concurring in the judgment:
I disagree with the majority’s analysis. Plaintiff Chevron 'challenges Hawaii Act 257, a rent control law limiting the amount an oil company can charge a dealer who leases a service station from the company. *1043In a motion for summary judgment, Chevron argued that Act 257 is unconstitutional. The district court found Act 257 unconstitutional, granted Chevron’s motion, and entered final judgment for Chevron. Defendant Cayetano timely appealed. We reverse and remand.
For the reasons that follow, I concur in the reversal of the grant of summary judgment to Chevron. I disagree, however, with the majority’s rationale and with the task the majority has set for the district court on remand.
I
There are two distinct tests of constitutionality potentially applicable to Act 257. The first is a “reasonableness” test normally applied to rent control laws. The second is a “substantially advances a legitimate state interest” test normally applied to zoning and land use regulations. Relying on our earlier decision in Richardson v. City and County of Honolulu, 124 F.3d 1150 (9th Cir.1997), the majority analyzes Act 257 under the second test. The problem in this case is not determining whether the majority has properly applied that test. The problem, rather, is determining whether the test should be applied at all.
An ordinary rent control law is constitutionally indistinguishable from a price control law. Rent control involves a price charged for real property, just as price control involves a price charged for personal property. The constitutional test for ordinary rent and price control laws is the same, regardless of whether the laws are challenged under the Due Process Clause or the Takings Clause. The test has been formulated in various ways, but it essentially requires that the law be “reasonable” and “not confiscatory.” A féw examples illustrate the point.
In Pennell v. City of San Jose, 485 U.S. 1, 11, 108 S.Ct. 849, 99 L.Ed.2d 1 (1988), the Supreme Court cited a rate regulation case in upholding a municipal rent control ordinance challenged under the Due Process Clause. The Court upheld the ordinance because it was not “ ‘arbitrary, discriminatory, or demonstrably irrelevant to the policy the legislature is free to adopt...'.’ Permian Basin Area Rate Cases, 890 U.S. 747, 769-770, 88 S.Ct. 1344, 20 L.Ed.2d 312 (1968).” In FCC v. Florida Power Corp., 480 U.S. 245, 253, 107 S.Ct. 1107, 94 L.Ed.2d 282 (1987), the Court upheld a rate regulation challenged under the Takings Clause as “not confiscatory.” Citing the same rate regulation case as Pennell, it wrote that a regulation is permitted under the Constitution to “ ‘limit stringently the return recovered on investment, for investors’ interests provide only one of the variables in the constitutional calculus of reasonableness.’” Id. (citing Permian Basin Area Rate Cases, 390 U.S. at 769, 88 S.Ct. 1344). In In re Permian Basin Area Rate Cases, the 1968 case cited in both Pennell and FCC v. Florida Power, the Court upheld a rate set by the Federal Power Commission without specifying whether the challenge was brought under the Due Process or Takings Clause: “any rate selected by the Commission from the broad zone of reasonableness ... cannot properly be attacked as confiscatory.” 390 U.S. at 770, 88 S.Ct. 1344. In Bowles v. Willingham, 321 U.S. 503, 517, 64 S.Ct. 641, 88 L.Ed. 892 (1944), the Court upheld a federal rent control law after applying the same test as for a price control law: “Of course, price control, the same as other forms of regulation, may reduce the value of the property regulated. But ... that does not mean that the regulation is unconstitutional.” Finally, in Federal Power Commission v. Natural Gas Pipeline Co., 315 U.S. 575, 585, 62 S.Ct. 736, 86 L.Ed. 1037 (1942), the Court upheld a rate regulation challenged under the Due Process Clause because it was “reasonable” and “not confiscatory.”
Beginning with Agins v. City of Tibron, 447 U.S. 255, 100 S.Ct. 2138, 65 L.Ed.2d 106 (1980), the Supreme Court developed a more stringent test of constitutionality for zoning .and land use regulation cases. The zoning ordinance in Agins severely limited development of privately owned land in order to preserve open *1044space for the community. The Court upheld the constitutionality of the ordinance against a regulatory taking challenge because the ordinance “substantially advance[d] legitimate state goals.” Id. at 261, 100 S.Ct. 2138. In Nollan v. California Coastal Commission, 483 U.S. 825,107 S.Ct. 3141, 97 L.Ed.2d 677 (1987), the California Coastal Commission required owners of a beachfront house to grant a public easement across their property as a condition for receiving a permit to rebuild the house. Citing Agins, the Court stated, “[0]ur verbal formulations in the takings field have generally been quite different [from those applicable to due process]. We have required that the regulation ‘substantially advance’ the ‘legitimate state interest’ sought be achieved, not that the State could rationally have decided that the measure adopted might achieve the State’s objectives.” Id. at 834 n. 3, 107 S.Ct. 3141 (emphasis in original; citations and internal quotations omitted). The Court further required that the easement have a connection, or “essential nexus,” to the harm that would be caused by rebuilding the house. Absent such a nexus, the required conveyance of an easement to the public would be nothing more than “extortion.” Id. at 837, 107 S.Ct. 3141. Finally, in Dolan v. City of Tigard, 512 U.S. 374, 114 S.Ct. 2309, 129 L.Ed.2d 304 (1994), the city of Tigard, Oregon, required the owner of a commercial building to dedicate a portion of her property for parking and floodplain protection as a condition of receiving a permit to expand the building. The Court repeated the test from Agins and refined the Nollan “essential nexus” test. “A land use regulation does not effect a taking if it ‘substantially advancefs] legitimate state interests’ and does not ‘den[y] an owner economically viable use of his land.’” Id. at 385, 114 S.Ct. 2309. Further, a “required dedication” from the landowner is constitutionally permissible if it bears a “rough proportionality” to the “nature and extent of the impact of the proposed development” for which the permit is sought. Id. at 391, 114 S.Ct. 2309.
The Supreme Court has applied the “substantially advances a legitimate state interest” test of Agins, and its refinement in Nollan and Dolan, only in cases of severe zoning limitations on the use of land {Agins) and required dedications by landowners as a condition of receiving building permits {Nollan and Dolan). See, e.g., City of Monterey v. Del Monte Dunes at Monterey, Ltd., 526 U.S. 687, 119 S.Ct. 1624, 1635-36, 143 L.Ed.2d 882 (1999) (“[W]e have not extended the rough proportionality test of Dolan beyond the special context of exactions — land-use decisions conditioning approval of development on the dedication of property to public use.... [T]his Court has [not] provided ... a thorough explanation of the nature or applicability of the requirement that a regulation substantially advance legitimate public interests outside the context of required dedications or exactions.”). In two cases, however, the Supreme Court has hinted that, in special circumstances, a rent control law might amount to a regulatory taking and might therefore be subject to the “substantially advances a legitimate state interest” test.
In 1988, the Court considered a San Jose, California, rent control ordinance in Pennell v. City of San Jose, 485 U.S. 1, 108 S.Ct. 849, 99 L.Ed.2d 1 (1988). The Court made clear that the ordinance, considered as a whole, should be analyzed under the normal reasonableness test:
The standard for determining whether a state price-control regulation is constitutional under the Due Process Clause is well-established: “Price control is ‘unconstitutional ... if arbitrary, discriminatory, or demonstrably irrelevant to the policy the legislature is free to adopt....’” Permian Basin Area Rate Cases, 390 U.S. 747, 769-770, 88 S.Ct. 1344, 20 L.Ed.2d 312 (1968).
485 U.S. at 11, 108 S.Ct. 849 (citation omitted). Applying this test, the Court sustained the rent control ordinance, holding that it represented “a rational attempt to accommodate the conflicting interests of *1045protecting tenants from burdensome rent increases while at the same time ensuring that the landlords are guaranteed a fair return on their investment.” Id. at 13, 108 S.Ct. 849.
The plaintiffs also brought a Takings Clause challenge to a specific provision in the ordinance that appeared to require a direct wealth transfer to a particular tenant based on the poverty of that tenant. Under the ordinance, a landlord seeking a rent increase of more than 8% was required to go before a hearing officer who was authorized to consider, among other factors, individual hardship. If the proposed increase above 8% constituted “ ‘an unreasonably severe financial or economic hardship on a particular tenant,’ ” the increase could be denied. Id. at 6, 108 S.Ct. 849. The landlords contended that denial of a proposed increase on that ground would constitute a taking, but the Court refused to decide the challenge, or even to specify a test for deciding it, because the provision had never been applied and a decision would therefore be “premature.” Id. at 9, 108 S.Ct. 849.
Next, in Yee v. City of Escondido, 503 U.S. 519, 112 S.Ct. 1522, 118 L.Ed.2d 153 (1992), the Court addressed a takings challenge to an Escondido, California, mobile home rent control ordinance brought by owners of a mobile home park. As background to their challenge, the park owners pointed out that, despite their name, mobile homes are not mobile; once placed in a park, only about one mobile home in 100 is ever moved. See id. at 523, 112 S.Ct. 1522. The park owners also pointed to California’s Mobilehome Residency Law, which severely limited their ability to terminate mobile home owners’ tenancies or prevent transfer of tenancies to purchasers of the mobile homes. The park owners contended that the rent control ordinance, when viewed against this background, amounted to a physical taking of • their property. See id. at 525, 112 S.Ct. 1522. The Court rejected this contention.
The park owners further contended that the ordinance constituted a regulatory taking, but the Court refused to consider the issue because it was not included in the grant of certiorari. However, in rejecting the plaintiffs’ claim of physical taking, the Court wrote the following:
[T]he effect of the rent control ordinance, coupled with the restrictions on the park owner’s freedom to reject new tenants, is to increase significantly the value of the mobile home. This increased value normally benefits only the tenant in possession at the time the rent control is imposed.... Petitioners are correct in citing the existence of this premium as a difference between the alleged effect of the Escondido ordinance and that of an ordinary apartment rent control statute. Most apartment tenants do not sell anything to their successors (and are often prohibited from charging “key money”), so a typical rent control statute will transfer wealth from the landlord to the incumbent tenant and future tenants. By contrast, petitioners contend that the Escondido ordinance transfers wealth only to the incumbent mobile home owner. This effect might have some bearing on whether the ordinance causes a regulatory taking, as it may shed some light on whether there is a sufficient nexus between the effect of the ordinance and the objectives it is supposed to advance. See Nollan v. California Coastal Comm’n. But it has nothing to do with whether the ordinance causes a 'physical taking.
Id. at 530, 112 S.Ct. 1522 (citation omitted).
A panel of this court in Richardson relied on this passage from Yee in evaluating a Honolulu rent control ordinance. The ordinance limited long-term ground rents for residential condominiums and allowed condominium-owners/ground-lessees to sell their condominiums, and their leaseholds, without restriction. It was clear that by selling their condominiums and leaseholds, the condominium owners could, like the mobile home owners in Yee, capture a pre*1046mium representing the present value of the difference between the controlled rent for the ground lease and the open market rent that would be charged in the absence of the ordinance. Because of this one-time wealth transfer to the current condominium owners, Richardson treated the ordinance as a regulatory taking and applied the “substantially advances a legitimate state interest” test from Agins, Nollan and Dolan: “A land use regulation ... does not effect a taking if it substantially furthers a legitimate state interest and does not deny the landowner economically viable use of his land. Dolan, 512 U.S. at 385, 114 S.Ct. 2309.” 124 F.3d at 1164. Richardson then struck down the ordinance because the ability of the owner to capture the premium by selling the condominium at an open market price meant that the ordinance did not “substantially further its goal of creating affordable owner-occupied housing in Honolulu.” Id. at 1166.
The panel today greatly expands the holding in Richardson. Following the Supreme Court’s suggestion in Yee, Richardson held that the “substantially advances a legitimate state interest” test was applicable in a case where it was clear that there was a premium resulting in a one-time wealth transfer from the landlord to the tenant. Absent such a transfer, the ordinance in Richardson would have been subject to the reasonableness test normally applied to rent control ordinances. In deciding whether to apply the “substantially advances a legitimate state interest test,” the majority in this case does not first ask how clear it is that such a premium will be captured by the lessee. Rather, it treats Richardson as creating a presumptive rule that rent control laws are to be evaluated under the “substantially advances a legitimate state interest” test rather than the reasonableness test: “We established in Richardson that land use regulations, including rent control ordinances like Act 257, do not effect a taking if the regulation ‘substantially furthers a legitimate state interest.’ Richardson, 124 F.3d at 1164.” Maj. Op. at 1034.
When the majority says that “rent ordinances like Act 257” are subject to the “substantially advances a legitimate state interest” test, it appears to mean that the mere possibility of a premium capture by an incumbent tenant is enough to render a rent control ordinance “like Act 257.” It writes, “The stipulated possibility that an incumbent dealer will be able to capture the value of the decreased rent in the form of a premium separates Act 257 from an ordinary rent control situation[.]” Maj. Op. at 1035 (emphasis added). But the possibility of premium capture exists under virtually all rent control ordinances. Even under ordinances under which subleasing and assigning are prohibited, subleasing and assigning (and resulting premium capture) are nonetheless often commonplace.
We do not know in this case whether the tenants will, in fact, capture a premium. But even without knowing this, the majority has determined that the “substantially advances a legitimate state interest” test should be used to test the constitutionality of Act 257. The majority might respond that it does not matter that, at the time of determining what test to apply, the Court does not know whether a premium will be captured. That is, if upon investigation it turns out that there is no captured premium, the ordinance will pass the test. In other words, no harm, no foul. The problem with this response is that the constitutional test applied by the majority is not phrased in terms of whether a premium is captured. Rather, the test asks whether the rent control ordinance “substantially advances a legitimate state interest.” Premium capture by the tenant is only one of many ways in which an ordinance can fail that test. Thus, if it turns out that there is, in fact, no premium capture, an ordinance may nonetheless be struck down because it fails for some other reason substantially to advance a legitimate state interest.
Even if our decision in Richardson was right, I believe that the majority is wrong *1047to expand it beyond the category of cases in which the existence of the premium capture is essentially beyond dispute. I believe that in expanding the holding of Richardson, the majority’s opinion undermines or even contradicts the Supreme Court’s decisions in ordinary rent control cases such as Pennell and Bowles v. Willingham, as well as threatens its decisions in price control cases such as FCC v. Florida Power and the Permian Basin Area Rate Cases.
II
If the majority confined itself to Richardson, it would not be able to apply the “substantially advances a legitimate state interest” test to the facts of this case. Under the suggested analysis in Yee, and as I read Richardson, the prerequisite to the application of that test is that there be a clear capture of the premium resulting from the rent control ordinance. Because there is no clear showing in this case that the premium will be captured by the lessees, the prerequisite for applying the test does not exist.
The factual foundation in this case is provided by a lengthy Stipulation of Facts filed in the district court. Stipulations 26 and 27 state that Chevron dealers can sell their dealerships (and associated leaseholds), and that the selling price is not limited either by Chevron’s dealer lease or supply contracts or by Act 257. However, such a sale is subject to two conditions. First, Stipulation 80 states that Chevron could object to the sale in “good faith,” as that term is defined by Hawaii Revised Statute § 486H-1: “The petroleum distributor shall not impose on a gasoline dealer by contract, rule, or regulation, whether written or oral, any standard of conduct that is not reasonable and of material significance to the franchise relationship.” It is not clear from the materials available to us whether Chevron would be. acting in good faith within this definition if it allowed a dealer to sell a dealership and leasehold only on condition that the premium resulting from Act 257 be passed on to the new dealer in calculating the sale price. Stipulation 34 suggests that such a condition might be in good faith: “The existing dealer at the time of the enactment of Act 257 may be able to sell his leasehold at a premium that derives from the value of the dealer’s leasehold interest, given the reduced rent imposed by Act, assuming that Chevron does not object in good faith when the selling dealer seeks Chevron’s consent to the assignment.” Second, Stipulation 26 states that Chevron may require the payment of an unspecified “transfer fee” as a condition of permitting the sale of a dealership. It is not clear from the materials before us whether such a transfer fee could 'include the premium resulting from Act 257.
Further, Act 257 only limits the amount of rent Chevron can charge its lessee-dealers. Chevron derives its revenue from them not only through rent, but also through the wholesale price for gasoline. Stipulation 15 states that Chevron requires its dealers, as a condition of their lease, to purchase Chevron-branded gasoline directly from Chevron. Stipulation 17 states, “Chevron recovers its expenses and investment costs of lessee dealer stations (e.g., ground-lease rent, real property taxes, ordinary maintenance, and depreciation) in Hawaii and throughout the United States through two principal revenue streams— rental revenue and earnings on Chevron gasoline sold through the stations.” Stipulation 9 states, “Under a supply contract, the lessee-dealer markets Chevron motor fuels, which the lessee dealer buys from Chevron, at a price unilaterally determined by Chevron. Chevron does not enter into a dealer lease unless the dealer simultaneously executes a supply contract with Chevron” (emphasis added).
It is thus entirely within Chevron’s power to prevent its lessee-dealers from capturing any premium resulting from Act 257. Chevron may have that power pursuant to its ability to object in good faith to a sale or to impose a transfer fee. Chevron certainly has that power pursuant to its ability unilaterally to increase the whole*1048sale price of the gasoline to its dealers. Indeed, the district court specifically discussed Chevron’s ability to charge more for its gasoline and thereby to capture the premium: “Defendant’s expert fails to explain why the oil company would not increase the wholesale price to simply offset the decrease in rent.... Neither Defendants nor Defendants’ expert have offered any reason why this is not a feasible, and even likely, result.”
I recognize that the district court also stated that “the Act ... allows incumbent dealers to capture the value of the decreased rent in the form of a premium,” but the court made the statement to show why the Act was not likely to achieve its purpose of lowering retail gasoline prices rather than to justify the application of the “substantially advances a legitimate state interest” test. The question under Richardson is not whether the terms of the Act themselves allow — that is, do not prohibit — capture of the premium; rather, the question is whether the Act creates a situation where we know the premium will, in fact, be captured. As the district court noted, the “feasible, even likely, result” is that Chevron will take that premium for itself in the form of higher wholesale prices charged to its dealers.
Ill
I fear that under the majority opinion virtually all rent control laws in the Ninth Circuit are now subject to the “substantially advances a legitimate state interest” test, and that this test may invalidate many of these laws. I will not undertake an extended analysis of the economic and social effects of rent control laws. Suffice it to say that the virtually unanimous opinion of economists is that, except in unusual and short-lived circumstances, they often do not achieve their stated purposes. They result in the creation of large and unwieldy bureaucracies. They do not subsidize the truly needy — the homeless and those in public housing; rather, they subsidize those already able to pay for their own housing, including many who can easily pay an open market price. They interfere with the normal play of free market forces, thereby creating incentives that result in reduced supplies of housing, reduced maintenance and repairs on existing housing, increased housing code violations, and increased transportation inefficiencies when tenants change schools or jobs but remain in rent-controlled housing.
The question before the judiciary is not the advisability of rent control laws but rather their constitutionality. Ever since its retreat from economic substantive due process at the end of the 1930s, the Supreme Court has essentially left it to the other branches of government to decide, in their political wisdom, whether to adopt rent and price controls. The Supreme Court’s hints in Pennell and Yee may signal a willingness to rethink this long-ago retreat, but at this point the Court has not yet done so.
I am not sure whether, in Richardson, we properly interpreted the Court’s hints in Yee in concluding that the “substantially advances a legitimate state interest” test used in zoning and land use regulation cases should have been applied to the rent control ordinance in that case. I am inclined to think that we did not. I am sure, however, that the majority in this case extends Richardson beyond current law.
IV
It is not clear that Hawaii’s Act 257 will result in the capture of a premium representing a one-time wealth transfer to dealers currently leasing stations from Chevron. Thus, even assuming that Richardson is good law, Act 257 should not be analyzed under the “substantially advances a legitimate state interest” test. Rather, Act 257 should be analyzed under the reasonableness test applicable to ordinary rent and price control laws. While I agree with the majority that the summary judgment granted to Chevron should be reversed, I disagree with the majority about the district court’s task on remand. In my view, the district court should apply the reasonableness test applied — until *1049today — to ordinary rent and price control laws.