The Indianapolis Airport Authority appeals from a decision invalidating the user fees that it imposes on airlines; the appellees are six airlines that among them carry more than 90 percent of the passengers who use the Indianapolis International Airport. The 15-year leases under which each of the airlines operated expired on August 31, 1980, and because the parties had been unable to agree on terms for new leases, the airport authority — a local governmental body established pursuant to the Indiana Airport Authorities Act, Ind.Code §§ 8-22-3-1 et seq. — enacted an ordinance (later amended) setting new fees effective September 1, 1980. When the airlines refused to pay the new fees, which were almost double those in the expired leases, and continued paying at the old level, the Authority brought this diversity action to collect *1265the difference between the new fees and the old — some $2 million. In the district court the airlines successfully defended their refusal to comply with the ordinance on the ground that the fee schedule in the ordinance was unreasonable on a vanety of statutory (state and federal) and constitutional grounds, and they also persuaded the court that they were holdover tenants entitled under Indiana law to contmue paying at the old lease rate until the Authority stopped accepting payments at that rate.
rr,, . . , ,, ,, . , The mam issue is whether the airport . ,,• v , , £ n authority, m setting a new schedule of fees j. ,, . r i, for the airlines, could disregard the revenues it obtains from airport concessionaires, in particular several car rental agencies and the operator of the airport’s parking lot. The ordinance allocates the annual costs of operating the airport among the different classes of user — mainly interstate airlines, operators of private planes (“general aviation”), and concessionaires — -largely on the basis of how much runway, hangar, terminal, and other indoor and outdoor airport space each class uses. (For services such as firefighting that have no fixed locale — obviously, the firemen and their equipment go wherever the fire is — a different method of allocation is used that we discuss later.) Since the concessionaires use much less space than the airlines, only a modest fraction of the airport’s costs was allocated to them — in round numbers, $100,000 to the car rental agencies and $900,000 to the operator of the parking lot — compared to $3 million to the airlines, The ordinance requires the airlines to pay landing fees and other charges calculated to yield the full $3 million, even though the airport gets from its concessionaires a rental income that greatly exceeds the costs allocated to the concessionaires — about $3.5 million from the ear rental agencies and the parking lot alone, compared to costs as we have said of about $1 million, The ordinance thus is calculated to yield the airport a total income substantially greater than its total costs, the excess being approximated by the difference between the airport costs allocated to the concessionaires and the airport rentals they pay.
The reasonableness of the concession rentals themselves is not in issue in this case — ordy bbe reasonableness of the fees cbarged the airlines. The basis of the airlineg, complaint about those feeSf however, ig that the airport ig required to and has faüed to take itg eoncession rentals into account ¡n determining what fees to impose on the air]jnes
The Indiana Airport Authorities Act authorizes airport authorities, such as the „ , ^ ’ appellant, To adopt a schedule of reason-ff , ’ , f „ , , able charges and to collect them from all „ ° . ?sers °f facilities and services within the district. IndCode § 8-22-3-n(9)' However’ reasonableness is not defmed m the statute’ and the statute has not been interpreted by the Indiana courts with reference to the issues in this case. The Federal Anti-Head-Tax Act forbids any state agency to levy or collect a tax, fee, head charge, or other charge, directly or indirectly, on persons traveling in air cornrnerce or on the carriage of persons travelhig in air commerce ...,’ 49 U.S.C. § 1513(a), other than “reasonable rental charges, landing fees, and other service charges from aircraft operators for the use °f airport facilities,” 49 U.S.C. § 1513(b). Again, reasonableness is not defined, but the statute has a history and a context that enable us to give meaning to the term. Airport authorities, to raise revenues, had taken to imposing “head taxes” on passengers emplaning at their airports. Congress decided that “the head tax is an unnecessary burden on interstate commerce, that it is discriminatory, and that it has a stifling effect on air transportation,” most of which, of course, is interstate. H.R.Rep. No. 157, 93d Cong., 1st Sess. 4 (1973). We may assume that what is unreasonable under the federal act is also unreasonable under the state act; but, in any event, if there is a clash, the federal act must of course prevail. Another federal act is invoked, the Airport and Airway Development Act of 1970, 49 U.S.C. § 1718(a)(1) (1976 ed.), now 49 U.S.C. § 2210(a)(1), *1266which requires that airports receiving federal subsidies — such as the Indianapolis airport — be “available for public use on fair and reasonable terms and without unjust discrimination____” It is unclear whether this act was intended to be enforceable by airport users, such as the appellee airlines; but it will not be necessary in this case to resolve this question, or determine whether the challenged user fees are unreasonable under this act.
Besides the Federal Anti-Head-Tax Act and important to understanding it, the appellees invoke the commerce clause of the Constitution (Art. I, § 8, cl. 3), which has been interpreted to forbid the states to discriminate against interstate commerce. See, e.g., Southern Pac. Co. v. Arizona, 325 U.S. 761, 767-68, 65 S.Ct. 1515, 1519, 89 L.Ed. 1915 (1945), and for this circuit’s most recent application of the clause W.C.M. Window Co. v. Bernardi, 730 F.2d 486, 493-96 (7th Cir.1984). Although the clause as written is a grant of authority to Congress to regulate interstate (and foreign) commerce rather than an independent limitation on state power, the Supreme Court, building on a dictum by John Marshall in Gibbons v. Ogden, 22 U.S. (9 Wheat.), 1, 197-209, 6 L.Ed. 23 (1824), early on interpreted the clause as prohibiting of its own force, without need for congressional action, state action that discriminates against interstate commerce. See, e.g., Cooley v. Board of Wardens, 53 U.S. (12 How.) 299, 319, 13 L.Ed. 996 (1851). Although controversial, see, e.g., Kitch, Regulation and the American Common Market, in Regulation, Federalism, and Interstate Commerce 7, 20-22 (Tarlock ed. 1981), this interpretation of the commerce clause can be defended on the practical ground that Congress is too busy — and maybe as James Madison feared too factionalized — to police every infringement of the policy (implied by a number of separate provisions of the Constitution) that the United States be a single “common market” for goods and services. See Eule, Laying the Dormant Commerce Clause to Rest, 91 Yale L.J. 425, 431-32 (1982). But this ground fails when Congress has exercised its regulatory power. “Once Congress acts, courts are not free to review state taxes or other regulations under the dormant Commerce Clause. When Congress has struck the balance it deems appropriate, the courts are no longer needed to prevent States from burdening commerce, and it matters not that the courts would invalidate the state tax or regulation under the Commerce Clause in the absence of congressional action.” Merrion v. Jicarilla Apache Tribe, 455 U.S. 130, 154, 102 S.Ct. 894, 910, 71 L.Ed.2d 21 (1982). Congress has acted here, in the Anti-Head-Tax Act, by forbidding airport authorities to charge unreasonable rates, directly or indirectly, to interstate airlines. Therefore, when those rates are challenged, the only question is whether they are consistent with the congressional policy.
We are asked by an amicus curiae to consider the bearing of the Convention on International Civil Aviation (“the Chicago Convention,” as it is known), 61 Stat. 1180. A treaty of the United States, see British Caledonian Airways Ltd. v. Bond, 665 F.2d 1153, 1159 n. 3 (D.C.Cir.1981), the Convention has the force of a federal statute. But it does not regulate airport fees. It establishes a Council with broad powers and the Council has recommended standards that require that landing fees and other airport charges be reasonable, but these standards are issued under a provision of the Convention, Article 55, relating to the “permissive” functions of the Council, and actions taken in discharge of those functions are not intended to have the force of law.
Two facts together persuade us that the district court was correct in concluding that the ordinance was unreasonable under the applicable state and federal standards in disregarding the airport’s concession revenues. The first is monopoly. We take judicial notice of the fact that only six airports in Indiana are served by airlines other than commuter airlines and that Indianapolis International Airport is the only one in the central part of the state except *1267for Purdue University Airport in Lafayette. See U.S. Dept, of Transportation, Federal Aviation Administration, National Airport System Plan: Revised Statistics 1980-1989, at 141 (1980). The Purdue airport is tiny compared to the Indianapolis airport; the 22,940 passengers who emplaned there in 1982 were less than 2 percent of the 1,382,-391 who emplaned at Indianapolis that year. U.S. Dept, of Transportation, Federal Aviation Administration, U.S. Airport Emplanement Activity for CY 1982, at V(II-R)-29 (July 1, 1983). As no one in this lawsuit has advanced the improbable proposition that the Purdue airport is a feasible substitute for most or even many of the passengers who now use the Indianapolis airport, we may assume that except for travelers whose origination or destination is near the borders of the state (and who can therefore use the airports just across the state line, in Chicago, Louisville, and Cincinnati), most people traveling by air to and from Indiana have to use the Indianapolis airport.
Therefore, unless forbidden to do so by state or federal law, the Indianapolis Airport Authority can charge a monopoly price for the use of its airport — that is, a price in excess of the cost of operating the airport (including debt service). Of course the sky is not the limit. If the Authority charged too high a price many people would stop using the airport. For example, some of those traveling out of the state would fly out of small airports in the state and switch planes at the nearest major airport in a neighboring state; or they would drive, or take a bus or train, to their destination. But for most people these would be such grossly inferior alternatives to using the Indianapolis airport that they would rather pay even a hefty premium than switch. And it would make no economic difference whether this premium was charged to the passenger directly as he came into or left the airport or to the airline that carried him. If it was charged to the airline, the airline could turn around and raise its ticket prices to passengers to and from Indianapolis; if it was charged to the passenger the airline could absorb the charge by reducing those prices. Whether airline or passenger ultimately bears the cost of an airport fee depends on the conditions of supply and demand rather than on who is assessed the charge. All this was recognized by the Congress that passed the Anti-Head-Tax Act. See S.Rep. No. 12, 93d Cong., 1st Sess. 22 (1973), U.S.Code Cong. & Admin.News 1973, p. 1434.
It should be clear without extended discussion that a monopoly price is an unreasonable price. Locational monopoly— the type of monopoly that the Indianapolis airport enjoys — is one of the traditional levers by which a state can (if not prevented) unreasonably burden interstate commerce, see Note, Airline Deregulation and Airport Regulation, 93 Yale L.J. 319, 322 (1983), and the Anti-Head-Tax Act was passed as we said earlier in order to prevent the placing of unreasonable burdens on interstate air transportation. The Senate Report refers to the “financial windfalls” that states or cities could obtain from imposing head taxes or equivalent taxes on the airlines or their passengers. See S.Rep. No. 12, supra, at 17. If the Indianapolis airport did not have monopoly power it could not extract revenues vastly in excess of its costs, which is what it has done by the combination of user fees and concession rentals shown on this record.
It is not enough for the airlines to show that the airport has monopoly power; it must also show that this power is being used to impose unreasonable rates, directly or indirectly, on the airlines or airline passengers, and not on other entities that are neither formal nor actual parties to this case. Here the second critical fact comes into play, which is that the people who use the concessions at the Indianapolis airport are, with rare exceptions, airline passengers. Although some airports adjacent to large cities (the Milwaukee airport for example) have meeting facilities that attract nonpassengers, the Indianapolis airport does not. The parking lot is used by em-planing passengers and by people picking up deplaning passengers. The car rental *1268agencies are used by emplaning and deplaning passengers, and likewise the food stands and newsstands. This means that when the airport charges a rental fee to concessionaires it is as if it were charging a landing fee to the airlines or imposing a head tax on the passengers. If a traveler is willing to pay $140 to fly from Indianapolis to (say) New York, it should be a matter of indifference to him whether he pays $100 for the ticket, $10 in head tax, and $30 for parking; or $120 for the ticket and $20 for parking, with no head tax. What matters to him is the total cost that he must incur to make the flight, rather than the form in which the cost is distributed among the various items that he must buy.
According to the Indianapolis Airport Authority's own figures, the annual cost to it of providing rental space and other services to the parking lot and the car rental agencies is only $1 million, yet it collects $3.5 million in annual rent from these concessions. The concessionaires recoup this expense in the prices they charge their customers — the airline passengers — and it is thus the passengers who end up paying the $2.5 million in net revenues that the Authority obtains from the concessionaires. When concession rentals — paid ultimately by the passengers or (in the form of reduced ticket prices) the airlines — that are more than three times the cost that the Authority itself allocates to the concessions are added to the airline user fees that also fall on either the airlines or their passengers, the result is an exaction that is wholly disproportionate to the costs to the airport of serving the airlines and their passengers, and is therefore unreasonable under the state and federal statutes. It is unreasonable whether done to evade the statutes, or to price discriminate against the more affluent passengers (heavy users of concession services), or for other reasons. And it is unreasonable even though we may assume that all of the Authority’s income must be plowed back into airport development. See Ind.Code §§ 8-22-3-11, -25, -28, -29. No one can say how much of the overcharge extracted from the airlines’ passengers will return to them in this way. The Authority might decide to use most of it for the benefit of general-aviation users, or pour it into “gold-plating”’ improvements that would give airline users a higher quality of airport services, at a higher price, than they wanted.
The problem of the Authority’s disregarding its concession income in setting user fees has analogies in conventional public utility regulation. Many regulated firms have unregulated affiliates. A good example (at least before the recent changes in the structure and regulation of the telephone industry) is the manufacture of telephones — which has never been a regulated activity — by affiliates of regulated telephone companies. A telephone company might be tempted to evade rate regulation by having its manufacturing affiliate charge exorbitant prices (nominally to it) for telephones and by passing on the overcharge to the telephone ratepayers in the form of higher rates for telephone service. The ratepayers would end up paying monopoly prices, despite regulation. The regulatory agencies, however, were alert to the danger and successfully asserted the power to limit the profits of the manufacturing subsidiaries. See, e.g., Illinois Bell Tel. Co. v. Illinois Commerce Comm’n, 55 Ill.2d 443, 483-84, 303 N.E.2d 364, 376 (1973); 1 Kahn, The Economics of Regulation 28 n. 20 (1970); cf. Smith v. Illinois Bell Tel. Co., 282 U.S. 133, 152-53, 51 S.Ct. 65, 70, 75 L.Ed. 235 (1930). No agency has regulatory authority over the rate practices of the Indianapolis Airport Authority; instead the duty of regulation falls to the courts in the enforcement of the state and 'federal statutes forbidding unreasonable rates. But this just means that we must imagine ourselves in the role of a regulatory agency (though with more circumscribed powers, because of the limitations that Article III places on federal judicial power, see, e.g., Federal Radio Comm’n v. General Elec. Co., 281 U.S. 464, 469, 50 S.Ct. 389, 390, 74 L.Ed. 969 (1930)) that is charged with preventing airport authorities from setting exorbitant rates to airlines or *1269their passengers. By charging its concessionaires rent far in excess of the cost to the Authority of providing them with space and services (while charging the airlines landing fees equal to their full costs), knowing that the concessionaires will try to pass on their rental expenses to the passengers (their customers), the Authority is doing the same kind of thing that the telephone companies would have been doing if they had been allowed to charge their subscribers indirectly for exaggerated costs that the companies had allocated to manufacturing telephones. True, it is not exactly the same thing. The telephone subscriber has to have a telephone if he wants telephone service, and in the heyday of telephone regulation he could get the telephone only from the telephone company. The passenger who uses Indianapolis International Airport does not have to use the parking lot or a car rental agency; there are other ways of getting to and from the airport. But the existence of alternatives just limits — it does not destroy — the Authority’s power to extract by indirection the profits that its locational monopoly makes possible. Its big surplus of concession income over the costs of providing space and other facilities and services to the concessionaires show this.
The Authority cites cases such as FPC v. United Gas Pipe Line Co., 386 U.S. 237, 243, 87 S.Ct. 1003, 1007, 18 L.Ed.2d 18 (1967), for the proposition that a utility’s income from unregulated activities (here, the concessions) should not be attributed to the regulated activities (the runways and other facilities used by the airlines). We have no quarrel with the proposition in the abstract but point out that implicit in it is the assumption that the customers for a utility’s unregulated services are different people from the customers for its regulated services. If they are the same people, the utility can (if permitted) evade the regulatory ceiling on rates to its regulated customers by charging them excessive prices for the unregulated services they buy. This is what the Authority has tried to do here, since it is the airline passengers, in their capacity as patrons of the airport concessions, or the airlines, who will in the end bear the cost of the heavy rentals that the Authority has charged its concessionaires. This point was missed in Raleigh-Durham Airport Authority v. Delta Air Lines, Inc., 429 F.Supp. 1069, 1079 (D.N.C.1976), the only decision that supports the Authority’s treatment of its concession revenues.
Actually, United Gas Pipe Line Co. strengthens our analysis. The issue there was whether, in deciding what cost of service allowance for federal income tax to permit a utility to take, the Federal Power Commission had to give the utility the allowance it would be entitled to if it filed a separate tax return, or could subtract the tax savings that the utility obtained by filing a consolidated return with affiliated corporations that were not regulated. The Court held that the latter course was proper, 386 U.S. at 243-45, 87 S.Ct. at 1007-08, even though it amounted to treating a tax savings obtained in part from unregulated activities as income to the regulated firm that had to be taken into account in setting its rates. In just the same way, the district judge here forced the airport authority to take account of concession rentals, nominally obtained from unregulated activities, in setting rates to the authority’s regulated customers (the airline defendants).
An amicus curiae defends the Authority’s approach as a method of preventing airport congestion. Suppose, it argues, the Authority’s concession revenues rose to a level where they covered the airport’s entire costs. Then, if our analysis is right, the landing fees and other charges to the airlines would have to be zero; the airlines would consequently have no incentive to economize on their use of the airport; and there would be congestion and resulting passenger delays that stiff landing fees could have avoided. But leaving aside the fact that the Authority introduced no evidence that the Indianapolis airport is crowded at the existing user fees (that is, the fees set in the expired leases), we point out that airport congestion is a problem of structure, not level, of rates. No airport is *1270congested all the time; congestion is a problem only at peak periods, such as Friday evenings. And the logical — we do not say the only — way to deal with congestion at peak periods is to charge a peak-period surcharge. See, e.g., 1 Kahn, supra, at 87-103. This is the point of Professor Eckert’s criticism of airport financing, which the amicus curiae cites. See Eckert, Airports and Congestion 20-27 (1972). He does not advocate allowing airport authorities to extract revenues greater than their costs; he wants a system of variable landing fees to induce the airlines to economize on their use of airports and airways. See also Levine, Landing Fees and the Airport Congestion Problem, 12 J. Law & Econ. 79, 91, 102-03 (1969). We have said nothing to prevent the Indianapolis Airport Authority from charging higher landing fees at some times of the day or week or year than at others or shifting from a system of financing its operations out of a combination of landing fees and concession rentals to landing fees alone. And it is not true, or at least not inevitable, that if the Authority reduced or even eliminated concession rentals, the concessionaires would simply pocket the considerable sums that they now pay the airport in rentals, and airline passengers would be no better off than they are now. There is direct competition among many of the concessionaires, such as the car rental agencies; and there is competition to become a concessionaire. See Omega Satellite Products Co. v. City of Indianapolis, 694 F.2d 119, 126-27 (7th Cir.1982).
Now it is conceivable, even in a regime of total reliance on landing fees, that if the Authority raised its peak-period fees it might have to lower its fees at other times in order to avoid making its income from the airlines excessive overall. But that would simply reinforce the desired effect of the peak-period surcharge in redirecting traffic to less busy periods. Moreover, no offsetting reduction would be necessary if the peak-period surcharge did not generate additional revenue, and it might not. Its purpose would be to reduce flights at the peak period rather than to make money, and the reduction in number of flights might offset the increase in the landing fee per flight.
It is also conceivable, however, that no matter what the structure (as distinct from level) of landing fees, the airport or its airspace would become unduly congested. If high landing fees were offset by reductions in concession rentals that were passed on to the concessionaires’ customers, the airport might become so attractive a place to travel to and from that it would be thronged with passengers despite the high ticket prices, reflecting high landing fees. In that unlikely event there would be an argument (we need not decide how compelling an argument) for allowing the airport to continue to charge those fees, even though it would obtain temporary monopoly profits by doing so, in order to limit usage and enable the airport to finance a rapid expansion of its facilities. See Levine, supra, at 88. But there is no suggestion that the Indianapolis airport is anywhere near that point.
The Authority used an invalid method of calculating airline landing fees and must therefore go back to the drawing board. But unless there were a single valid method, we could not tell the Authority what fees it must charge; and no one says there is. We emphasize, too, that the powers of a federal court in regulating rates are more limited than those of an administrative agency. We can invalidate an unreasonable rate, but we cannot fix the reasonable rate; that is a legislative or administrative rather than a judicial function. Reagan v. Farmers Loan & Trust Co., 154 U.S. 362, 397-98, 14 S.Ct. 1047, 1054, 38 L.Ed. 1014 (1894). We therefore do not hold that the Authority must use the “single cash register” approach, in which the entire airport is treated as a single cost center, rather than the “multiple cash register” approach, illustrated by this case, in which the airport is divided into different revenue-producing centers each of which must pay its own way. The vice of the ordinance, at least so far as the case before us is concerned, is not that it makes the *1271airlines pay the costs allocable to them, or even that it makes the concessionaires pay so much more than the costs allocable to them (for no concessionaires are parties to this suit); it is that, by a combination of airline user fees and concession rentals, the airport authority has imposed on the airlines and their passengers a cost for the use of the airport that greatly exceeds a reasonable estimate of the costs that the airlines impose on the airport.
Although we need take no position on the single versus multiple “cash register” question we do need to consider three other objections that the airlines make to the ordinance, because they are independent of the Authority’s refusal to take account of concession revenues in setting user fees. The first is that the ordinance should have allocated the costs of firefighting services in proportion to the number of fires experienced by each class of users of the airport’s facilities over some reasonable period of time rather than mainly to the airlines. We disagree. The airport has elaborate firefighting facilities not in order to enable it to respond to a grease fire in a hamburger stand or a car fire in the parking lot but to protect the airlines and their passengers should a plane catch on fire in a crash or other accident. Airport costs that would not be avoided if a particular class of users stopped using the airport (i.e., the concessionaires) are not costs imposed on the airport by those users, and therefore are not properly allocable to them. Cf. Illinois v. ICC, 722 F.2d 1341, 1346 (7th Cir.1983). Thus the bulk of the firefighting costs are properly allocable to the airlines. On similar grounds we think the Authority introduced enough evidence — if barely enough — to support its allocation of security costs to the airlines, particularly in light of the absence of any contrary evidence introduced by the airlines. (This is a minor item; the costs of the anti-hijacking program are recovered in a separate fee to the airlines that is not contested.)
The Authority allocated $400,000 of its costs to the airport’s general-aviation users but charged no landing fee to them. It points out that the cost of collecting landing fees from general-aviation users would be very high — maybe as much as half the fee — and that in lieu of a landing fee it imposes a charge per gallon of aviation fuel consumed by general-aviation users at the airport. There is no objection to substituting a flowage fee (as it is called) for a landing fee in order to economize on the costs of collection, but the flowage fee must yield revenues commensurate with the costs allocated to general-aviation users. This flowage fee does not. It generates only $250,000 in revenues. Although the ordinance nearly doubled the landing fees and other charges to the airline users of the airport, it left unchanged the flowage fee that the Authority had been charging to general-aviation users since 1971. The difference in the Authority’s treatment of airlines and private planes — making the former pay for the full costs (and more!) that they impose on the airport, but, through inaction, allowing the latter to get away with paying little more than half of the costs they impose — has not been justified. And since flights by private planes are more likely to be intrastate than airline flights are, the effect of leaving the flowage fee unchanged has been to shift some of the costs imposed by local users of the airport to its interstate users, who are, along with many of their customers, nonresidents of Indiana. This is just the sort of discrimination Congress wanted to prevent in the Anti-Head-Tax Act.
The only other issue that requires discussion is whether the airlines are holdover tenants. If they are, then whatever new ordinance the airport authority enacts in response to our decision will not be retroactive to the date of the expiration of the leases. Several early Indiana cases discussed in Employers’ Liability Assurance Co. v. Enos Coal Corp., 457 F.2d 402, 406-07 (7th Cir.1972), hold that if a tenant does not vacate the premises at the expiration of his lease but continues to make, and the landlord continues to accept, payment of the rental fixed in the lease, *1272the tenant is automatically a holdover tenant whose obligations to the landlord are discharged by payment of that rental. But a more recent case indicates that Indiana has come into line with the prevailing view that the presumption that the tenancy continues “may be rebutted by proof of a contrary intention on the part of the landlord alone ____” Speiser v. Addis, 411 N.E.2d 439, 441 (Ind.App.1980). Of course the Authority is wrong to argue that Speiser, a decision of the Indiana Appellate Court, “overruled” the earlier cases, which include two decisions of the Indiana Supreme Court, Harry v. Harry, 127 Ind. 91, 92, 26 N.E. 562 (1891), and Lautman v. Miller, 158 Ind. 382, 63 N.E. 761 (1902). But just as an intermediate federal appellate court may properly decline to follow a U.S. Supreme Court decision when convinced that the Court would overrule the decision if it had the opportunity to do so, see, e.g., Norris v. United States, 687 F.2d 899, 902-04 (7th Cir.1982); Browder v. Gayle, 142 F.Supp. 707, 717 (M.D.Ala.) (three-judge court), aff’d per curiam, 352 U.S. 903, 77 S.Ct. 145, 1 L.Ed.2d 114 (1956); United States v. Girouard, 149 F.2d 760, 765 (1st Cir.1945) (dissenting opinion), rev’d, 328 U.S. 61, 66 S.Ct. 826, 90 L.Ed. 1084 (1946), so may intermediate state appellate courts decline to follow earlier state supreme court decisions for the same reason — especially when almost a century has passed since the earlier decisions. And if we think the intermediate state appellate court has made a correct or even, perhaps, just a defensible prediction of what the state supreme court would do if the question were put to it, then we are bound to follow its ruling in a diversity case or any other case where the issue is one of state law. See, e.g., Garris v. Schwartz, 551 F.2d 156, 158 (7th Cir.1977); 19 Wright, Miller & Cooper, Federal Practice and Procedure § 4507 at pp. 94-95 (1982).
Although the usual way of creating a holdover tenancy is by an explicit communication from the landlord that he is electing to treat the tenant as a holdover, “The conduct of the landlord ... may amount to an assertion of the right of election. This would be the case if an offer by the tenant to continue paying the usual rent was accepted without reservations.” 2 Restatement (Second) of Property (Landlord and Tenant) § 14.4, comment e (1977). But here the rent was accepted with reservations. The Authority told the airlines that it would not extend the leases unless they agreed to substantial increases in the rental rate. The airlines knew that the Authority had the legal power unilaterally to impose new terms (provided they were reasonable).'' And beginning on September 1, 1980, when the ordinance went into effect, they were fully apprised of the Authority’s position that the charges fixed in the ordinance were the rates that it would seek to collect from the airlines. By continuing to use the airport in these circumstances the airlines assumed the risk that either the ordinance would be held valid and they would have to pay the fees fixed in it or, if the ordinance was held invalid and some substitute enactment fixing reasonable fees was promulgated, they would have to pay those fees instead.
To summarize, we agree with the district judge that the ordinance is invalid in disregarding concession income and in continuing the old flowage fee for general-aviation users, but we disagree with some of his other conclusions and remand the case for further proceedings consistent with this opinion. The Authority raises some minor procedural and evidentiary issues in this appeal that are without merit and require no discussion. The parties shall bear their own costs in this court.
Affirmed in Part, Reversed in Part, and Remanded.