Opinion for the Court filed by Circuit Judge SILBERMAN.
SILBERMAN, Circuit Judge:
Petitioners City of Los Angeles and the Air Transport Association of America, an airline trade association, each challenge the Department of Transportation’s Final Policy Regarding Airport Rates and Charges. We grant the petitions for review and remand.
I.
Airports are required by statute to charge aeronautical users reasonable fees.
Section 511 of the Airports and Airways Improvements Act, codified at 49 U.S.C. § 47107 (1994), requires an airport that accepts federal grant money (or land) to assure that the airport will be available for public use on reasonable conditions and without unjust discrimination; this assurance has been interpreted to include a requirement that the airport’s fees be reasonable.
See New England Legal Found. v. Massachusetts Port Auth.,
883 F.2d 157, 169-70 (1st Cir.1989). Similarly, the Anti-Head Tax Act allows a publicly owned airport to collect only reasonable landing fees and charges from aeronautical users.
See
49 U.S.C. § 40116(e)(2) (1994);
Northwest Airlines v. County of Kent,
510 U.S. 355, 366, 114 S.Ct. 855, 862-63, 127 L.Ed.2d 183 (1994). Although the Department of Transportation (DOT) has adjudicated disputes over the reasonableness of airport fees under these statutes,
see, e.g., New England Legal Found.,
883 F.2d at 163-66, until recently it never promulgated regulations defining what it thought to be reasonable airport fees. Airlines and airports did, however, take these issues to court.
See, e.g., Northwest Airlines,
510 U.S. 355, 114 S.Ct. 855;
New England Legal Found.,
883 F.2d 157;
Indianapolis Airport Auth. v. American Airlines,
733 F.2d 1262 (7th Cir.1984),
disapproved by Northwest Airlines.
In August 1994, Congress enacted § 113 of the Federal Aviation Administration Authorization Act, codified at 49 U.S.C. § 47129 (1994), which required the Secretary of Transportation to publish “final regulations, policy statements, or guidelines establishing — ... the standards or guidelines that shall be used by the Secretary in determining under this section whether an airport fee is reasonable.”
Id.
§ 47129(b)(2). That section also created an expedited administrative procedure in which an airline may challenge before the Secretary the reasonableness of an airport fee.
See City of Los Angeles Dep’t of Airports v. Department of Transp.,
103 F.3d 1027, 1030 (D.C.Cir.1997)
(LAX I).
In such proceedings, only the reasonableness of the challenged fee is open to question; the Secretary “shall not set the level of the fee.” 49 U.S.C. § 47129(a)(3).
The Secretary adopted an “Interim Policy” in February 1995 which required that aeronautical fees be capped by the cost to airports of providing aeronautical facilities and services. The interim policy also required that airfield assets be valued at their historic
cost in determining the total costs that could be recovered from fees. Airports complained that the Interim Policy would force them to change their operating methods and would damage their ability to finance improvements, since they had commonly based fees for certain aeronautical facilities (such as terminals) on something other than historic cost.
In June 1996, after receiving comments on a “Supplemental Proposed Policy,” the Secretary published a regulation entitled the “Final Policy Regarding Airport Rates and Charges.” The Final Policy, unlike the Interim Policy, distinguishes between airfield and non-airfield fees. For airfield fees— aeronautical fees charged for the use of the runways, taxiways, ramps, aprons, and roadway land — the Final Policy maintains the Interim Policy’s requirements that aeronautical fees be capped at actual cost to the airport, and that airfield assets be valued according to their historic cost. For non-airfield fees — aeronautical fees charged for the use of all other aeronautical facilities and services, including terminals, hangars, cargo space, and maintenance — the Final Policy instead permits airports to use “any reasonable methodology.”
According to the Secretary, requiring that assets be valued at historic cost in setting airfield fees is consistent with current practice, permits airports to recover their out-of-pocket costs, and, since historic cost is easily determined from accounting records, will help avoid controversies and be easier to administer. The Secretary rejected fair market valuation as a methodology for valuing assets in determining airfield fees because the Secretary thought that airports do not have “opportunity costs” in airfield assets since they are committed to continued operations. The Secretary also thought that the Final Policy’s treatment of non-airfield fees is consistent with past practice (which had not resulted in disputes), and would permit pricing to reflect the differences among non-airfield facilities. The Secretary believed it unlikely that airports can or will exercise market power in setting non-airfield fees.
The Final Policy also permits airports to include in airfield fees a charge for imputed interest on those aeronautical funds reinvested in the airfield,
unless
the invested funds derive from airfield fees. A charge for imputed interest, according to the Secretary, lets airports recover their opportunity costs in the airfield (if any) and compensates for inflation. But imputed interest cannot be charged for airfield revenues reinvested in the airfield because the Secretary believes airfield users might then be financing airfield investment twice: once in the form of otherwise reasonable airfield fees that turned out to generate excess revenue, and once on the imputed interest charge on the investments made with that revenue.
Petitioners attack the Final Policy from both sides. The Air Transport Association of America (ATA), on behalf of its members, challenges what it sees as the Final Policy’s complete deregulation of non-airfield fees. The City of Los Angeles challenges the Final Policy’s requirement of historic cost for airfield fees.
II.
Anterior analytically to the main dispute between the two petitioners and the Secretary — whether the regulation’s
disparate treatment of airfield and non-airfield fees is arbitrary and capricious under the Administrative Procedure Act (APA) — is the nature of the Secretary’s authority and obligation to issue regulations (guidelines) which set forth a particular methodology for determining the reasonableness of fees. Los Angeles contends the Secretary has no such power, whereas ATA claims that not only does he have that authority, he must use it in a meaningful manner for both airfield and non-airfield fees.
Los Angeles argues that the statute obliges the Secretary to take a relatively passive role: he “may only determine whether [a] fee is reasonable or unreasonable and shall not set the level of the fee.” 49 U.S.C. § 47129(a)(3). Emphasizing the differences that exist among airports regarding airfield and non-airfield facilities and financial struc
tares, it contends that Congress, understanding those differences, intended that the Secretary’s role be limited to determining whether a particular airport’s fees meet an overall (bottom line) reasonableness standard.
Cf. Federal Power Comm’n v. Natural Gas Pipeline Co.,
315 U.S. 575, 586, 62 S.Ct. 736, 743, 86 L.Ed. 1037 (1942). Each airport is therefore entitled to use
any
reasonable methodology, and the Secretary lacked authority to require airports to use historic cost as a basis for airfield fees.
We think Los Angeles’ position on this issue is not very persuasive. Section 113, it will be recalled, obliges the Secretary to publish “final regulations, policy statements, or guidelines establishing — ... the standards or guidelines that shall be used by the Secretary in determining under this section whether an airport fee is reasonable.” 49 U.S.C. § 47129(b)(2). That provision certainly implies that Congress intended the Secretary to fashion a quasi-legislative uniform approach to measuring the reasonableness of airport fees. That the Secretary is not authorized actually to set the fee itself means that each airport necessarily has some discretion in the application of the Secretary’s guidelines, but it hardly can be read to mean that the Secretary was not authorized to require a particular methodology for
all
airports — so long as that methodology is itself reasonable.
The more difficult question, posed by ATA, is whether the Secretary’s approach to non-airfield fees meets its obligation under § 113. The Secretary’s “guideline” seems to be missing a “line.” The regulation merely states that
any
reasonable methodology will serve as a basis for non-airfield fees. That concept does not seem to add much — if anything — to the statutory requirement that airport fees be reasonable.
It may well be that the Secretary is not required to insist on a single methodology. Perhaps he could indicate that several different methodologies, depending on the circumstances, would be authorized. We need not decide whether the Secretary’s treatment of non-airfield fees actually violates his statutory obligation to set forth guidelines, however, because as will be seen, we conclude that in any event his approach is arbitrary and capricious.
III.
Even if the Secretary’s regulation as applied to non-airfield fees satisfied § 113’s publication obligation, ATA asserts that it amounts to
de facto
deregulation of those fees — permitting airports to use their market power to charge airlines excessive amounts. Not surprisingly, ATA wants the Secretary’s guidelines for valuing assets in determining airfield fees, based on historic cost, extended to the non-airfield facilities. ATA complains that DOT has not imposed any sort of meaningful restraint on non-airfield revenues, leaving fees to be set by the market. That posture, it is argued, is by definition inconsistent with the reasonableness requirement in the Anti-Head Tax Act, and violates the Secretary’s duty under § 113. ATA claims that it is well accepted that an agency responsible for ensuring reasonable rates in industries with anticompetitive conditions may not rely on the market to accomplish that goal.
See, e.g., Federal Power Comm’n v. Texaco, Inc.,
417 U.S. 380, 397-99, 94 S.Ct. 2315, 2326-27, 41 L.Ed.2d 141 (1974);
Farmers Union Cent. Exch. v. FERC,
734 F.2d 1486, 1509-10 (D.C.Cir.),
cert. denied,
469 U.S. 1034, 105 S.Ct. 507, 83 L.Ed.2d 398 (1984);
cf. Elizabethtown Gas Co. v. FERC,
10 F.3d 866, 870 (D.C.Cir.1993). Certainly that is so regarding non-airfield fees because, according to ATA, airports enjoy virtually the same market power over non-airfield services — terminals, hangers, cargo space maintenance, etc. — that they do over the airfields. At minimum, therefore, the distinction drawn by the Secretary between the charges for the two types of assets is arbitrary and capricious.
If ATA is correct that the Secretary’s regulation imposes no real restraint on airport charges for non-airfield assets and that airports exercise the same, or virtually the same, market power over those assets, then it seems to us that the Secretary’s tight regulation of airfield fees — tying them to assets valued at historic cost and capping them
at actual cost — is something of an illusion. Most aeronautical users, it would seem, need both landing fields and non-airfield services. The entire package usually is sold to aeronautical users in one overall fee. Under the “one lump” theory of antitrust law — the proposition that there is only one monopoly rent to be gained from the sale of an end product,
see Western Resources, Inc. v. Surface Transp. Bd.,
109 F.3d 782, 784 (D.C.Cir.1997) (citing 3 Phillip Areeda & Donald F. Turner, Antitrust Law § 725b, at 199 (1978)) — any market power that an airport wants to exercise over airfield fees theoretically may be exercised by increasing the non-airfield portion of the fee.
The Secretary takes no position on the question whether airports have market power over non-airfield facilities, but it is his belief that even if an individual airport does have such power, it will not use it. His explanation, set forth in the basis for the Final Policy and in the briefs before us, is that publicly owned airports do not seek to maximize profits because they do not have the same economic incentives as private investors. And airport authorities may not divert revenues from airports to fund general government facilities, so there is little reason to charge airlines excessive prices. Moreover, even if an airport authority had a desire to build the world’s fanciest airport,
see Indianapolis Airport Auth.,
733 F.2d at 1268, it would be somewhat constrained by competition among airports, at least to be the location of passenger and cargo hubs.
The Secretary thought past experience supported this position: that negotiations for non-airfield fees had not generated controversies, and non-airfield fees had typically produced reasonable results by agreement between airlines and the airports. But as ATA observes, past litigation had sometimes involved total airport fees, combining both airfield and non-airfield fees.
See Northwest Airlines v. County of Kent,
738 F.Supp. 1112, 1115 (W.D.Mich.1990),
aff'd,
955 F.2d 1054 (6th Cir.1992),
aff'd,
510 U.S. 355, 114 S.Ct. 855, 127 L.Ed.2d 183 (1994);
cf. Indianapolis Airport Auth.,
733 F.2d at 1265. Be that as it may, experience prior to the Secretary’s regulation is not a particularly reliable guide since airports may well have thought that the statutory limitation (reasonableness) on all airport fees prior to 1994 implied some sort of cost-based methodology.
See, e.g., New England Legal Found.,
883 F.2d at 169;
Indianapolis Airport Auth.,
733 F.2d at 1271. For the first time, as ATA points out, the Secretary’s regulation, allowing
any
reasonable methodology, gives airports much greater assurance that they are free to charge more for non-airfield fees, and there is little reason to expect they will not. Indeed, by linking airfield fees to historic cost and capping airfield fees at total cost but permitting any reasonable methodology and imposing no cap for non-airfield fees, the regulation certainly implies that non-airfield fees need not be cost-based. Nor is the fact that airlines typically have reached agreement with airports as to these fees of any special significance. Airlines have incentives to enter into agreements with airports, even if they are monopolists. And an individual airline may not have a large incentive to challenge the reasonableness of a monopoly rent, since its competitors at an airport would also be paying the same monopoly rent (by virtue of the prohibition against discriminating among aeronautical users), and any benefit of the litigation would be dispersed among the airlines.
To be sure, the Final Policy allows airlines to challenge the reasonableness of an airport’s methodology, so presumably an airline can challenge a non-airfield fee it believes is excessive. But the regulation provides that the Secretary will “not [] consider complaints about the reasonableness of fees set by agreement if filed by parties to the agreement,” nor does he “expect to investigate routinely fees set by agreement.”
Airlines
are presumably free not to enter into agreements with airports, thereby remaining able to challenge particular non-airfield fees, but we are told that there are certain business reasons that induce airlines to have fee agreements in place. At oral argument, counsel for the Secretary seemed to indicate that airlines would be permitted to enter into agreements under protest and then file complaints under § 113. But that explanation seems inconsistent with the regulation’s language, which states that “[t]he threat of a complaint could discourage airport proprietors from putting forward their best offers in negotiations.” It is thus not at all clear whether the Final Policy provides airlines with an adequate remedy to challenge non-airfield fees.
Aside from these procedural ambiguities, the Final Policy provides no real guidance as to how the Secretary will determine reasonableness. As we have noted, § 113 may actually require the Secretary to set forth a full quasi-legislative standard rather than developing those standards through a case-by-case approach. Even if the Secretary has more flexible authority under the statute, his regulation surely is inadequate under the APA. The Final Policy indicates that the Secretary will consider “extraordinary situationfs]” (whatever that may mean) in the expedited proceedings under § 113; that the Secretary will consider the reasonableness of non-airfield fees “over the long term and not on the basis of a single year’s results”; and that the Secretary will investigate only where there is “a ease of progressive accumulation of surplus aeronautical revenue.” That the Secretary believes that only a long-term accumulation of surplus aeronautical revenue is evidence of unreasonable non-airfield fees implies that the only legal limit to airport charges is the ability to spend revenues fast enough on airport improvements, including those for non-aeronautieal concessions. That approach could lead to competition to build the Taj Mahal of airports.
See Indianapolis Airport Auth.,
733 F.2d at 1268.
It may well be that, as the Secretary suggests, airports would face some restraint on the exercise of any market power that they might have. But we do not think the Secretary has adequately explained how those generally unspecified restraints will ensure, in the absence of meaningful guidelines, that non-airfield fees are reasonable. Even assuming, however, that the Secretary on remand can with greater precision identify these restraints and justify notions of reasonableness of some or all aeronautical fees using concepts less restrictive than historic cost, there remains the most serious logical problem with his regulation — which we simply cannot accept. His explanation for the distinction drawn between airfield and non-airfield fees is internally inconsistent.
Cf. Motor Vehicle Mfrs. Ass’n of United States v. State Farm Mut. Auto. Ins. Co.,
463 U.S. 29, 48, 103 S.Ct. 2856, 2869, 77 L.Ed.2d 443 (1983). The obvious difficulty, for instance, with the Secretary’s reliance on his perception that public airports are not profit maximizers is that the proposition, if correct, would apply equally to airfield charges, so it hardly can be relied upon to support a different approach to non-airfield fees. Even the specific alleged restraint on the airports’ use of market power to charge excessive non-airfield fees — that airports compete to be hub locations — would, if significant, also restrain airfield charges. Surely airlines, when determining where to locate hub facilities, are not indifferent to airfield charges, since they channel their flights in and out of those hub facilities. We are obliged therefore to vacate the rule and remand to the Secretary.
IV.
Los Angeles makes in part the same point: that the Secretary’s disparate treatment of airfield and non-airfield charges is arbitrary and capricious, albeit from the perspective of applauding the Secretary’s approach to non-airfield charges. The record does not reveal why Los Angeles is the sole airport owner to challenge the regulation. We know only that it is one of the few (if not the only) municipalities that stopped taking federal grant money — which, as noted, carries with it an obligation to continue using airport facilities as an airport — in the early 1990s.
In any event, Los Angeles’ primary attack is directed at the Secretary’s requirement of valuing assets at historic cost in calculating airfield charges. The Secretary was careful not to make the same mistake that caused us to reject his prior determination that Los Angeles was obliged to use historic cost as the basis for airfield fees.
See LAX I,
103 F.2d at 1032. Los Angeles then, as now, argued that it should be entitled to use the fair market value of the land — which in downtown Los Angeles is quite dear — as the basis for those charges because it accurately reflects the opportunity cost of devoting LAX to airport use. In the prior proceeding, the Secretary had simply assumed, erroneously we held, that the Anti-Head Tax Act and § 113 require historic cost valuation. This time he did not. He did observe that since airports are obliged to use their property as an airport, the concept of opportunity cost, and therefore fair market value, does not quite fit. (That logic, however, may not apply to Los Angeles.)
The Secretary explained why he thought historic cost was generally preferable to other valuation bases such as fair market value: it is simply easier to administer and less likely to generate controversies. We observed in
LAX I,
however, that valuing land (the principal airfield asset) does not present the same difficulties as valuing, for example, utility plants and equipment, for “there is no need to reconstruct a hypothetical asset in order to account for technological changes, and there is often (indeed, perhaps usually) a ready market in parcels sufficiently comparable for a professional appraiser to extrapolate with some confidence.” 103 F.3d at 1033;
cf. Duquesne Light Co. v. Barasch,
488 U.S. 299, 308-09, 109 S.Ct. 609, 615-16, 102 L.Ed.2d 646 (1989). Nor has the Secretary actually given any reason to think controversies will be less likely to occur using historic cost. In
LAX I,
the airlines did not contest the accuracy of the appraisal obtained by Los Angeles for the land at LAX, 103 F.3d at 1033; in a later proceeding the airlines’ appraisal of the land at LAX was actually
higher
than the airport’s; nor is there a guarantee that historic cost will cause few or no disputes. And, so far as we can tell, airports need only appraise the market value of the land once in order to bring it into the rate base.
See id.
Although permitting a methodology other than historic cost valuation might, in the Secretary’s words, “turn landing fee disputes into debates between real estate appraisal experts with [the Secretary] in the role of referee,” it is not apparent why this justifies requiring historic cost: the Secretary can adopt any number of fairly simple procedures, such as the “final offer” device used in baseball arbitration,
see Southern Pacific Transp. Co. v. ICC,
69 F.3d 583, 585 (D.C.Cir.1995), for administering such disputes.
Even if ease of determination were an appropriate ground for requiring historic cost valuation of assets in determining airfield fees, however, the Secretary’s use of it here undermines his treatment of non-airfield fees. By permitting non-airfield fees
to
be set according to any reasonable methodology, the Final Policy leaves airports free to choose any valuation method, including fair market value; the regulation thus permits airports, in the context of non-airfield fees, to use methodologies that the Secretary has determined are too difficult to use in the context of airfield fees. As Los Angeles puts it, the Secretary simply has not explained why fair market valuation may be appropriate for other portions of the airport, but too difficult to use in valuing airfield assets.
Y.
There remains the Secretary’s treatment of “imputed interest” which, if anything, is even more puzzling to us than the rest of his rule. The Secretary’s Final Policy permits an airport to include in its airfield fees a charge for imputed interest on funds the airport reinvests in its airfield (unless, of course, debt service costs are already included as explicit interest). But airports are only permitted to include such imputed interest in airfield fees insofar as the reinvested funds stem from non-airfield fees, not from airfield fees. The Secretary’s rule permits the for
mer, but not the latter, because he reasoned that to charge this cost of capital to the airlines for funds derived from airfield fees would be double charging: the airlines would get stuck with both the excess revenue and the imputed interest on that excess.
ATA objects to paying any imputed interest as part of airfield fees no matter the source of the funds, claiming in effect that, if it is double charging to charge imputed interest on the funds derived from airfield fees, it is equally double charging to include imputed interest on funds stemming from non-airfield fees. We agree, in a sense, because we do not understand what the Secretary means by double charging. Airports receive revenues from essentially three sources: airfield fees, non-airfield fees, and concession leases. Although airports cannot invest outside the airport, they can reinvest, and indeed may be obliged to reinvest, in all three airport activities — certainly at least the first two. And they have discretion as to the object of their investment, which makes the revenue airports receive from the three sources fungible. Therefore, as the Secretary suggested, there is an opportunity cost in reinvesting in one activity rather than the other two: the forgone revenue that would be received had the airport invested those funds in the other two. There may be a case to be made, similar to the Secretary’s argument regarding most airports’ lack of opportunity costs generally, that because airports do not have wide discretion as to where they put surplus funds (they cannot invest off airport), their cost of capital (opportunity cost) is less than would be true of another business, but we frankly do not understand the distinction the Secretary makes between funds that stem from airfield fees and those that come from non-airfield fees.
* *
*
* * *
The Secretary seems to have come up with a rule that represents an overall rough compromise between the interests of most airports and carriers. Unfortunately it is much too rough to withstand judicial review under the APA. Accordingly, we vacate the rule and remand this proceeding to the Secretary.