of the House Select Committee on Aging support its position "that in fact, what the petitioners profess their protocol does, cannot yet be done." FAA Order at 13. We, like the petitioners, have difficulty seeing how the mere existence of two protocols supports the FAA's argument. Any suggestion of a lack of consensus is undermined by the prior panel's unanimous support for the petitioners' protocol, which one of the members of the 1985 panel helped to develop. Indeed, another member of the earlier panel has represented to Congress that the two protocols "are essentially equivalent." Letter from Dr. T. Franklin Williams to the Honorable Edward R. Roybal, Chairman of the House Select Committee on Aging (Sept. 23, 1986), reprinted in Joint Appendix at 220.
If the FAA's conclusion that the petitioners' testing plan would not screen out all of the risks associated with aging rested in any large part on these questionable claims, its position might not be sustainable. However, we believe that the other evidence upon which the FAA relied provides a substantial basis for its conclusion with respect to the first issue.
B.
Turning to the second formulation of the petitioners' claim, which introduces the decrease in risk associated with experience into the equation, we find the FAA's rejection of the petitioners' position far less persuasive. The order's entire discussion of the effect of experience on its analysis of the public's interest in strict adherence to the age sixty rule is confined to the following two paragraphs:
The petitioners have failed to show that what they request is in the public interest. The FAA does not agree that it is in the public interest to suppose an increase in safety will result from the use of their services as pilots beyond age 60, thereby outweighs the potential safety hazard to the public from increased risk of incapacitation and diminished performance.
Petitioners argue that the loss of experienced pilots to the airlines because of retirements mandated by the Age 60 rule is resulting in a shortage of pilots which is forcing the airlines to lower their standards and to hire less experienced and qualified pilots. Thus, petitioners argue that granting exemptions to the Age 60 rule would ease this shortage and increase the level of safety. However, the exhibits presented by the petitioners in support of this argument make it clear that current lowering of minimum experience requirements at some airlines is the result of a number of factors; including fewer civilian pilot trainees, fewer pilots leaving the military, increasing numbers of senior pilots retiring voluntarily before age 60, and the expansion of the airlines after deregulation.
FAA Order at 11.
This represents notably less than the response to which the petitioners are entitled. On the second issue, the FAA fails to present findings of fact supported by substantial evidence (or even by the presumably lesser quantum of evidence required to avoid arbitrariness and capriciousness) or to identify the governing principles and set forth a "rational connection between the facts found and the choice made." Burlington Truck Lines, Inc., 371 U.S. at 168, 83 S.Ct. at 245. First, the FAA cites no evidence to justify its offhand dismissal of the petitioners' evidence on experience. Petitioners offer comparisons, based on data taken from the FAA Statistical Handbook of Aviation and National Transportation, of accident rates by age group for pilots with commercial and air transport certificates for the period from 1982 through 1985. Joint Appendix at 207-08. These data show below average accident rates for pilots in the 55-year old to 59-year old age group and the 60-plus age group (a group that presumably piloted only flights not subject to the age sixty rule). These data, taken together with related, less persuasive submissions, hardly make an airtight case for the petitioners; there is no discussion of the statistical significance of these comparisons or of possible distortions due to differences across age cohorts in factors such as miles flown, routes, schedules or equipment used. However, the comparisons are at least suggestive; they warranted something more than the FAA's summary dismissal of the experience argument as based on mere supposition. The FAA does address the petitioners' claim that rigid enforcement of the age sixty rule reduces the average experience level among active pilots. But its observation that other factors have contributed to an overall decline in experience has no apparent connection to the effect on safety of refusing exemptions to particular older pilots. A person charged with maintaining the water level in a cistern could not ordinarily point to the existence of inaccessible leaks as a logical justification for ignoring the accessible ones.
In its brief, the FAA seeks to fill this gap with a citation to a district court's finding of fact in a twenty-eight year old preliminary injunction case. See Air Line Pilots Ass'n, Int'l v. Quesada, 182 F. Supp. 595, 596-97 (S.D.N.Y.) ("[T]here is increasing evidence that as men grow older, they unfortunately experience deterioration for which maturity, experience, judgment and skill cannot compensate adequately."), aff'd, 276 F.2d 892 (2d Cir. 1960), cert. denied, 366 U.S. 962, 81 S.Ct. 1923, 6 L.Ed.2d 1254 (1961). Even if it were permissible for the FAA's brief to correct for factual deficiencies in its order, we would doubt the relevance of a district court's summary of a record produced in 1960.
In their briefs petitioners purport to draw support for their experience claim from (1) a purported correlation between increasing median age for pilots (supported only by data from United Air Lines operations) and decreases in accidents per million aircraft miles (supported by data from all United States air carriers), see Joint Appendix at 147-48; and (2) a docket submission from K. Warner Schaie, Professor of Human Development and Psychology, criticizing the FAA's rigid enforcement of the age sixty rule during a period when airlines are encountering a shortage of experienced pilots, reprinted in Joint Appendix at 247.
The FAA might have attempted to justify its rejection of the petitioners' second argument without refuting their factual claims about experience by arguing that the public interest standard encompasses administrative and economic concerns as well as safety concerns. At some point, the relevance of these concerns seems logically inescapable. Cf. Starr, 589 F.2d at 312 (discussing possible administrative justifications for no-exemption policy with respect to age sixty rule). But in the order before us the legal response to the petitioners' experience argument is no better developed than the factual response. The FAA's order recites the statutory requirement set forth in section 601(b), 49 U.S.C. App. § 1421(b) (1982), that the FAA, in formulating its regulations, give "full consideration" to air carriers' duty to maintain the "highest possible degree of safety in the public interest." FAA Order at 9. It also paraphrases section 601(c), 49 U.S.C. App. § 1421(c) (1982), which authorizes exemptions from the regulations if the Administrator "finds that such action would be in the public interest." Id. The order also summarizes the regulation governing exemptions, but this regulation sheds little light on what factors, other than safety, might enter into the FAA's vision of the public interest and how any necessary tradeoffs might be made.
Section 107 of the Airline Deregulation Act of 1978 underscored the FAA's safety mandate, stating that Congress intended for deregulation to go forward with "no diminution of the high standard of safety in air transportation." 49 U.S.C.App. § 1307(a) (1982).
Section 11.25 of the FAA regulations directs petitioners to state
the reasons why the granting of the request would be in the public interest and . . . why the exemption would not adversely affect safety or the action to be taken by the petitioner to provide a level of safety equal to that provided by the rule from which the exception is sought.
14 C.F.R. § 11.25 (1988).
The scant discussion of the exemption standard that the order does contain suggests that, at least in the present context, safety concerns predominate over administrative and economic factors in the FAA's application of the section 601(c) public interest standard. Near the conclusion of the order, the FAA indicates that its "foremost duty, as mandated by Congress in the Act, is to promote safety of flight for civil aircraft." FAA Order at 16. The order appears to base its denial of the exemptions purely on safety concerns. This may or may not be the FAA's considered view of the public interest standard for exemptions. In fact, as the paragraphs quoted at the outset of this discussion indicate, the petitioners' experience argument received very little consideration at any level — a failing for which petitioners may share some of the blame since their petition did not clearly separate the two distinct arguments as clearly as we have attempted to do here.
Petitioners also maintain that the FAA's policy of refusing to grant exemptions to the age sixty rule runs counter to a broadly applicable principle that individualized assessment must precede disqualification from any employment on account of age. Petitioners purport to derive this broad principle from Western Air Lines, Inc. v. Criswell, 472 U.S. 400, 105 S.Ct. 2743, 86 L.Ed.2d 321 (1985). Western Air Lines, however, dealt with challenges to a jury verdict under the Age Discrimination in Employment Act ("ADEA"). The Supreme Court held that an airline had failed to justify its policy of requiring flight engineers to retire at age sixty under the ADEA's narrow bona fide occupational qualification ("BFOQ") exception to the general bar on forced retirement before age 70. Western Air Lines, like other cases involving ADEA challenges to age cutoffs for flight personnel, distinguished between the validity of the age sixty rule as an exercise of the FAA's authority under the Federal Aviation Act and the validity of applying similar restrictions to flight personnel outside the ambit of Part 121 under the BFOQ exception. 472 U.S. at 417-18, 105 S.Ct. at 2753-54; EEOC v. Boeing Co., 843 F.2d 1213, 1218-19 (9th Cir. 1988); see also Trans World Airlines, Inc. v. Thurston, 469 U.S. 111, 123 n. 17, 105 S.Ct. 613, 622 n. 17, 83 L.Ed.2d 523 (1985). The same distinction has been recognized in three prior cases reviewing refusals of petitions for exemption to the age sixty rule. Keating, 610 F.2d at 613; Rombough, 594 F.2d at 898-99; Starr, 589 F.2d at 313-14. Because the consistency of the age sixty rule with the ADEA is not presently before us, the principles governing jury consideration of BFOQ questions are irrelevant here.
We are unprepared to say, without a more developed account of the FAA's view of the legal standard governing exemption determinations, what the permissible scope of administrative and economic factors might be. As the agency charged with administering the Federal Aviation Act, the FAA's interpretation, once articulated, will be entitled to considerable deference. This deference, however, does not lighten the FAA's burden to establish that its decision is logically connected to a sustainable reading of its mandate. Moreover, such deference should not be equated with a license to issue inconsistent determinations. See Airmark Corp. v. FAA, 758 F.2d 685, 691-95 (D.C. Cir. 1985). We note that the FAA has shown an increased willingness in recent years to issue "special" medical certificates under section 67.19 of the regulations (the functional equivalent of a second exemption mechanism) to pilots otherwise disqualified by episodes of heart disease or alcoholism. See 14 C.F.R. §§ 67.13, .19 (1988). Part of the rationale for a statutory exemption procedure, no doubt, is to allow the FAA flexibility in the application of its rules; this flexibility arguably entails authority to determine that some rules can be relaxed more freely than others. See Starr, 589 F.2d at 313. The statutory justifications for these distinctions, however, should be just as susceptible to rational explanation as the bases for any other agency decision. We think it essential that the construction of its statutory responsibilities under which the FAA seeks to justify its ultimate decision on the present exemption make sense of these special issuances as well.
IV.
The FAA's consideration of the petition for exemption at issue here was incomplete. The FAA adduced substantial evidence supporting its rejection of the contention that the petitioners' protocol, combined with existing methods of operational testing, would screen out all increased risks of incapacitation or undetected skill deterioration among pilots older than sixty. However, the FAA failed to set forth a sufficient factual or legal basis for its rejection of the petitioners' claim that older pilots' edge in experience offsets any undetected physical losses. We therefore vacate the denial of the exemptions and remand to the FAA for further proceedings to provide findings and explanations addressing the deficiencies we have noted and for other appropriate proceedings not inconsistent with this order.
VACATED AND REMANDED.