Volume 1 of 2

                 FOR PUBLICATION
 UNITED STATES COURT OF APPEALS
      FOR THE NINTH CIRCUIT

DANIEL GUGGENHEIM; SUSAN              
GUGGENHEIM; MAUREEN H. PIERCE,
                                            No. 06-56306
             Plaintiffs-Appellants,
                v.                           D.C. No.
                                          CV-02-02478-FMC
CITY OF GOLETA, a municipal
                                             OPINION
corporation,
              Defendant-Appellee.
                                      
       Appeal from the United States District Court
           for the Central District of California
     Florence Marie Cooper, District Judge, Presiding

                   Argued and Submitted
            April 7, 2008—Pasadena, California

                 Filed September 28, 2009

   Before: Alfred T. Goodwin, Andrew J. Kleinfeld, and
               Jay S. Bybee, Circuit Judges.

                Opinion by Judge Bybee;
                Dissent by Judge Kleinfeld




                           13803
13808           GUGGENHEIM v. CITY OF GOLETA




                        COUNSEL

Mark D. Alpert, Robert S. Coldren (argued), and C. William
Dahlin, Santa Ana, California, for the plaintiffs-appellants.

Julie Hayward Biggs and Amy E. Morgan (argued), Los
Angeles, California, for defendant-appellee.


                         OPINION

BYBEE, Circuit Judge:

   Daniel Guggenheim and others bring a facial challenge to
the City of Goleta’s mobile home rent control ordinance.
                 GUGGENHEIM v. CITY OF GOLETA              13809
Guggenheim argues that the ordinance, which effects a trans-
fer of nearly 90 percent of the property value from mobile
home park owners to mobile home tenants, constitutes a regu-
latory taking under Penn Central Transportation Co. v. New
York City, 

(1978). We have fielded such chal-
lenges before, but have never reached the merits of the tak-
ings claim. See, e.g., Equity Lifestyle Props., Inc. v. County of
San Luis Obispo (“Equity Lifestyle“), 

, 1190
n.11 (9th Cir. 2008); Carson Harbor Vill. Ltd., v. City of Car-
son, 

, 475-77 (9th Cir. 1994), overruled on other
grounds by WMX Techs., Inc. v. Miller, 

, 1136
(9th Cir. 1997) (en banc); Levald, Inc. v. City of Palm Desert,

, 686-89 (9th Cir. 1993); Sierra Lake Reserve v.
City of Rocklin, 

, 955 (9th Cir. 1991), vacated,

(1992).

   To determine whether a taking has occurred we must
decide several issues. We must first determine whether the
mobile home park owners have standing to bring this case.
Additionally, we must consider whether this case is ripe under
Williamson County Regional Planning Commission v. Hamil-
ton Bank of Johnson City, 

(1985). If so, then we
must determine whether the city ordinance constitutes a regu-
latory taking under Penn Central. We also address challenges
to the ordinance under the Due Process and Equal Protection
Clauses.

   The district court did not address either the standing or
ripeness questions due to the unusual procedural history of the
case, but implicitly found the case was properly brought. The
district court found that no taking had occurred. For the rea-
sons explained below, we agree with the district court that this
case is properly brought and ripe for decision, but we disagree
with the district court on the merits of the takings claim.
Because we find that a taking has occurred, we reverse and
remand to the district court to determine what compensation
is due. We affirm the district court’s judgment on the due pro-
cess and equal protection claims.
13810           GUGGENHEIM v. CITY OF GOLETA
                               I

                               A

   The Takings Clause of the Fifth Amendment, made appli-
cable to the states through the Fourteenth Amendment, see
Chicago, B. & Q.R. Co. v. Chicago, 

, 236
(1897), provides that “private property [shall not] be taken for
public use, without just compensation.” The Takings Clause
“does not prohibit the taking of private property, but instead
places a condition on the exercise of that power.” First
English Evangelical Lutheran Church of Glendale v. County
of Los Angeles, 

, 314 (1987). The Takings
Clause was drafted so as “not to limit the governmental inter-
ference with property rights per se, but rather to secure com-
pensation in the event of otherwise proper interference
amounting to a taking.” 

The Takings Clause
“ ‘bar[s] Government from forcing some people alone to bear
public burdens which, in all fairness and justice, should be
borne by the public as a whole.’ ” Lingle v. Chevron U.S.A.
Inc., 

, 537 (2005) (quoting Armstrong v. United
States, 

, 49 (1960)).

   To determine whether a mobile-home rent control ordi-
nance constitutes a taking under the Constitution, we must
first understand some unique characteristics of mobile homes.
“The fact that these homes can be moved does not mean that
they do move.” JOHN STEINBECK, TRAVELS WITH CHARLEY: IN
SEARCH OF AMERICA 96 (Penguin Books 1986) (1962). As
described by the Supreme Court:

    The term “mobile home” is somewhat misleading.
    Mobile homes are largely immobile as a practical
    matter, because the cost of moving one is often a sig-
    nificant fraction of the value of the mobile home
    itself. They are generally placed permanently in
    parks; once in place, only about 1 in every 100
    mobile homes is ever moved . . . . A mobile home
                GUGGENHEIM v. CITY OF GOLETA             13811
    owner typically rents a plot of land, called a “pad,”
    from the owner of a mobile home park. The park
    owner provides private roads within the park, com-
    mon facilities such as washing machines or a swim-
    ming pool, and often utilities. The mobile home
    owner often invests in site-specific improvements
    such as a driveway, steps, walkways, porches, or
    landscaping. When the mobile home owner wishes
    to move, the mobile home is usually sold in place,
    and the purchaser continues to rent the pad on which
    the mobile home is located.

Yee v. City of Escondido, 

, 523 (1992) (citation
omitted).

   The County of Santa Barbara, California (the “County”),
first enacted its Rent Control Ordinance (the “RCO”) in 1979,
and amended it in 1987. In 2002, the City of Goleta incorpo-
rated within the County. As required by California law, the
new City of Goleta immediately adopted by reference the
County’s code in its entirety, including the RCO, as its provi-
sional new code. See CAL. GOV’T CODE § 57376 (2008); City
of Goleta Ordinance No. 02-01. About two months later, the
City re-adopted by reference most provisions of the County
code, including the RCO, as permanent city ordinances. City
of Goleta Ordinance No. 02-17.

   The statement of “Purpose” in the RCO has remained
unchanged since the RCO was first passed by the County in
1979. The purpose was to prevent mobile home park owners
from charging exorbitant rents to exploit local housing short-
ages and the fact that mobile home owners could not easily
move their homes:

    A growing shortage of housing units resulting in a
    critically low vacancy rate and rapidly rising and
    exorbitant rents exploiting this shortage constitutes
    serious housing problems affecting a substantial por-
13812              GUGGENHEIM v. CITY OF GOLETA
      tion of those Santa Barbara County residents who
      reside in rental housing . . . . Especially acute is the
      problem of low vacancy rates and rapidly rising and
      exorbitant rents in mobile home parks in the County
      of Santa Barbara. Because of such factors and the
      high cost of moving mobilehomes, . . . the board of
      supervisors finds and declares it necessary to protect
      the owners and occupiers of mobilehomes from
      unreasonable rents while at the same time recogniz-
      ing the need for mobile home park owners to receive
      a fair return on their investment and rent increases
      sufficient to cover their increased costs.

RCO § 11A-1.1

   The RCO limits any increases in mobile home rents on an
annual basis to 75 percent of the increase in the local Con-
sumer Price Index (“CPI”). RCO §§ 11A-5(a)(2), 11A-
5(a)(3), 11A-5(g). This increase is referred to as the “auto-
matic increase.” Mobile home park owners may also increase
the rent by an additional amount to pass through increased
operating costs, capital expenses, and capital improvements.
This increase is referred to as the “discretionary increase.”
RCO § 11A-5(f)(1); 11A-6. The RCO sets out an arbitration
process by which park owners must work with the mobile
home owners and an arbitrator to determine the total amount
of the permissible rent increase for each year. RCO §§ 11A-4,
11A-5. The arbitrator must follow a complicated formula to
  1
    Because the City of Goleta adopted the RCO “by reference,” the
authoritative source of the RCO is found in the County code. See City of
Goleta Ordinance No. 02-17 (adopting most provisions of the County code
“by reference” and stating that “[w]henever ‘County’ or ‘County of Santa
Barbara’ is used in the County Code . . . it shall mean the City of
Goleta.”). A copy of Santa Barbara’s current version of the RCO may be
found at http://www.bpcnet.com/codes/stbarb/, by clicking on the link
titled “Chapter 11 Mobilehomes.” The citations in this opinion refer to the
RCO as amended in 1987.
                GUGGENHEIM v. CITY OF GOLETA             13813
determine the amount of any increase in excess of the auto-
matic increase:

    (1) First, grant one-half of the automatic increase to
    management as a just and reasonable return on
    investment. The arbitrator shall have no discretion to
    award additional amounts as a just and reasonable
    return on investment;

    (2) Next, grant one-half of the automatic increase to
    management to cover increased operating costs. The
    arbitrator shall have no discretion to award less than
    this amount for operating costs.

    (3) Next, add an amount to cover operating costs, if
    any, in excess of one-half of the automatic increase.
    The arbitrator shall have discretion to add such
    amounts as are justified by the evidence and other-
    wise permitted by this chapter.

    (4) Next, add an amount to cover new capital
    expenses. Where one-half of the automatic increase
    is more than the actual increase in operating costs for
    the year then ending, the arbitrator shall offset the
    difference against any increases for new capital
    expenses.

    (5) Next, add an amount to cover old capital
    expenses. Where one-half of the automatic increase
    is more than the actual increase in operating costs for
    the year then ending, the arbitrator shall offset the
    difference against any increase for old capital
    expenses unless such difference has already been
    used to offset an increase for a new capital expense
    or another old capital expense . . . .

    (6) Finally, add an amount to cover increased costs
    for capital improvements, if any. The arbitrator shall
13814            GUGGENHEIM v. CITY OF GOLETA
    have discretion to add such amount as is justified by
    the evidence and otherwise permitted by this ordi-
    nance.

RCO § 11A-5(I). The RCO also contains a vacancy control
provision, which limits the permissible rent increase to 10
percent when a unit is sold. RCO § 11A-14. In sum, the RCO
mandates that a “just and reasonable return” for the park own-
ers must always be less than or equal to exactly one half of
75 percent of the annual increase of the CPI. The RCO per-
mits park owners to go to arbitration to pass through addi-
tional costs, but such costs must be re-captured without any
return on investment. In the event a tenant sells his or her unit,
the park owners are entitled to a one-time rent increase of 10
percent; subsequent increases are capped by the regular for-
mula.

                                B

                                1

   Appellants Daniel Guggenheim, Susan Guggenheim, and
Maureen H. Pierce (collectively, the “Park Owners”) pur-
chased the Ranch Mobile Estates mobile home park (“the
Park”) in 1997, at which time the Park was located in an unin-
corporated part of the County. At the time of the purchase,
therefore, the Park was subject to the County’s RCO as
amended in 1987. When the City incorporated in 2002, the
Park fell within the new city’s jurisdiction. Because the City
adopted the RCO by reference, the Park continued to be sub-
ject to the RCO after the City incorporated.

   A month after the City incorporated, the Park Owners
brought suit in federal court, alleging only facial challenges
to the RCO. The Park Owners claimed, inter alia, violations
of the Takings Clause, the Due Process Clause, and the Equal
Protection Clause. The Park Owners also raised complex state
law claims, claiming the City failed to follow proper proce-
                    GUGGENHEIM v. CITY OF GOLETA                     13815
dures required by the California Government Code when it
enacted the RCO. Apparently, the Park Owners initiated the
lawsuit in 2002, even though they purchased the Park in 1997,
because they claimed the City adopted the RCO without any
“hearings or studies or investigations as to whether the Coun-
ty’s Ordinance was needed or appropriate for the City.” The
Park Owners’ complaint represented that they “had attempted
to meet with the City officials-elect to discuss the City’s
potential adoption of” the County’s RCO, and had “applied to
the City for relief from the potential vacancy control restric-
tion in the County Ordinance[,] but it was nevertheless
adopted without any change by Defendant City.” The Park
Owners complained that when it adopted the RCO, “[t]he City
failed to review the County Code or make any findings on
whether there was a purpose or need” for the RCO in the cur-
rent real estate market.

   The district court stayed the viable federal claims under the
Pullman doctrine, to permit the resolution of certain complex
state law claims that might “moot or narrow the constitutional
questions.” San Remo Hotel v. City and County of San Fran-
cisco, 

, 1104 (9th Cir. 1998). The parties settled
their state law claims after litigating in Santa Barbara Supe-
rior Court, and then returned to federal court for a second time.2

                                     2

  Back in federal court, the Park Owners moved for partial
summary judgment. The district court reviewed the undis-
puted facts and the affidavits and documents proffered by the
  2
    Of particular relevance to this appeal, the parties stipulated that there
was a gap in time when no rent control ordinance was in effect over the
Park. This stipulated fact was necessary to support the timeliness of the
Park Owners’ facial challenges to the RCO. The statute of limitations for
a facial takings claim begins to run with the passage of the challenged law.
See Equity 

. The supposed “gap in time” clari-
fied that the City’s RCO, for purposes of this litigation, was enacted in
2002. Thus, the Park Owners’ suit, initiated in 2002, was timely.
13816           GUGGENHEIM v. CITY OF GOLETA
parties. The court found that during the time the Park Owners
owned the Park, housing costs in the City increased approxi-
mately 225 percent. Because of the RCO, the rents charged by
the Park Owners did not keep pace with this increase. The
below-market rents resulted in the ability of mobile home
owners to sell their homes at a significant premium (the trans-
fer premium). The district court found, based on a report pro-
vided by the Park Owners, that the transfer premium
amounted to, on average, 88 percent of the sale price. “In
other words,” the district court found, “an average mobile
home worth $12,000 would sell for approximately $100,000.”
The district court found that “the uncontroverted facts . . .
establish the existence of a premium,” and that even “[t]he
City has acknowledged the existence of such a premium.”

   The district court granted summary judgment on the takings
claim in favor of the Park Owners on October 29, 2004. At
the time the district court made its determination, the law in
the Ninth Circuit was that a government regulation effected a
taking if such regulation did not “substantially advance” legit-
imate state interests. See, e.g., Agins v. City of Tiburon, 

, 260 (1980); Richardson v. City and County of
Honolulu, 

, 1165-66 (9th Cir. 1997) (holding
that a condominium rent control ordinance that permits
incumbent condominium owners to capitalize the net present
value of reduced land rent will not substantially further its
goal of creating affordable owner-occupied housing and thus
constitutes a taking). The district court found it undisputed
that the RCO effected a one-time wealth transfer from the
Park Owners to the incumbent tenants, and that the RCO
failed to substantially advance its stated purpose of providing
affordable housing. The court found, therefore, that the RCO
was an unconstitutional regulatory taking and the Park Own-
ers were entitled to just compensation. The City timely
appealed.

  On May 23, 2005, while the case was on appeal, the
Supreme Court decided Lingle v. Chevron U.S.A. Inc., 544
                    GUGGENHEIM v. CITY OF GOLETA                       

(2005). Lingle repudiated the “substantially
advances” theory upon which the Park Owners had prevailed.3
In light of this development, the parties stipulated to vacate
the district court’s judgment and return for what would now
be their fourth round of litigation before a trial court.

                                      3

   After some renewed pre-trial litigation, the district court
issued a series of summary judgment rulings in which it found
in favor of the City on each of the Park Owners’ remaining
constitutional claims. On April 5, 2006, the district court
denied the Park Owners’ motion for partial summary judg-
ment, finding that the Park Owners were not entitled to judg-
ment as a matter of law as to whether the RCO constituted a
regulatory taking under Penn Central Transportation Co. v.
New York City, 

(1978). The court reviewed both
parties’ expert reports and found that the evidence as to the
economic impact of the regulation was “mixed”:

      Although [the Park Owners] have enjoyed a rate of
      return comparable to other real estate investments,
      [the Park Owners’] evidence tends to suggest that
      they would have earned more—perhaps much more
      —in the absence of the RCO.

The district court also denied the Park Owners’ motion for
summary judgment on their substantive due process and equal
protection claims. The parties continued to prepare for trial—
designating experts, agreeing to witness and exhibit lists, and
filing motions in limine.
  3
   The Court found that the “substantially advances” theory “prescribes
an inquiry in the nature of a due process, not a takings, test, and that it has
no proper place in our takings jurisprudence.” 

. See
generally Crown Point Dev., Inc. v. City of Sun Valley, 

,
854-56 (9th Cir. 2007) (discussing Lingle’s reasoning and its impact on
Takings Clause jurisprudence).
13818            GUGGENHEIM v. CITY OF GOLETA
   On July 27, 2006, the district court sua sponte issued an
Order to Show Cause why the court should not, on its own
motion, enter summary judgment in favor of the City. On
September 6, 2006, after reviewing the parties’ responses, the
district court entered summary judgment in favor of the City
on all of the Park Owners’ remaining causes of action. The
court stated:

    Because this is a facial challenge to the ordinance in
    question, the evidence [the Park Owners] seek to
    present at trial vis[-]a[-]vis their Fifth Amendment
    [T]akings [C]lause claim is irrelevant. To facially
    attack the ordinance as an uncompensated “taking,”
    [the Park Owners] must demonstrate that the mere
    enactment of the ordinance constitutes a taking.

The court then complained that the Park Owners had “imper-
missibly attempted to convert this action, de facto, into an as-
applied challenge.” The district court did not, however, iden-
tify which evidence it found “irrelevant” or “impermissible”
in a facial takings claim. The district court also did not make
explicit whether it incorporated its April 5 analysis of the Park
Owners’ Penn Central claim into its final judgment or
whether it entered final judgment solely on the ground the
Park Owners were barred from presenting evidence in a facial
challenge. The Park Owners appealed in a timely manner.

                               II

   This case has already been litigated through three full
rounds at the trial level, including one in state court and two
in federal court, producing one victory for the Park Owners,
one for the City, and one tie (the settlement). Accordingly, it
may come as a surprise that before we reach the merits of the
Park Owners’ appeal, we must consider whether the plaintiffs
have standing to bring this case and whether this case is ripe
for decision.
                   GUGGENHEIM v. CITY OF GOLETA                   13819
                                   A

   Under Article III, our power to adjudicate is limited to
“cases” and “controversies.” U.S. CONST. art. III, § 2, cl. 1.
Accordingly, we are not authorized to decide a dispute
“merely because a party requests a court of the United States
to declare its legal rights, and has couched that request for
forms of relief historically associated with courts of law in
terms that have a familiar ring to those trained in the legal
process.” Valley Forge Christian Coll. v. Ams. United for
Separation of Church and State, Inc., 

, 471
(1982). Rather, we must first determine whether a litigant has
“standing” to bring suit in the federal forum for his alleged
injury.4 See 

   The Supreme Court has defined standing generally as “the
question of . . . whether the litigant is entitled to have the
court decide the merits of the dispute or of particular issues.”
Warth v. Seldin, 

, 498 (1975). We have recog-
nized that a plaintiff must at minimum present a suit with
“three elements” in order to satisfy us that this question can
be answered affirmatively. Colwell v. Dep’t of Health &
Human Servs., 

, 1121-22 (9th Cir. 2009) (quot-
ing Lujan v. Defenders of Wildlife, 

, 560 (1992)).
A plaintiff must first “have suffered an injury in fact—an
invasion of a legally protected interest which is (a) concrete
and particularized and (b) actual or imminent, not conjectural
or hypothetical.” 

(internal quotations
omitted). “Second, there must be a causal connection between
the injury and the conduct complained of—the injury has to
be fairly traceable to the challenged action of the defendant,
and not the result of the independent action of some third
party not before the court.” 

marks and
alterations omitted). Finally, “it must be ‘likely,’ as opposed
  4
   Although neither party addressed standing, “[b]ecause standing is a
necessary element of federal jurisdiction, we raise the issue of standing
sua sponte.” Carson Harbor 

.
13820            GUGGENHEIM v. CITY OF GOLETA
to merely ‘speculative,’ that the injury will be ‘redressed by
a favorable decision.’ ” 

   There is no question that the latter two elements of the
standing inquiry are satisfied by the Park Owners. The Park
Owners submitted a comprehensive analysis of the effects of
the RCO which demonstrated that the RCO reduced the rents
the Park Owners could collect by approximately $10,000 per
year. See infra n.14. The link between the Park Owners’
injury and the RCO is thus not “tenuous” but “fairly trace-
able” to the City’s action. See Tyler v. Cuomo, 

,
1132-33 (9th Cir. 2000). If we were to determine that the
RCO effected a taking, the Park Owners are due compensa-
tion for their loss—thus, it is not “merely speculative . . . that
the injury will be redressed by a favorable decision.” 

   [1] Nevertheless, we must still determine whether the Park
Owners have an “actual injury”—that they have “alleged such
a personal stake in the outcome of the controversy as to assure
that concrete adverseness which sharpens the presentation of
issues upon which the court so largely depends for illumina-
tion of difficult constitutional questions.” Baker v. Carr, 

, 204 (1962). This “actual injury” requirement “tends
to assure that the legal questions presented to the court will
be resolved, not in the rarified atmosphere of a debating soci-
ety, but in a concrete factual context conducive to a realistic
appreciation of the consequences of judicial action.” Valley

.

   Although in this case there is every indication that the Park
Owners (and the City, who never raised the question of stand-
ing) believed they had a personal stake in the outcome of the
controversy—indeed, both parties litigated the merits of the
claim several times over—we have previously denied stand-
ing to similar plaintiffs bringing facial takings challenges
against rent control ordinances. See Equity 

; Carson Harbor 

-76. In Carson
                 GUGGENHEIM v. CITY OF GOLETA               13821
Harbor Village, we held that a mobile home park owner who
had purchased the regulated property after the allegedly
unconstitutional ordinance was passed did not have standing
to state a facial takings 

. We reasoned
that the park owner’s claims “necessarily rest on the premise
that an interest in property was taken from all mobile home
property owners upon the statute’s enactment.” 

because “[i]n a facial taking, the harm is singular and
discrete, occurring only at the time the statute is enacted . . . .
[b]ecause [the plaintiff] did not own the property when the
statutes were enacted and when the alleged facial takings
occurred, it has incurred no injury entitling it to assert a facial
claim.” 

Equity Lifestyle, we dismissed a
mobile home park owner’s facial takings claim because “the
injury is treated as having occurred to the previous landown-
er” who occupied the property at the moment the allegedly
offending statute was 

.

   We have also noted, without deciding the issue, that the
Supreme Court’s opinion in Palazzolo v. Rhode Island, 

(2001), calls into question the principle that a subse-
quent property owner does not have standing to assert a facial
challenge to a statutory enactment thought to effectuate a tak-
ing. See Equity 

n.11, 1193 n.15.
In Palazzolo, the Court rejected the notion that only the land-
owner at the time of the statute’s enactment could assert a
valid takings claim under Penn Central. 

(“[A takings claim] is not barred by the mere fact that title
was acquired after the effective date of the state-imposed
restriction.”). The Court remarked that “[a] law does not
become a background principal for subsequent owners by
enactment itself.” Id; see Equity 

n.11. These statements, as we have previously noted, cast
doubt on Carson Harbor Village’s rationale for denying
standing to subsequent purchasers—they indicate that a sub-
sequent purchaser may have a stake in a facial suit against the
regulation. See 

v. CITY OF GOLETA
   [2] In this case, however, even if the rule of Carson Harbor
Village survives Palazzolo, the Park Owners satisfy Article
III’s case or controversy requirements. Although the Park
Owners purchased the burdened property in 1997, eighteen
years after the County first passed the RCO and ten years
after it was amended, the City adopted the RCO in 2002, after
the Park Owners were in possession of the Park. Additionally,
the parties stipulated that there was some time period between
the City’s incorporation and the City’s adoption of the RCO
in which no rent control ordinance was in effect. 

n.2. Thus, assuming that Carson Harbor Village is still good
law, even though the Park Owners might not have standing to
challenge the County’s use of the RCO, they are precisely the
sort of plaintiffs Carson Harbor Village envisioned bringing
a facial challenge to the City’s RCO. See Carson Harbor Vil-

(“[F]acial [takings] claims necessarily
rest on the premise that an interest in property was taken from
all mobile home property owners upon the statute’s enact-
ment.”). We therefore find that the Park Owners have stand-
ing to bring their takings claim.

                               B

   [3] “[A] takings claim must [also] . . . comply with timeli-
ness requirements. It must be filed neither too early (unripe)
nor too late (barred by a statute of limitations).” Equity Life-

. In Williamson County Regional Plan-
ning Commission v. Hamilton Bank of Johnson City, 

, 185 (1985), the Supreme Court held that a takings claim
is not ripe until the property owner has attempted to obtain
just compensation for the loss of his or her property through
the procedures provided by the state for obtaining such com-
pensation and been denied. 

Williamson also set
forth an additional hurdle, applicable only to as-applied chal-
lenges: the property owner must have received a “final deci-
sion” from the appropriate regulatory entity as to how the
challenged law will be applied to the property at issue. 

The latter requirement is not applicable here because
                GUGGENHEIM v. CITY OF GOLETA              13823
the Park Owners have raised only a facial challenge. “Facial
challenges are exempt from the [“final decision”] prong of the
Williamson ripeness analysis because a facial challenge by its
nature does not involve a decision applying the statute or reg-
ulation.” Hacienda Valley Mobile Estates v. City of Morgan
Hill, 

, 655 (9th Cir. 2003) (citation omitted).

   The district court found that this case was ripe, although on
a slightly different theory. When the Park Owners filed suit
in federal district court, they had approached the City of
Goleta to ask for relief from the RCO, but had not brought an
inverse condemnation suit in a California court. Thus, the
Park Owners had failed to satisfy Williamson’s first prong,
that the property owners exhaust state remedies. The district
court found that the Park Owners’ facial challenges were ripe
nevertheless because of a narrow exception to the Williamson
requirement. At the time the Park Owners brought this suit, a
claim that a law constituted a taking because it did not “sub-
stantially advance” the purpose of that law was exempt from
the Williamson requirement. See 

. The
Court had created the exception for challenges on the “sub-
stantially advances” theory because “this allegation does not
depend on the extent to which petitioners are deprived of the
economic use of their particular pieces of property or the
extent to which these particular petitioners are compensated.”

(citing Keystone Bituminous Coal Ass’n v. DeBene-
dictis, 

, 495 (1987)). Thus, the Park Owners
were permitted to litigate their claims through federal court;
they eventually prevailed on the “substantially advances” the-
ory.

   When the Supreme Court repudiated the “substantially
advances” theory in Lingle, presumably it closed this theory’s
loophole in the Williamson requirements. See San Remo
Hotel, L.P. v. City and County of San Francisco, 

, 345-46 & n.25 (2005) (stating that Williamson did not
reach “substantially advances” claims, but noting that after
Lingle, such claims were foreclosed). Therefore, once Lingle
13824            GUGGENHEIM v. CITY OF GOLETA
was decided and the parties stipulated to vacate the first judg-
ment of the district court and return to that court to litigate the
Park Owners’ remaining claims, the ripeness of these claims
was unclear.

   After returning to district court, the City failed to raise the
issue of ripeness to the district court’s attention, and instead
proceeded to defend its RCO on the merits. The district court,
too, declined to raise the issue of ripeness and proceeded to
grant summary judgment in favor of the City on the merits.
In the parties’ filings on appeal to this court, neither party
raised the issue of ripeness. Unsure of whether this case was
ripe, and unsure of whether we had a duty to raise the issue
ourselves, we asked the parties to discuss ripeness at oral
argument. In addition to presenting the legal arguments we
discuss below, the parties both represented that the Park Own-
ers had not brought an inverse condemnation action in a Cali-
fornia court. See generally Kavanau v. Santa Monica Rent
Control Bd., 

(Cal. 1997).

                                1

   In order to determine whether the Park Owners’s claims are
ripe under Williamson, and, if so, whether they have satisfied
the Williamson requirements, we must look closely at Wil-
liamson and its progeny.

   As the Park Owners contended at oral argument, William-
son has come under scrutiny since it was decided. Counsel for
the Park Owners accused Williamson of having effectively
“closed the federal courthouse doors” to litigants seeking to
vindicate an important right embedded in the Fifth Amend-
ment of the United States Constitution. He was not the first
to level this accusation. In fact, the Supreme Court has
already acknowledged that the practical effect of Williamson
is that plaintiffs alleging violations of the Takings Clause will
almost never have the opportunity to litigate their federal
claims in federal court. See San Remo Hotel, 545 U.S. at
                    GUGGENHEIM v. CITY OF GOLETA                     13825
337-41, 344-48. Williamson requires plaintiffs to go first to
state court, where they are likely to generate a ruling on the
merits of their takings claim from the state court that in turn
will have preclusive effect should they opt to return to federal
court. 

Rehnquist, joined by three members of the
Court, wrote specially in San Remo Hotel to explain why he
believed Williamson may have been wrongly decided. See 

352 (Rehnquist, C.J., concurring in the judgment) (“I
joined the opinion of the Court in Williamson County. But
further reflection and experience lead me to think that the jus-
tifications for its state-litigation requirement are suspect,
while its impact on takings plaintiffs is dramatic.”). The con-
curring Justices stated:

      Williamson County all but guarantees that claimants
      will be unable to utilize the federal courts to enforce
      the Fifth Amendment’s just compensation guarantee.
      The basic principle that state courts are competent to
      enforce federal rights and to adjudicate federal tak-
      ings claims is sound, and would apply to any number
      of federal claims. But that principle does not explain
      why federal takings claims in particular should be
      singled out to be confined to state court, in the
      absence of any asserted justification or congressional
      directive.

(citations omitted). Nevertheless, unless and until
the Court decides to reconsider this issue, Williamson is the
law by which we are bound.5
  5
   “We are free to muse, however,” Clement v. City of Glendale, 

, 1093 n.4 (9th Cir. 2008), as to an additional reason why Williamson
may be incorrect. See San Remo 

(Rehnquist, C.J.,
concurring in the judgment) (suggesting that it might be appropriate to
revisit this issue if the “court below has addressed the correctness of Wil-
liamson”). Williamson reasoned that “[t]he Fifth Amendment does not
13826               GUGGENHEIM v. CITY OF GOLETA
                                     2

   Although the Court has so far declined to reconsider Wil-
liamson, it has with some frequency continued to clarify and
modify the doctrine. These modifications provide the frame-
work by which we must determine the application of William-
son to the unusual case before us.

   Most importantly, the Court has explicitly held that the Wil-
liamson requirements are merely prudential requirements. In
Lucas v. South Carolina Coastal Council, the Court stated
that the Williamson requirements were prudential: “Lucas has
properly alleged Article III injury in fact in this case,” and the
fact that he had not satisfied Williamson “goes only to the
prudential ‘ripeness’ of Lucas’s challenge, and for the reasons
discussed we do not think it prudent to apply that prudential
requirement here.” 

, 1012-13 (1992). Similarly,
in Suitum v. Tahoe Regional Planning Agency, the Court dis-
tinguished constitutional ripeness under Article III and pru-
dential ripeness, and stated that Williamson ripeness is
grounded exclusively in prudential considerations. 

, 733-34 & n.7 (1997) (stating that it was undisputed that
the case “presents a genuine ‘case or controversy’ sufficient

proscribe the taking of property; it proscribes taking without just compen-
sation,” and therefore, until the property owner has actually sought and
been denied just compensation in the state court, the Fifth Amendment has
not been violated. 

. With this analysis, the
Court read an exhaustion requirement into the definition of the constitu-
tional injury. With all due respect, we do not think the Constitution
requires this result. “[P]rivate property [is] taken for public use, without
just compensation,” U.S. CONST. amend. V, at the very moment the gov-
ernment takes the private property and fails to compensate the property
owner. The property owner who files an unsuccessful inverse condemna-
tion claim in state court and then somehow manages to have the claim
heard in federal court is owed just compensation from the date of the gov-
ernment’s action taking the property without compensating the owner, not
the date on which the property owner received the final decision in state
court denying the taking claim.
                 GUGGENHEIM v. CITY OF GOLETA              13827
to satisfy Article III,” and considering only whether the plain-
tiff’s case satisfied the prudential requirements). The Court
itself called the Williamson requirements “prudential ripeness
principles.” Id.; see also San Remo 

(Rehnquist, C.J., concurring in the judgment) (noting that the
Court may have purported to “divin[e]” the Williamson
requirements from the text of the Fifth Amendment but later
had held them to be merely prudential).

   The Court’s clarification that Williamson created mere
“prudential requirements” is crucial to our analysis for two
reasons. First, if Williamson were grounded in Article III ripe-
ness, we would be required to raise the issue sua sponte even
though neither party raised it. See Poland v. Stewart, 

, 1104 (9th Cir. 1997) (“An appellate court has a duty to
consider sua sponte whether an issue is ripe for review, in
order to ensure that proper subject matter jurisdiction exists to
hear the case.”). Because Williamson has been held to be
merely a set of prudential, exhaustion-type requirements,
although we asked the parties for their views, we were not
obligated to raise the issue. Compare Day v. McDonough, 

, 202, 205, 209-10 (2006) (holding that AEDPA’s
statute of limitations was akin to an exhaustion requirement,
that it could be waived by the state, and that “district courts
are permitted, but not obligated, to consider, sua sponte” the
issue), with John R. Sand & Gravel Co. v. United States, 

, 752-54, (2008) (holding that the statute of limita-
tions for cases in the United States Court of Federal Claims,
28 U.S.C. § 2501, is akin to a jurisdictional requirement, and
therefore the Court of Appeals for the Federal Circuit was
obligated to raise the timeliness issue despite the govern-
ment’s waiver of it). As Lucas clearly illustrates, some tak-
ings cases will have undisputably satisfied Article III
jurisdictional requirements but will have failed to satisfy the
Williamson prudential requirements. The Court has held that
where constitutional ripeness requirements have otherwise
been met, a court may consider whether to excuse the failure
to satisfy prudential requirements without concern of exceed-
13828            GUGGENHEIM v. CITY OF GOLETA
ing its Article III jurisdiction. 

(cit-
ing practical concerns to justify reaching the merits).

    [4] Second, because Williamson exhaustion is prudential
only, the requirement may be waived or forfeited. See 

, 205 (holding that the state may waive objec-
tions to AEDPA’s statute of limitations, which, like an
exhaustion requirement, is nonjurisdictional); Queen of
Angels/Hollywood Presbyterian Med. Ctr. v. Shalala, 

, 1482 (9th Cir. 1995) (holding that “Medicare’s admin-
istrative exhaustion requirements are jurisdictional in nature”
but may be waived by the Secretary expressly or “involuntar-
ily, through a mistake or omission”); LaDuke v. Nelson, 

, 1323 & n.4 (9th Cir. 1985) (distinguishing
between Article III and prudential standing requirements, not-
ing that one prudential requirement was waived by the plain-
tiff, and finding the requirement did not bar the suit because
the underlying justifications for that prudential limitation
were absent); see also Zhong v. U.S. Dep’t of Justice, 

, 117-25 (2d Cir. 2007) (holding that issue exhaus-
tion in the context of immigration petitions is not “truly ‘juris-
dictional’ in the Article III sense” but rather a prudential
administrative exhaustion requirement, and therefore that the
defense may be waived and that the court has discretion to
review issues not exhausted where it deems the underlying
prudential concerns have been satisfied (citation omitted)).
Here, in its post-Lingle filings before the district court and its
filings on appeal to this court, the City of Goleta forfeited the
claim that this case was not ripe for review by failing to raise
it.

   We note that there is tension in our decisions on this point.
As we recently observed, “[a]lthough the Supreme Court has
described takings claims ripeness as addressing prudential
rather than Article III considerations . . . our Circuit has ana-
lyzed takings claim ripeness as raising both prudential and
Article III considerations.” McClung v. City of Sumner, 

, 1224 (9th Cir. 2008) (citing cases). In particular,
                   GUGGENHEIM v. CITY OF GOLETA                     13829
there is tension between language in three of our decisions. In
Hacienda Valley Mobile Estates v. Morgan Hill, 

(9th Cir. 2003), we concluded that “[b]ecause . . . Hacienda’s
claim is not ripe, we affirm the district court’s dismissal for
lack of subject matter jurisdiction.” 

In that case, we
did not otherwise discuss whether Williamson embraced both
jurisdictional and prudential requirements, nor did we discuss
the impact on Williamson of Palazzolo, Suitum, or Lucas. By
contrast, in Richardson v. City & County of Honolulu, 

(9th Cir. 1997), we recited that “[i]f a claim is
unripe, federal courts lack subject matter jurisdiction and the
complaint must be dismissed,” and we then determined that
the landowners’ takings claim was “not ripe” and “prema-
ture.” 

1161-62 (quotation marks and citation
omitted), 1162. Although the claim was not ripe under Wil-
liamson, we reached the merits anyway and upheld the consti-
tutionality of the ordinance. 

We thus treated the
ripeness inquiry as prudential only. In our latest effort in this
area, McClung v. City of Sumner, we noted the conflict in our
cases and then treated the jurisdictional concerns as an aspect
of Article III and the prudential concerns as the sole inquiry
under 

(“[W]e do not resolve
whether this claim is ripe under the standards articulated in
Williamson, and instead assume without deciding that the tak-
ings claim is ripe in order to address the merits of the appeal.”).6

   [5] In light of the Supreme Court’s unmistakable pro-
nouncements, we think that McClung and Richardson repre-
sent the more considered view. In this case, as in McClung,
there is no question that the Park Owners have satisfied Arti-
  6
    Much of the confusion stems from dicta in our decision in Southern
Pacific Transportation Company v. City of Los Angeles, 

(9th
Cir. 1990), where we noted in the context of a takings claim that “ripeness
is more than a mere procedural question; it is determinative of jurisdic-
tion” and remarked that the lack of ripeness “may be raised sua sponte if
not raised by the parties.” 

Southern Pacific was decided previ-
ous to the Supreme Court’s decisions in Palozzolo, Suitum, and Lucas, and
this language is inconsistent with those decisions.
13830            GUGGENHEIM v. CITY OF GOLETA
cle III requirements, including ripeness. We have held, in
another context, that “the [Article III] ripeness inquiry con-
tains both a constitutional and a prudential component.”
Thomas v. Anchorage Equal Rights Comm’n, 

,
1138 (9th Cir. 2000) (en banc) (emphasis added); see also

(“[R]ipeness doctrine reflects both
constitutional and prudential considerations.”). We stated in
Thomas that the Article III component of the ripeness inquiry
“can be characterized as standing on a timeline,” and that the
real question in cases presenting questions of Article III ripe-
ness is whether “there exists a constitutional ‘case or contro-
versy,’ [and] that the issues presented are ‘definite and
concrete, not hypothetical and abstract.’ 

,
1139 (quotations omitted). See also 

. Where it is clear that a takings plaintiff has standing—
including “standing on a timeline”—and has “presented a
genuine case or controversy sufficient to satisfy Article III,”
the further questions under Williamson of whether a plaintiff
has “received a final decision regarding the application of the
challenged regulations to the property at issue” and whether
the he has “sought compensation through the procedures the
State has provided for doing so” are merely “prudential ripe-
ness requirements.” 

n.7, 734 (internal
alterations and quotation marks omitted). In this case, then,
where the Park Owners have obviously presented a live case
or controversy, 

Part II.A, it is clear that any further
questions under Williamson do not raise the spectre of an
Article III jurisdictional bar. See 

(noting that because “[p]laintiffs’ stake in the legal issues is
concrete rather than abstract . . . the ripeness requirement of
Article III is satisfied.”).

                               3

   Having reviewed the Williamson jurisprudence, we find
that we may reach the merits of the Park Owners’ takings
claim. For the following reasons, “we do not think it prudent
                   GUGGENHEIM v. CITY OF GOLETA                     13831
to apply that prudential requirement here.” 

.

    First, the City forfeited its claim that the case was not ripe
for decision. Because the Williamson requirements are “pru-
dential ripeness principles,” 

-34, and
not Article III jurisdictional limitations, they may be waived
or forfeited. After Lingle prompted the parties to stipulate to
vacate the initial judgment of the district court and return for
litigation at the trial level, the City had the burden to raise any
remaining prudential concerns under Williamson. Instead, the
City was content to continue litigating the claims on the mer-
its. The City expressed no doubts that the record in the case
was fit for a decision by a federal court. Moreover, the City
was not concerned that the Park Owners’ failure to initiate an
inverse condemnation action in state court left any doubt as
to whether the state had yet compelled the City to provide just
compensation to the Park Owners. The City did not, in fact,
mention ripeness at all until prompted by an order from this
court to discuss the issue at oral argument.

   At oral argument, the City acknowledged that as part of its
appeal it had not even considered whether Williamson pre-
vented us from reaching the Park Owners’ takings claims, and
first came to the position that Williamson did so after receiv-
ing our order to be prepared to discuss Williamson at oral
argument. The City argued to us that the case was not ripe
under Williamson, and offered as supporting authority our
opinions in Equity Lifestyle Properties7 and Carson Harbor
Village. Neither of these cases, nor in fact Williamson itself,
appeared in the City’s appellate brief.8 In answer to our ques-
   7
     Equity Lifestyle Props., Inc. v. County of San Luis Obispo (“Equity
Lifestyle Properties”), 

(9th Cir. 2007), withdrawn and super-
seded on denial of rehearing by Equity 

.
   8
     The opinion in Equity Lifestyle Properties was not filed until Septem-
ber 17, 2007, and so could not have been included in the City’s brief, ini-
tially filed March 12, 2007. However, Williamson and Carson Harbor
Village had been decided several years earlier, as had numerous other Wil-
liamson cases, and Equity Lifestyle Properties could have been offered in
a letter under FED. R. APP. P. 28(j).
13832               GUGGENHEIM v. CITY OF GOLETA
tions, the City expressly conceded that the Williamson
requirements were merely prudential and not Article III juris-
dictional requirements. The City argued that its claim was not
waived because we could still exercise our discretion to find
that the case was not ripe because of prudential consider-
ations. The City’s argument lacks merit, however. Because
the Williamson requirements are merely prudential, the claims
can be waived. The fact that we may exercise our discretion
to find the claims unripe does not change the fact that the
claims are waivable, and that in this case the City forfeited them.9

   The fact that the City forfeited its ripeness claim has an
additional, evidentiary implication. It confirms our belief that
the record in this case is eminently ripe for review. William-
son could have resurfaced at the time that Lingle implicitly
foreclosed the Park Owners’ exception from Williamson
based on the “substantially advances” theory. It did not. It
would certainly seem counter-intuitive to us now to think that
a case that had at that point already been litigated through
three rounds—two in federal court and one in state court—
could suddenly become “unripe.” The fact that the City failed
to notice this as well suggests to us that any concerns meant
to be protected by Williamson had been sufficiently protected
by the unusual and lengthy development of the case. See

(holding that the purpose of Wil-
liamson is to develop the record in order to understand the
effect of the challenged regulation).
   9
     Moreover, Equity Lifestyle and Carson Harbor Village do not control
our decision here. Those cases used the general principles of Williamson
and found the property owners’ as-applied takings claims had not been
properly exhausted. See Equity 

-92; Carson
Harbor 

. Neither case dealt with the application
of Williamson in a case where the government failed to raise the ripeness
issue on appeal after the district court implicitly found the takings claims
ripe. Similarly, neither case presented the question of whether we could
find that the property owners had substantially satisfied the prudential con-
cerns embodied in Williamson after three rounds of trial-level litigation in
state and federal court.
                 GUGGENHEIM v. CITY OF GOLETA              13833
   Second, we find that the Park Owners have substantially
satisfied the Williamson requirements. “[I]t is important to
bear in mind the purpose” that the Williamson requirement
serves. 

Williamson held that a facial takings claim
could not be ripe until the property owner has “unsuccessfully
attempted to obtain just compensation through the procedures
provided by the State for obtaining such 

. Here, the Park Owners did, in fact, take this case
to state court. Although they did not file a formal inverse con-
demnation proceeding, they litigated and settled several state
law issues relevant to the alleged taking with the City, includ-
ing issues necessary to establish the timeliness of the takings
claim. They then returned to federal court, without having
been compensated for the taking of their property. There is no
doubt that they have now unsuccessfully attempted to obtain
just compensation through procedures provided by the State.
See 

is just no question that the case is fit for
review. The parties have now litigated this case through three
full rounds at the trial court level. There is no doubt that there
is sufficient evidence in the record to “determine whether
[the] regulation goes ‘too far.’ ” 

(quoting 

). In addition, as we dis-
cuss below, it is undisputed that: (1) the RCO has caused a
significant loss of value to the Park Owners’ property; and (2)
neither the City of Goleta nor the State of California has ever
offered compensation for this loss in value. To the contrary,
the City has litigated against providing compensation continu-
ously since 2002.

  [6] Given the Park Owners’ substantial compliance with the
Williamson requirements, and the City’s forfeiture of the ripe-
ness claim, we believe that Lucas compels us to reach the
merits of this case. “In these circumstances, we think it would
not accord with sound process to insist that [the Park Owners]
pursue the late-created” need to file a formal inverse condem-
nation action in state court “before [their] takings claim can
13834              GUGGENHEIM v. CITY OF GOLETA
be considered ripe.” 

. Just like Lucas,
the Park Owners “[have] properly alleged Article III injury in
fact in this case.” 

to have filed a formal
inverse condemnation claim while already in state court “goes
only to the prudential ‘ripeness’ of [the Park Owners’] chal-
lenge, and for the reasons discussed we do not think it prudent
to apply that prudential requirement here.” 

                                    III

   [7] Having held that we may reach the merits of the Park
Owners’ takings claims, we now turn to those claims.10 As we
have recently summarized, the Supreme Court has identified
three basic categories of regulatory takings claims:

       [1] where government requires an owner to suffer a
       permanent physical invasion of property, see Loretto
       v. Teleprompter Manhattan CATV Corp., 

(1982); [2] where a regulation deprives an
       owner of all economically beneficial use of property,
       see Lucas v. South Carolina Coastal Council, 

(1992); and [3] where the Penn Central
       factors are met, Penn Central Transp. Co. v. New
       York City, 

(1978).

Crown Point Dev., Inc. v. City of Sun Valley, 

,
855 (9th Cir. 2007); see 

. On appeal,
the Park Owners raise only a facial challenge under Penn
Central.
  10
     We review a district court’s ruling on summary judgment that no reg-
ulatory taking has occurred de novo. Gammoh v. City of La Habra, 

, 1122 (9th Cir. 2005). “We must determine, viewing the evi-
dence in the light most favorable to the non-moving party, whether there
are any genuine issues of material fact and whether the district court cor-
rectly applied the substantive law.” 

and citation
omitted).
                GUGGENHEIM v. CITY OF GOLETA             13835
  As described in Lingle,

    The Court in Penn Central acknowledged that it had
    hitherto been unable to develop any set formula for
    evaluating regulatory takings claims, but identified
    several factors that have particular significance. Pri-
    mary among those factors are the economic impact
    of the regulation on the claimant and, particularly,
    the extent to which the regulation has interfered with
    distinct investment-backed expectations. In addition,
    the character of the governmental action— for
    instance whether it amounts to a physical invasion or
    instead merely affects property interests through
    some public program adjusting the benefits and bur-
    dens of economic life to promote the common good
    —may be relevant in discerning whether a taking has
    occurred. The Penn Central factors—though each
    has given rise to vexing subsidiary questions—have
    served as the principal guidelines for resolving regu-
    latory takings claims that do not fall within the phys-
    ical takings or Lucas 

(internal quotation marks and citations
omitted). We must first address each factor in turn, and then
weigh the factors together, in what has famously been
described as an “essentially ad-hoc, factual inquir[y].” Penn

. Before we can apply the Penn Cen-
tral factors, however, we must consider the viability of a
facial challenge under Penn Central, and determine what facts
we may consider when engaging in Penn Central’s ad-hoc
factual inquiry.

                              A

   The Park Owners have brought only a facial challenge to
the RCO under Penn Central—they have not brought a corol-
lary as-applied claim. Unlike an as-applied challenge, which
asserts that a statute or regulation “by its own terms,
13836            GUGGENHEIM v. CITY OF GOLETA
infringe[s] constitutional freedoms in the circumstances of the
particular case,” United States v. Christian Echoes Nat’l Min-
istry, Inc., 

, 565 (1972), a facial challenge
alleges that the statute or regulation is unconstitutional in the
abstract: that “no set of circumstances exists under which the
[a]ct would be valid.” United States v. Salerno, 

,
745 (1987).

   The Park Owners’ decision to refrain from an as-applied
challenge has two important consequences. First, as noted
above, the decision exempts the Park Owners from the “final
decision” prong of Williamson. See Hacienda Valley Mobile

(“Facial challenges are exempt from
the [“final decision”] prong of the Williamson ripeness analy-
sis because a facial challenge by its nature does not involve
a decision applying the statute or regulation.”). Second, the
Park Owners’ decision to cast their Penn Central claim as a
facial challenge places limits on the types of evidence that can
be considered in adjudicating the claim. “In facial takings
claims, our inquiry is limited to ‘whether the mere enactment
of the [regulation] constitutes a taking.’ ” Tahoe-Sierra Pres.
Council, Inc. v. Tahoe Reg’l Planning Agency, 

,
773 (9th Cir. 2000) (quoting 

); see
also Hodel v. Va. Surface Mining & Reclamation Ass’n, 

, 295 (1981). More specifically, in a facial challenge
“we look only to the regulation’s general scope and dominant
features, rather than to the effect of the application of the reg-
ulation in specific circumstances.” 

(internal quotation marks and citation omitted).

   For this reason, the Supreme Court has noted that property
owners bringing a facial takings challenge “face an uphill bat-
tle.” 

n.10; see 

. The fact that the Park Owners have characterized their
facial challenge under Penn Central creates further complica-
tions. In a typical Penn Central claim, the court must consider
factors that will usually not be found in the text of the statute,
such as the economic impact on the claimant and the claim-
                   GUGGENHEIM v. CITY OF GOLETA                     13837
ant’s investment-backed expectations. Nevertheless, when
adjudicating a facial challenge, the court must be careful not
to simply look at “the effect of the application of the regula-
tion in specific circumstances.” 

. The Park Owner’s facial Penn Central claim requires us
to address this apparent paradox: we must confront the ques-
tion of whether a facial challenge under Penn Central is actu-
ally a viable legal claim; and if we determine that it is, we
must then consider what evidence the Park Owners may pre-
sent to prove their claim.

                                    1

   In the district court’s summary judgment ruling of April 5,
2006, the court reviewed the record and engaged in a detailed
Penn Central analysis. Each party had proffered an expert
report in support of its position: the Park Owners proffered a
report by Dr. John M. Quigley,11 and the City responded with
a report by Mr. William Thomsen.12 In its April ruling,
although the district court did not rely on the detailed figures
presented in either report, the district court did credit the core
findings of each report (which we discuss infra). The district
court found in favor of the City under Penn Central. Subse-
quently, the district court issued another summary judgment
ruling on September 6, 2006, in which it purported to address
the Park Owners’ remaining claims. The district court then re-
affirmed its ruling that the Park Owners had not prevailed
under Penn Central. The district court’s ruling was ambigu-
ous, however, as to the basis for its decision. The court was
unclear as to whether it was simply re-incorporating and re-
  11
      Dr. Quigley is a professor of economics, business, and policy at the
University of California, Berkeley and serves as the Director of the Berke-
ley Program on Housing and Urban Policy. He served as President and
Director of the American Real Estate and Urban Economics Association
and has been a member of the National Academy of Sciences/National
Research Council Committee on National Urban Policy.
   12
      Mr. Thomsen is an MBA/CFA with the accounting firm of Grobstein,
Horwath & Company, LLP.
13838           GUGGENHEIM v. CITY OF GOLETA
affirming the Penn Central analysis applied in its April ruling,
or whether it now based its ruling on the new ground that the
Park Owners were precluded from presenting any of the evi-
dence the court had relied on in April because the Park Own-
ers brought only a facial challenge. The district court stated
that the evidence the Park Owners sought to present “at trial”
was “irrelevant” to a facial challenge, and complained that the
Park Owners had “impermissibly attempted to convert this
action, de facto, into an as-applied challenge.” Because of the
necessarily “ad hoc” nature of a Penn Central challenge, if
the district court was adopting a rule that a property owner
may present no evidence of the effect of a regulation on his
property in a facial challenge, the court would essentially be
adopting the rule that there is no such thing as a facial chal-
lenge under Penn Central.

   Similarly, the City’s position to our court on the meaning
of a facial Penn Central challenge is ambiguous. The City has
never argued that a facial challenge under Penn Central is not
a viable legal claim. On the contrary, the City devoted much
of its briefing and oral argument to defending the district
court’s April ruling on the Park Owners’ Penn Central facial
challenge, including the court’s reliance on the core conclu-
sions of the two parties’ expert reports. In defending the con-
clusion of the district court on appeal, the City argues:

    [T]he district court concluded that absent the Ordi-
    nance, Park Owners would have achieved higher
    rates of return. This conclusion credits Park Owners’
    economic evidence and essentially agreed with Park
    Owners that the Ordinance had an economic impact
    on their business operation. It is difficult to imagine
    how the court’s analysis and conclusion regarding
    the Ordinance’s economic impact can be found lack-
    ing.

Elsewhere in its brief, however, the City complains that the
Park Owners have introduced so much evidence as to try to
                 GUGGENHEIM v. CITY OF GOLETA              13839
turn a facial challenge into an as-applied challenge. The City
does not point out which evidence is proper and which is
impermissible in a facial challenge.

   [8] Both logic and Supreme Court precedent support our
conclusion that a facial challenge under Penn Central must
exist as a viable legal claim. Certainly it is apparent that a
facial challenge is easier to mount under either Loretto or
Lucas. It is far easier to prove that a regulation effects a phys-
ical invasion or that it denies an owner of all economically
viable use of his property without considering evidence
beyond the face of the regulation than it is to demonstrate that
the regulation’s effect satisfies the multi-factor test of Penn
Central. However, we have recently described the Loretto and
Lucas tests as categorical “exceptions to the application of the
regulatory takings test” as set forth in Penn Central. Scheehle
v. Justices of the Sup. Ct. of Ariz., 

, 894 (9th Cir.
2007); see 

(“Outside these two rela-
tively narrow categories . . . , regulatory takings challenges
are governed by the standards set forth in Penn Central.”). In
fact, the Supreme Court has emphasized that per se takings
claims are disfavored, whereas Penn Central claims are pre-
ferred. See Tahoe-Sierra Pres. Council, Inc. v. Tahoe Reg’l
Planning Agency, 

, 321, 339, 342 (2002). It
would seem incongruous indeed if only the disfavored excep-
tions to Penn Central could be brought as facial challenges,
where a claim under the general rule of Penn Central could
not.

   [9] Supreme Court precedent also demonstrates the viabil-
ity of a facial challenge under Penn Central. In Keystone, the
Court emphasized the difficulty of prevailing on a facial chal-
lenge under Penn Central, and ultimately concluded that the
mere enactment of the challenged statute did not effect a tak-
ing. See 

. The Court’s ruling
implicitly recognizes that a facial Penn Central challenge is
feasible. Moreover, in Keystone, the Court considered the lim-
ited evidence that the property owners had proffered, includ-
13840              GUGGENHEIM v. CITY OF GOLETA
ing the actual tonnage of coal that the challenged statutes
prevented the owners from removing, and the percentage of
total coal in the mine that the restricted tonnage represented.
See 

& n.24. The Court found that the property
owners’s facial challenge under Penn Central failed because
the evidence the property owners provided was insufficient to
demonstrate economic harm in any significant amount. 

Court found against the property owners not
because the Court was not permitted to consider the evidence
provided, but rather because the property owners’ evidence
did not show that the mere enactment of the statute amounted
to a taking.13 Keystone suggests that a facial Penn Central
challenge is difficult, but viable. Similarly, in Connolly v.
Pension Benefit Guaranty Corp., the Court considered and
rejected a facial Penn Central challenge to the withdrawal lia-
bility provisions of the Multiemployer Pension Plan Amend-
ments Act of 1980. 

, 213, 224-28 (1986); see
also 

(holding that the property
owners’ facial takings claim should have been brought under
Penn Central); 

(ruling on a facial
challenge under Penn Central). Keystone and Connolly dem-
onstrate that a facial challenge under Penn Central may be
difficult, but the mere fact that Penn Central requires an ad-
hoc multi-factor balancing test does not bar a facial challenge.

                                    2

  The fact that the Court’s precedents approve of a facial
challenge under Penn Central requires us to consider what
kinds of evidence beyond the text of the challenged regulation
  13
    To be precise, in Keystone, the property owners expressly appealed
only a facial challenge because they wanted to avoid the expense of pro-
ducing the detailed evidence they believed would be necessary to mount
an as-applied challenge. See 

. Thus, the issue
before the Supreme Court was not how much evidence a property owner
may produce in a facial challenge, it was how little evidence the property
owner could produce and still prevail in a facial challenge. The Court
found the property owners had not produced enough. See 

                 GUGGENHEIM v. CITY OF GOLETA              13841
the reviewing court may consider. A facial challenge seeks to
prove that “the ‘mere enactment’ of the [regulation] consti-
tutes a taking.” 

(quoting 

). Property owners “thus face an uphill battle
in making a facial attack on [a regulation] as a taking.” 

In reviewing a facial challenge under the Takings
Clause, we “look only to the regulation’s general scope and
dominant features, rather than to the effect of the application
of the regulation in specific circumstances.” 

(quotation marks omitted). In a takings case,
however, we are cognizant that the text of the regulation itself
will rarely describe the actual economic effect on property
owners in concrete terms. Thus, the very nature of a takings
inquiry would seem to require that we consider some evidence
outside of the text of the statute. See 

n.24 (discuss-
ing the tension between the demands of a facial challenge and
the necessity of demonstrating the economic impact of the
regulation in a takings claim, and suggesting that there are
still open questions as to what kinds of information may be
considered in a facial takings claim); Garneau v. City of Seat-
tle, 

, 807-08 (9th Cir. 1998) (evaluating a facial
regulatory takings claim, and stating that plaintiffs have the
burden of “introducing evidence of the economic impact of
the enactment . . . on their property”).

   The proper inquiry in a facial challenge is not whether the
property owners can demonstrate that property has been taken
without providing evidence beyond the text of the regulation;
the inquiry is whether the “mere enactment” of the regulation
constitutes a taking. See 

(quotation
marks omitted). Thus, in a takings claim, we must look not
only at what the statute says, but also at what its mere enact-
ment does. See 

. At a minimum,
we must look to the general economic principles that allow us
to interpret the statute’s effect, so that we may understand the
regulation’s general scope and dominant features. Cf. 

(reviewing academic literature to under-
stand the economic effects of a mobile home rent control ordi-
13842           GUGGENHEIM v. CITY OF GOLETA
nance); 

(using economic
principles to understand the impact of a coal mining regula-
tion where Pennsylvania law creates unique, severable inter-
ests in land related to coal mining); 

(using academic literature to develop an analysis to
determine whether a temporary moratorium on development
effects a taking). In addition, there must be a way to under-
stand the economic impact on the complaining property
owner. A property owner who is not permitted at least to pre-
sent evidence that proves that he has actually suffered the
kind of economic harm of which he complains would be pre-
cluded from even proving his own standing to bring the claim
—the property owner must be permitted to adduce evidence
that he has suffered the injury for which he seeks redress. See

; Pennell v. City of San Jose, 

, 6-7 (1988) (requiring the property owners to allege stand-
ing to bring a takings claim by alleging that they are likely to
suffer economic injury by enforcement of the challenged ordi-
nance). Thus, even in a facial challenge, the court may con-
sider evidence related to the individual property owner that
illustrates the economic impact that the mere enactment of the
statute had on that owner and proves that the owner has suf-
fered the injury of which he complains. See 

(considering evidence of the actual tonnage of
coal the regulations rendered unremovable); 

(stating that plaintiffs bringing a facial chal-
lenge “must show that the value of their property diminished
as a consequence” of the regulation); 

n.2 (providing an example using exact dollar amounts as
“illustrative” of the economic impact of the regulation in a
facial challenge).

   In this case, the Park Owners submitted evidence of the
effect that the mere enactment of the RCO had on their prop-
erty. The Park Owners principally relied on the report by Dr.
Quigley. The City did not object to the use of this report; on
the contrary, the City responded by producing its own expert
report by Mr. Thomsen. The district court reviewed both par-
                GUGGENHEIM v. CITY OF GOLETA             13843
ties’ expert reports in preparation for its summary judgment
ruling in April. In conducting its analysis, the district court
did not rely on the detailed information provided in each
report about the actual economic impact of the RCO on each
particular mobile home within the Park, nor did it rely on the
information about the actual impact on the Park as a whole.
Instead, the district court relied on the core findings of the
expert reports and the general findings taken from economic
studies and academic literature about the effects of mobile
rent control ordinances generally. On appeal to this court, the
City has defended the district court’s analysis and its use of
core findings from each party’s expert report. It has argued,
however, that attempts by the Park Owners to provide evi-
dence beyond the core findings of the Quigley Report is an
impermissible attempt to convert a facial challenge into an as-
applied challenge. The City has not identified which evidence
would be so property-specific as to be impermissible in a
facial challenge.

   We need not, however, determine the exact boundaries
between permissible and impermissible kinds of evidence to
support a facial challenge. The City has defended the district
court’s use of core findings from each party’s report. There-
fore, we will confine ourselves to review of these same core
findings in our review of the Park Owners’ facial Penn Cen-
tral challenge. We will provide additional figures from the
Quigley Report only for purposes of demonstrating that the
Park Owners have suffered the actual economic injury of
which they complain and illustrating in concrete terms the
economic impact that the “mere enactment” of the RCO had
in Goleta. In addition, we may consider the district court’s
undisputed factual findings about property values in the City
of Goleta, as these values affect the entire City, and thus
everyone subject to the City’s RCO, and are not specific to
the RCO’s application to the Park Owners. With these limita-
tions in mind, we consider the three factors of the Penn Cen-
tral analysis.
GUGGENHEIM v. CITY OF GOLETA           13845
                               Volume 2 of 2
13846            GUGGENHEIM v. CITY OF GOLETA
                                B

   The three factors described in Penn Central are: (1) the
economic impact of the regulation on the claimant; (2) the
extent to which the regulation has interfered with investment-
backed expectations; and (3) the character of the governmen-
tal action. We consider each in turn.

                                1

   [10] The first consideration under Penn Central is the “eco-
nomic impact of the regulation on the claimant.” 

-39 (quoting Penn 

). There
is no mathematical formula provided by the Constitution, but
“if [the] regulation goes too far it will be recognized as a tak-
ing.” Pa. Coal Co. v. Mahon, 

, 415 (1922). By
definition, under Penn Central, the property owners need not
show a complete deprivation of all economically viable use of
the property. Deprivation of all economically viable use
would entitle the property owners to just compensation under
Lucas, and there would be no need to apply a Penn Central
analysis. See 

(“Where a regulation
places limitations on land that fall short of eliminating all eco-
nomically beneficial use, a taking nonetheless may have
occurred . . . .” ). In sum, to prevail under Penn Central, the
property owner must demonstrate a loss of value that may be
less than 100 percent, but high enough to have “go[ne] too
far.” 

is a broad consensus that a mobile home rent
control ordinance like the RCO causes a wealth transfer from
the mobile home park owners to the incumbent mobile home
tenants. The Quigley Report explained how the RCO affects
the mobile home market. Mobile homes have a divided own-
ership. A park owner owns the real estate, consisting of the
home sites, while the home itself is owned by the tenant who
rents the site. When a jurisdiction enacts a rent control ordi-
nance, the right to occupy a mobile home site at a below-
                    GUGGENHEIM v. CITY OF GOLETA                     13847
market rent acquires its own intrinsic value distinct from the
value of the land. The owner of a given mobile home at the
time the RCO is passed will capitalize this value (equal to the
present value of the future stream of rent discounts) into the
selling price of the home. This is referred to as the “transfer
premium.” A new mobile home tenant, anxious to acquire the
right to regulated, below-market rent, pays the transfer pre-
mium in the form of a higher purchase price for the home.
The net effect is that the cost of the home and the rental site
is approximately the same whether there is an RCO or not; the
difference is that under the RCO, the value of the capitalized
rent is paid to the mobile home owner instead of to the park
owner.14 Accordingly, in the end, the RCO does not actually
  14
     The Quigley Report illustrated these points with figures for the aver-
age property in its sample of dwellings sold in the Park during the relevant
period. The Report estimated that, based on comparable land rental rates,
the annual unregulated market rental rate of the average site in the Park
would be $13,344. The RCO-regulated rental rate on the average site is
only $3,256. Thus, a home owner pays roughly $10,000 less in annual rent
to the Park Owners. This annual savings, however, is reflected in the sell-
ing price of the mobile home. The Report estimated that the average
mobile home, but for the RCO, would be worth $14,037. Because of the
RCO, however, the average mobile home sold for $119,091. This differ-
ence equals $105,054, or a full 88 percent of the entire sale price, and rep-
resents the net present value to the mobile home owner of being able to
save roughly $10,000 a year in rent.
   As a hypothetical, the Quigley Report then calculated what would hap-
pen if a mobile home owner financed the average home with a typical
mortgage product used for these kinds of purchases. The Quigley Report
found that, under the RCO, the average annual housing-related payments
of the purchaser would be: $13,968 in loan repayment plus $3,256 in regu-
lated rent. Without the RCO, because the same home would have sold for
$105,054 less, but rent would have been more, the average annual
housing-related payments would be: $1,646 in loan repayment plus
$13,344 in rent. As the Quigley Report noted, in this particular example,
the mobile home owner would actually be paying more annually under the
City’s RCO than he would in an unregulated market. This is due in part
to the fact that mobile home mortgage products tend to have higher inter-
est rates and their purchasers often have low asset levels or weaker credit
histories. In general, however, the Quigley Report suggests that mobile
13848               GUGGENHEIM v. CITY OF GOLETA
decrease housing costs at all for the new tenant. If a new ten-
ant purchases the home, the new tenant will have to pay an
amount equal to the rental discount in the form of the transfer
premium.

   The Quigley Report summarizes the effect of the RCO:

     For every dollar by which housing costs are reduced
     through lower mobile home rents, consumers are
     forced to pay higher purchase prices for these mobile
     homes. These two effects roughly cancel. Thus, the
     principal effect of the rent control regulation is to
     inhibit increases in the supply of affordable housing
     in the market and consequently to increase rents in
     the local economy. The principal costs are borne by
     those consumers who otherwise would have been
     able to reside in lower cost housing in the region.

The Quigley Report estimated that the RCO forced the Park
Owners to rent the entire Park at close to an 80 percent dis-
count below the market rate. The RCO has resulted in transfer
premiums of approximately 90 percent of the sale price of
mobile homes, enjoyed by the incumbent tenants.

  The district court credited the Quigley Report’s findings
and found that the RCO causes a wealth transfer from the
Park Owners to their tenants. The district court found that
housing costs in the City of Goleta increased “approximately
205% from 1997 to 2003, and increased another 21.1% in
2004. The rent on the rent-controlled spaces in the Park [has]

home owners will end up paying roughly the same amount in annual
expenses whether or not the RCO is in effect. The difference is in who
captures the value of the rent-controlled site in the Park. Without the RCO,
the Park owners receive roughly $10,000 more a year in rent. With the
RCO, the incumbent mobile home owners receive a one-time premium of
$105,054, captured in the sale value of their home.
                   GUGGENHEIM v. CITY OF GOLETA                     13849
not kept up with the increase in housing costs.” The court
found:

       The RCO has resulted in what is known as “transfer
       premiums” in the sale of mobile homes. These trans-
       fer premiums constitute approximately 90% of the
       sale price of mobile homes in the Park. No provi-
       sions in the RCO prevents the seller of a mobile
       home from capturing transfer premiums.

More simply, “an average mobile home worth $12,000 would
sell for approximately $100,000.” The district court con-
cluded that “the uncontroverted facts . . . establish the exis-
tence of a premium.” Indeed, it found that even “[t]he City
has acknowledged the existence of such a premium.”15 The
Supreme Court observed the same wealth transfer phenome-
non in Yee:

       [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 nor-
       mally benefits only the tenant in possession at the
       time the rent control is imposed. Petitioners are cor-
       rect in citing the existence of this premium as a dif-
       ference between the alleged effect of the
       [challenged] ordinance and that of an ordinary apart-
       ment 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
  15
    The City’s own expert, William Thomsen, recognized the existence of
the transfer premium in his report: “residents who departed the Park and
were able to sell their homes at a premium have received an additional
benefit in that the capitalized economic benefit of this rent control could
then be used to finance the purchase of another home or otherwise help
defray occupancy costs elsewhere.”
13850              GUGGENHEIM v. CITY OF GOLETA
       landlord to the incumbent tenant and all future ten-
       ants. By contrast, petitioners contend that the [chal-
       lenged] ordinance transfers wealth only to the
       incumbent mobile home 

(internal citation omitted). The Court in Yee,
however, left open the question of whether the wealth transfer
constitutes a regulatory taking under Penn Central because
the only issue before the Court was whether the wealth trans-
fer constituted a per se taking under Loretto. See 

cases have observed the wealth transfer effect as
well, but the posture of those cases or differences in takings
law when those cases were decided made it unnecessary to
reach the question of whether the wealth transfer effected a
regulatory taking under Penn Central. See, e.g., 

; Levald, Inc. v. City of Palm Desert, 

, 685-89 (9th Cir. 1993); Hall v. City of Santa Bar-
bara, 

, 1276-77, 1279 (9th Cir. 1987) (“[T]he
tenant gets an interest that he can liquidate and take with him
when he leaves the property, or even the City of Santa Barba-
ra.”), overruled on other grounds by 

;
see also Chi. Bd. of Realtors, Inc. v. City of Chicago, 

, 741-42 (7th Cir. 1987) (separate opinion of Posner,
J., with whom Easterbrook J., joined) (detailing empirical
studies and economic analyses showing that rent control regu-
lations reduce the quantity and quality of affordable housing).16
  16
    Academic literature has also discussed the wealth transfer created by
mobile home rent control ordinances. See, e.g., William A. Fischel,
Exploring the Kozinski Paradox: Why Is More Efficient Regulation a Tak-
ing of Property?, 67 CHI.-KENT. L. REV. 865, 872-75 (1991); Werner Z.
Hirsch & Joel G. Hirsch, Legal-Economic Analysis of Rent Controls in a
Mobile Home Context: Placement Values and Vacancy Decontrol, 35
UCLA L. REV. 399, 405, 423-31 (1988); Richard A. Epstein, Rent Control
and the Theory of Efficient Regulation, 54 BROOK. L. REV. 741, 758-59
(1988).
                    GUGGENHEIM v. CITY OF GOLETA                      13851
   [12] The wealth transfer from the Park Owners to their ten-
ants is a naked transfer accomplished by the mere enactment
of the RCO. By taking the value of the Park Owners’ mobile
home sites and transferring it to the Park’s incumbent tenants,
the RCO has effected “the distribution of resources or oppor-
tunities to one group rather than another solely on the ground
that those favored have exercised the raw political power to
obtain what they want.” Cass R. Sunstein, Naked Preferences
and the Constitution, 84 COLUM. L. REV. 1689, 1689 (1984).
In the classic naked transfer, the government takes property
from A to give to B for the sole benefit of B. See Kelo v. City
of New London, 

, 477 (2005) (“[I]t has long been
accepted that the sovereign may not take the property of A for
the sole purpose of transferring it to another private party B,
even though A is paid just compensation.”). In this case, the
RCO works slightly differently, as the government does not
act as a fiscal intermediary. Because of the divided ownership
of mobile homes—the Park Owners own the real estate and
the tenants own the home itself—the transfer can be effected
directly by the mere enactment of the RCO. The RCO takes
wealth from A, the Park Owners, and transfers it to B, the
incumbent tenants, who reap the benefits in the form of
mobile homes worth several times their original value.17

   Incumbent tenants are not the only group that benefit from
the City’s passage of the RCO. The RCO also benefits
another group: those who would like to support affordable
   17
      As a result, the RCO is unlikely to increase the availability of afford-
able housing in the City of Goleta, for the widely-recognized reasons sum-
marized in the Quigley Report. The RCO only affects a small portion of
the total housing market in the City, and because of the potential to capi-
talize the value of the regulated rent into the sale price of the mobile
home, even within the mobile home market, the RCO does not actually
generate mobile home sites that are cheaper to live on than they would be
if rents were unregulated. It is easy to see why mobile home tenants would
encourage the City to adopt the County’s RCO without further investiga-
tion as to whether such regulation was necessary in the real estate market
of 2002.
13852               GUGGENHEIM v. CITY OF GOLETA
housing initiatives without paying for it themselves, for exam-
ple, owners and developers of other forms of housing such as
apartments that might otherwise be forced to provide subsi-
dized housing, and taxpayers who want to subsidize afford-
able housing without actually increasing their own tax
liability to pay for it. See 

(Scalia, J.,
concurring in part and dissenting in part) (“The politically
attractive feature of [rent control] regulation is not that it per-
mits wealth transfers to be achieved that could not be
achieved otherwise; but rather that it permits them to be
achieved ‘off budget,’ . . . .” ) Thus, the City, “solely on the
ground that those favored have exercised the raw political
power to obtain what they want,” has taken from A to give to
B, both for the benefit of B (the incumbent tenants) and for
a larger group, who does not wish to support affordable hous-
ing through more politic means. The Takings Clause does not
prohibit this use of the police power, see 

, but the Takings Clause does not ask us to pretend that
such a naked transfer does not cause a severe, observable eco-
nomic impact on the property owner whose property has been
conscripted for the public’s use.

  The City’s principal argument in response is that, even con-
ceding the wealth transfer, the RCO’s economic impact on the
Park Owners does not amount to a Penn Central taking
because the Park Owners can still earn a return on their invest-
ment.18 The City supplied some evidence in the Thomsen
   18
      The City also claimed that incumbent tenants do not necessarily bene-
fit from the one-time wealth transfer in the form of the “transfer premium”
because the transfer is not realized until the tenants sell their homes, and
they do not all sell their homes. This claim is irrelevant to the point that
real wealth has been transferred. Even if an incumbent tenant does not sell
his mobile home, he may have realized value in it. He might, for example,
be permitted to borrow against the increased value of the home created by
the RCO while he remained in the home. To use the district court’s fig-
ures, an incumbent tenant who purchased his home before the passage of
the RCO for $12,000 could, after the passage of the RCO, take out a home
equity loan against a $100,000 house.
                GUGGENHEIM v. CITY OF GOLETA             13853
Report to show that the Park Owners have earned, depending
on the analysis, roughly 10 percent on their investment annu-
ally. According to the report, this return is, again depending
on the analysis, comparable to or occasionally better than the
return on investment earned by real estate investment trusts
and other kinds of investments according to national indices.

  The district court credited both the Park Owners’ evidence
of the wealth transfer and the City’s evidence of return on
investment. Reviewing the reports together it found:

    Considering all this evidence, a reasonable inference
    that may be drawn is that although Plaintiffs have
    received a rate of return on investment comparable
    to other real estate investments, and although they
    have enjoyed a significant appreciation in value of
    their property, Plaintiffs could have received higher
    rates of return in the absence of the [regulations].

The district court concluded that the wealth transfer was
greater than any return on investment:

    The evidence of the rate of return is mixed. Although
    Plaintiffs have enjoyed a rate of return comparable
    to other real estate investments, Plaintiffs’ evidence
    tends to suggest that they would have earned more—
    perhaps much more—in the absence of the RCO.

Nevertheless, the district court reasoned that because the Park
Owners could receive some return on investment—even
though it was less, perhaps even substantially less, than their
wealth transfer loss—the Park Owners had not suffered a reg-
ulatory taking.

   [13] We disagree with the district court’s reasoning. The
fact that the Park Owners earned some return on investment
is not, as the district court reasoned, the end of their Penn
Central claim. Even if the Park Owners earned some return
13854               GUGGENHEIM v. CITY OF GOLETA
on investment, a taking may have occurred. If the Park Own-
ers could show that the RCO denies them all return on invest-
ment, they could, of course, prevail on a per se takings claim
under Lucas, and we would not have to labor through the
Penn Central analysis. Penn Central thus practically assumes
that the Park Owners may be able to earn some return on
investment. Our challenge under Penn Central is to figure out
what loss of potential return on investment, greater than zero
but less than 100 percent, is significant enough to constitute
a regulatory taking. See 

. The
district court thus erred in the conclusion that because plain-
tiffs can realize a “rate of return comparable to other real
estate investments,” the Park Owners have not suffered signif-
icant economic harm. Cf. 

(“The city’s
argument that [the mobile home park owners] are adequately
compensated by the rents they receive is irrelevant to the
determination of whether a taking has occurred . . . . Whether
compensation is adequate is an inquiry separate from whether
there has been a taking.”).19

   [14] The Park Owners may have enjoyed a positive rate of
return, perhaps even a rate of return comparable to some other
real estate investments, but the district court found, and nei-
ther the City nor the Thomsen Report denies, that the Park
Owners “would have earned more—perhaps much more” if
not for the RCO. Although the “much more” does not appear
   19
      The dissent offers a different rationale from the city and the district
court. Judge Kleinfeld argues that “[t]here is nothing in the record to sup-
port the notion that the Guggenheims’ interest in the trailer park was worth
more before than after the City reenacted the County ordinance.” Dissent-
ing Op. at 13880. As with the question of return on investment, this point
is better addressed as a question of the taking compensation due rather
than whether there was a taking. As Judge Kleinfeld admits, “[i]f this were
a new rent control ordinance . . . this might be an actionable case.” Dis-
senting Op. at 13879-80. It is a new ordinance, which is what makes this
case ripe for review. 

at 13822-34. We simply disagree that “re-
adoption was merely a ministerial re-enactment . . . [and] had no economic
impact on the Guggenheims.” Dissenting Op. at 13879-80.
                    GUGGENHEIM v. CITY OF GOLETA                       13855
to have been reduced to a total dollars-and-cents loss, the dis-
trict court also found—again without contradiction—that the
loss could be as high as almost 90 percent of the sale price on
a site-by-site, home-by-home basis. To illustrate this impact,
the Quigley Report did estimate possible losses for individual
units in the Park, and some of the figures run upwards of
$100,000 per site. By any measure, that is a significant eco-
nomic transfer from the Park Owners to the tenants, one that
must be characterized as a loss for the Park Owners. Cf.
Cienega Gardens v. United States, 

, 1343 (Fed.
Cir. 2003) (finding that an extraction of 96 percent of the
property’s value was severe enough to constitute a taking
under Penn Central). The undisputed evidence shows that the
mere enactment of the RCO has caused a significant eco-
nomic loss for the Park Owners. This factor weighs heavily
in the Park Owners’ favor.

                                      2

   [15] The next consideration is “the extent to which the reg-
ulation has interfered with distinct investment-backed expec-
tations.” 

(quoting Penn 

). Here, it is undisputed that the RCO was passed
in Santa Barbara County in 1979 and amended in 1987, and
that the Park Owners purchased the Park in 1997. The pur-
chase was eighteen years after the RCO was first passed by
the County, but five years before the City of Goleta adopted
the RCO in 2002. We agree with the finding of the district
court, therefore, that the Park Owners “got exactly what they
bargained for when they purchased the Park—a mobile-home
park subject to a detailed rent control ordinance.”20 Thus, we
   20
      The parties stipulated in their state-court settlement agreement that the
RCO, originally a County ordinance, was not in effect for a brief period
during the City’s process of incorporation, as we have previously 

n.2. This fact is relevant to the timeliness of the suit. Nonetheless,
for the purposes of considering the Park Owners’ investment-backed
expectations, the district court found that the RCO had, for all practical
purposes, been in effect “unchanged in substance, for all times relevant to
the present action.”
13856            GUGGENHEIM v. CITY OF GOLETA
take pause at the notion that the Park Owners can claim that
the challenged regulation took between 80 and 90 percent of
the value out of their rental park when, apparently, this value
had been extracted before they purchased the park.

   Our analysis of this issue is controlled by Palazzolo. In that
case, a corporation owned property at the time the govern-
ment enacted the challenged 

.
Palazzolo came into possession of the property in 1978 when
the corporation’s charter was revoked and title to the property
passed, by operation of law, to Palazzolo as the sole share-
holder. 

At that time, the property was already sub-
ject to the regulation that designated the property as part of
protected “coastal wetlands” upon which development would
be limited. 

Island Supreme Court held that
Palazzolo could not, therefore, bring a takings claim because
“[a] purchaser or a successive title holder like [Palazzolo] is
deemed to have notice of an earlier-enacted restriction and is
barred from claiming that it effects a taking.” 

As
the Supreme Court described the state high court’s reasoning,
“by prospective legislation the State can shape and define
property rights and reasonable investment-backed expecta-
tions, and subsequent owners cannot claim any injury from
lost value. After all, they purchased . . . with notice of the lim-
itation.” 

Court reversed:

    The State may not put so potent a Hobbesian stick
    into the Lockean bundle . . . . Were we to accept the
    State’s rule, the postenactment transfer of title would
    absolve the State of its obligation to defend any
    action restricting land use, no matter how extreme or
    unreasonable. A State would be allowed, in effect, to
    put an expiration date on the Takings Clause. This
    ought not to be the rule. Future generations, too,
    have a right to challenge unreasonable limitations on
    the use and value of land.
                   GUGGENHEIM v. CITY OF GOLETA                    13857

Further, the Court pointed out that “[t]he
State’s rule would work a critical alteration to the nature of
property, as the newly regulated landowner is stripped of the
ability to transfer the interest which was possessed prior to the
regulation. . . . A blanket rule that purchasers with notice have
no compensation right when a claim becomes ripe is too blunt
an instrument to accord with the duty to compensate for what
is taken.” 

   The Court’s concern, that a rule precluding post-enactment
purchasers from bringing a regulatory taking claim would
undesirably insulate the government from liability and allow
the state to “secure a windfall for itself,” 

is particu-
larly salient on the facts before us. In 2002, the City of Goleta
adopted the County’s RCO, created to manage housing prob-
lems as they existed in 1979, apparently without any formal
consideration of whether the problems still existed. Were the
fact that the Park Owners purchased the Park when the
County RCO was already in existence sufficient to bar their
takings claim, the City of Goleta would be insulated from lia-
bility for the effects of adopting the RCO when the City
incorporated in 2002. All of the existing park owners at that
time had bought their parks when the land was still part of
unincorporated Santa Barbara County. Unless any of these
park owners had purchased their park prior to the original
RCO enactment in 1979, all the park owners would have pur-
chased with notice of the original RCO. By its own theory,
the City was free to adopt the law with complete impunity,
notwithstanding its obvious effects.

  The Palazzolo Court explained why subsequent property
owners do not lose their right to challenge the government’s
actions:
  21
    The Court limited its reasoning to regulatory takings claims; physical
takings claims resulting from a state’s direct condemnation of property
were distinguished as properly brought only by the property owner at the
time of the 

.
13858           GUGGENHEIM v. CITY OF GOLETA
    Nollan v. California Coastal Comm’n, 

    (1987), presented the question whether it was consis-
    tent with the Takings Clause for a state regulatory
    agency to require oceanfront landowners to provide
    lateral beach access to the public as the condition for
    a development permit. The principal dissenting opin-
    ion observed it was a policy of the California Coastal
    Commission to require the condition, and that the
    Nollans, who purchased their home after the policy
    went into effect, were “on notice that new develop-
    ments would be approved only if provisions were
    made for lateral beach access.” 

(Brennan,
    J., dissenting). A majority of the Court rejected the
    proposition. “So long as the Commission could not
    have deprived the prior owners of the easement with-
    out compensating them,” the Court reasoned, “the
    prior owners must be understood to have transferred
    their full property rights in conveying the lot.” 

n.2.

(internal citations omitted). The
Court also rejected analogies between purchasing a property
subject to a challenged land-use regulation and purchasing a
property whose contours are shaped by background principles
of state law:

    It suffices to say that a regulation that otherwise
    would be unconstitutional absent compensation is
    not transformed into a background principle of the
    State’s law by mere virtue of the passage of title.
    . . . A regulation or common-law rule cannot be a
    background principle for some owners but not for
    others.
    . . . A law does not become a background principle
    for subsequent owners by enactment itself. Lucas did
    not overrule our holding in Nollan, which, as we
    have noted, is based on essential Takings Clause
    principles.
                   GUGGENHEIM v. CITY OF GOLETA                   13859

-30. The Court concluded by
remanding for consideration of Palazzolo’s Penn Central
claim, stating, “[t]hat claim is not barred by the mere fact that
title was acquired after the effective date of the state-imposed
restriction.” 

   [16] We have held that Palazzolo permits property owners
who have purchased property subject to the regulations they
challenge to bring regulatory takings claims under Penn Cen-
tral. See Equity 

(rejecting the
county’s argument that the property owner could not bring a
takings claim because the owner acquired its interest in the
property after the ordinance was passed because “a regulatory
takings claim ‘is not barred by the mere fact that title was
acquired after the effective date of the state-imposed restric-
tion’ ” (quoting 

)); Daniel v.
County of Santa Barbara, 

, 383 (9th Cir. 2002)
(holding that “[t]he Court’s decision last Term in Palazzolo
indicates that in some circumstances a purchaser may have a
valid takings claim even if his or her purchase price was dis-
counted to reflect existing land-use regulations,” and that
Palazzolo applied to regulatory but not physical takings
claims). Our sister circuits have also observed in dicta that
Palazzolo permits post-enactment purchasers to prevail on
regulatory takings claims. See Philip Morris, Inc. v. Reilly,

, 34 n.5, 37 (1st Cir. 2002) (en banc) (describing
the Palazzolo holding as “whether property is acquired before
or after a regulation is enacted does not completely determine
the owner’s reasonable investment-backed expectations”);
Abbott Labs. v. CVS Pharmacy, Inc., 

, 860 (7th
Cir. 2002) (analogizing from Palazzolo to find that a plaintiff
  22
     Finally, we note that even before Palazzolo, the Supreme Court per-
mitted property owners who purchased property subsequent to the enact-
ment of the challenged regulation to bring regulatory takings claims. In
Penn Central itself, one of the appellants, Union General Properties,
acquired its leasehold interest in Grand Central Terminal in 1968, a year
after the Terminal was designated as a landmark in 1967. Penn 

.
13860               GUGGENHEIM v. CITY OF GOLETA
company’s claims against another survive the plaintiff’s
acquisition by another entity).

   [17] Palazzolo left open the question of how to apply the
“investment-backed expectations” analysis to property owners
who purchased subject to the regulation. It merely remanded
the case with instructions to address the merits of Palazzolo’s
claim under Penn 

. Our sister circuits
have yet to address the issue. Penn Central will not aid us
because it never supplied “any ‘set formula’ ” in the first
place. Penn 

. Instead, it “identified
several factors that have particular significance” in what the
Court described as an “ad hoc, factual inquir[y].” 

we must continue to consider “[t]he economic
impact of the regulation on the claimant” and the “character
of the governmental action,” 

must not deem a reg-
ulatory takings claim forfeited simply because the property
changed hands after the regulations went into effect.23

   [18] We read Palazzolo to mean that even though the Park
Owners purchased the Park in a regulated state similar to the
one imposed by the City, the Park Owners may still prevail
under Penn Central. How we are to apply Penn Central post-
Palazzolo is less clear. The question of investment-backed
expectations yields mixed results. On the one hand, as the dis-
trict court found, the Park Owners’ “expectations of the value
of the Park when purchased, as well as the income to be
received from the Park, should have been, at all times, tem-
pered by the knowledge that the RCO would have an adverse
  23
     We thus do not disagree with the dissent’s statement that Palazzolo
does not suggest that the mere “transfer of title revives dead claims.” Dis-
senting Op. at 13880. But as Justice Kennedy stated in Palazzolo, “the
postenactment transfer of title” ought not “absolve the State of its obliga-
tion to defend any action restricting land use, no matter how extreme or

(“[A] regulation that otherwise would be
unconstitutional absent compensation is not transformed into a back-
ground principle of the State’s law by mere virtue of the passage of
title.”).
                    GUGGENHEIM v. CITY OF GOLETA                     13861
effect on their investment.” On the other hand, when the Park
Owners acquired the property, they also arguably acquired the
prior owner’s interest in the property, including the right to
bring a takings action. See 

; see
also CAMSI IV v. Hunter Tech. Corp., 

, 85
(Cal. Ct. App. 1991) (“the harm implicit in a tortious injury
to property is harm to the property itself, and thus to any
owner of the property once the property has been injured and
not necessarily to a particular owner”). At the very least, the
Park Owners have the right to bring a takings action based on
the City’s 2002 adoption of the RCO.

   [19] These two interests are in tension and are, in some
respects, self-referential: The new owner’s investment-backed
expectation depends on the value of any takings claim, but
whether there is a regulatory taking turns on the owner’s
investment-backed expectations. In other words, in this con-
text we cannot address the investment-backed expectation
prong of Penn Central without referring to the merits of the
takings claim, but in order to decide the takings claim, we
must determine the Park Owners’ investment-backed expecta-
tions.24 There is no easy way out of this conundrum. For now
we will acknowledge the dilemma: the Park Owners took pos-
session of the Park knowing that it was subject to the Coun-
ty’s (but not the City’s) RCO. They also assumed ownership
  24
    Judge Kleinfeld is concerned that the current mobile home tenants
may not have received a windfall from the City’s adoption of the RCO
because they invested in reliance on the City’s ordinance. Dissenting Op.
at 13881-82. These are fair concerns that might be implicated if the City
repealed the RCO in the future. See 

(O’Connor, J., concurring). On the other hand, as Justice O’Connor
explained, the alternative is that “the State wields far too much power to
redefine property rights upon passage of title.” Id.; see also 

(Scalia, J., concurring). We do not think concerns over windfalls (or not)
go to whether there has been a taking.
  These are difficult questions, ones that—so far as we know—are
uncharted. To the extent these questions are relevant, they should be
addressed by the district court in the first instance.
13862               GUGGENHEIM v. CITY OF GOLETA
with some hope that they would be able to challenge the RCO
under the Takings Clause and, as they have done here, on
equal protection, due process and state law grounds. We con-
clude, therefore, that the question of investment-backed
expectations is not determinative but must be considered in
tandem with the economic impact of the regulation on the
Park Owners, and the character of the governmental action.
See 

(O’Connor, J., concurring)
(“[I]nterference with investment-backed expectations is one
of a number of factors that a court must examine.”).

                                     3

   [20] The final consideration is “the character of the govern-
mental action.” 

(quoting Penn Cen-

). We have seen two divergent
interpretations of this test, both of which appear to derive
from different portions of Penn Central. We consider each in
turn.25

   One test, applied less frequently in practice, considers
“whether [the governmental action] amounts to a physical
invasion or instead merely affects property interests through
‘some public program adjusting the benefits and burdens of
economic life to promote the common good.’ ” 

(quoting Penn 

). The
  25
     The district court applied yet a third version, whether there was any-
thing “sinister in the purpose of,” or “suspect” or “pretextual” about the
regulations. The court’s use of this test in a Penn Central analysis was in
error, although the test may be relevant to a due process or equal protec-
tion claim. The district court evidently imported this requirement from our
discussion in Armendariz v. Penman, 

(9th Cir. 1996) (en
banc), overruled in part on other grounds as stated in Crown 

. That discussion, in which we considered whether the city’s
stated purpose in passing a housing code was a pretext for other, less noble
purposes, was in the context of an Equal Protection claim that the housing
code unfairly targeted certain property owners. See 

. We did not consider the city’s purpose when undertaking our
takings claim analysis.
                 GUGGENHEIM v. CITY OF GOLETA              13863
application of this test to our case is controlled by Yee, in
which mobile home park owners claimed that a rent control
ordinance almost identical to the RCO amounted to a physical
taking under Loretto. 

. The Supreme
Court held that the rent control ordinance did not amount to
the imposition of a physical invasion. 

however,
proceeded to state in no uncertain terms that the fact that the
regulations caused a one-time wealth transfer from landlord to
the incumbent tenants “might have some bearing on whether
the ordinance causes a regulatory taking.” 

   The district court thought “the character of the governmen-
tal action is less like a per se taking and more like a permissi-
ble shifting of economic benefits and burdens.” We disagree.
Although we understand that the RCO does not amount to a
physical taking, the RCO is substantially more like a “regula-
tory taking,” 

, than a “mere[diminution
of the Park Owners’] property interests through ‘some public
program adjusting the benefits and burdens of economic life
to promote the common good.’ ” 

(quoting Penn 

). The RCO is quite
unlike zoning or other restrictions that apply broadly to busi-
nesses and residences and inevitably restrict the property’s
uses. The Court has explained that its various formulations of
the test for regulatory takings “(reflected in Loretto, Lucas,
and Penn Central) . . . aim[ ] to identify regulatory actions
that are functionally equivalent to the classic taking.” 

effects a transfer of the right to rents for the use of the
property from the Park Owners to the tenants. The Park Own-
ers may own the property on which the mobile homes rest, but
under the RCO the tenants have the right to convey the home
with the right to remain on the site at a much-reduced rent.
This looks much more like a classic taking than a mere regu-
latory burden. This iteration of the “character of the govern-
mental action” test weighs in favor of the Park Owners.

  The second, more frequently applied iteration of the “char-
acter of the governmental action” test considers whether the
13864              GUGGENHEIM v. CITY OF GOLETA
challenged regulation places a high burden on a few private
property owners that should more fairly be apportioned more
broadly among the tax base. See Armstrong v. United States,

, 49 (1960) (“The [Takings Clause] was designed
to bar Government from forcing some people alone to bear
public burdens which, in all fairness and justice, should be
borne by the public as a whole.”); see also 

(discussing Armstrong with approval); First English
Evangelical Lutheran Church of Glendale v. County of Los
Angeles, 

, 318-19 (1987) (applying Armstrong in
a regulatory takings claim); Penn 

.

   We find Cienega Gardens persuasive as to the application
of the Armstrong analysis in this case. 

.
Cienega Gardens found a Penn Central taking where two fed-
eral statutes abrogated property developers’ contractual rights
to prepay their forty-year mortgage loans after twenty years.
See 

The effect of the statutes was to prevent
the developers from exiting the low-rent housing programs in
which they were required to participate while carrying the
loans. See 

These statutes led to a 96 percent loss
of return on equity for the developers. 

Cienega
Gardens found that the government action at issue placed the
expense of low-income housing on a few private property
owners (those who had previously participated in the federal
loan program but now wanted to exit), instead of distributing
the expense among all taxpayers in the form of incentives for
developers to construct more low-rent apartments. 

-39.26
  26
     The City argues that Cienega Gardens involved an abrogation of the
plaintiffs’ contractual property rights whereas this case involves an abro-
gation of the Park Owners’ right to charge market rental rates. This dis-
tinction is not relevant here. Regulatory takings cases necessarily involve
economic analyses, in which the formal characteristics of the transaction
are less relevant than the economic substance. For example, in this case,
the fact that Park Owners are not allowed to raise rents could also be con-
sidered an abrogation of contract rights—their right to contract for annual
                    GUGGENHEIM v. CITY OF GOLETA                     13865
   [21] Here, the RCO applies only to mobile home park own-
ers. The district court found that the City did not impose com-
parable costs on any other property owners in the City, except
as a condition of new development.27 The City has singled out
the Park Owners and imposed solely on them a burden to sup-
port affordable housing. We find the Federal Circuit’s reason-
ing persuasive and applicable to the facts of this case:

     Unquestionably, Congress acted for a public purpose
     (to benefit a certain group of people in need of low-
     cost housing), but just as clearly, the expense was
     placed disproportionately on a few private property
     owners. Congress’ objective in passing ELIHPA[28 ]
     and LIHPRHA[29 ]—preserving low-income housing
     —and method—forcing some owners to keep
     accepting below-market rents—is the kind of
     expense-shifting to a few persons that amounts to a
     taking. This is especially clear where, as here, the
     alternative was for all taxpayers to shoulder the bur-
     den. Congress could simply have appropriated more
     money for mortgage insurance and thereby induced

market-based rent increases. Similarly, the case could be analogized (cre-
atively) to a land-use extraction case: the Park Owners are only permitted
to operate a mobile home park in exchange for an agreement to rent it at
80 percent below existing market rates (which in turn could be analogized
as an extraction that they may rent 20 percent of the park at full market
rates if they agree to permit 80 percent of the tenants to live rent-free).
See, e.g., 

(suggesting that a mobile home rent control
ordinance may be analogized to a land-use extraction and referencing Nol-
lan v. California Coastal Commission, 

(1987)).
   27
      For new developers in Goleta, the burden is substantially less severe.
Although the Park Owners must rent their entire property at an 80 percent
discount, new developers are only required to make 20 percent of their
housing available at below-market rates.
   28
      The Emergency Low Income Housing Preservation Act of 1987, 12
U.S.C. § 17151, note (1988).
   29
      The Low-Income Housing Preservation and Resident Homeownership
Act of 1990, 12 U.S.C. §§ 4101 et seq. (1994).
13866           GUGGENHEIM v. CITY OF GOLETA
    more developers to build low-rent apartments in the
    public housing program to replace housing, such as
    the plaintiffs’, that was no longer part of the pro-
    

; see also Pa. 

(“In
general it is not plain that a man’s misfortunes or necessities
will justify his shifting the damages to his neighbor’s shoul-
ders . . . . [A] strong public desire to improve the public con-
dition is not enough to warrant achieving the desire by a
shorter cut than the constitutional way of paying for the
change.”).

   [22] We do not doubt that the City’s objective in passing
the RCO was to increase the availability of low-cost housing.
Singling out mobile home park owners, however, and forcing
them to rent their property at a discount of 80 percent below
its market value, “is the kind of expense-shifting to a few per-
sons that amounts to a taking.” Cienega 

. Moreover, the City has numerous alternatives for
supporting affordable housing—such as tax incentives, low-
cost loans, rent supports, or vouchers—without directing the
burden at such a limited group. In sum, taking account of the
“character of the governmental action” test in this case also
weighs strongly in the Park Owners’ favor.

                               C

   [23] Having reviewed each factor individually, we must
weigh them together. We conclude that the RCO has caused
substantial economic hardship to the Park Owners. Property
values in the area have increased by 225 percent in the time
that the Park Owners have owned the Park, yet the Park Own-
ers have not been permitted to increase rents beyond 75 per-
cent of the annual increase in the CPI. This is a zero-sum
game; loss to the Park Owners has become gain to their ten-
ants. The RCO has forced the Park Owners to rent their prop-
erty at an 80 percent discount below the market value,
                   GUGGENHEIM v. CITY OF GOLETA                  13867
resulting in transfer premiums equal to approximately 90 per-
cent of the selling price of a mobile home. Thus, the savings
created by these below-market rents are transferred directly
into the pockets of the incumbent mobile home tenants, who
can now sell their mobile homes for almost ten times their
purchase price. See 

. Next, we agree with
the district court that the RCO has not strongly interfered with
the Park Owners’ investment-backed expectations because the
Park Owners purchased the Park when the Park was already
regulated. Nevertheless, the mere fact that the Park Owners
bought the Park in its regulated state does not mean that the
City has not taken property by regulation or that the Park
Owners cannot bring such a claim. See 

-28. Finally, we conclude that the RCO looks more like a
classic taking than a mere shifting of benefits and burdens,
see 

, and that the RCO singles out mobile
home park owners and forces them to bear a burden of pro-
viding affordable housing in the City that should fairly be
born by the taxpayers as a whole. See 

.

   [24] On balance, the City’s RCO “goes too far” and consti-
tutes a regulatory taking under the Fifth and Fourteenth
Amendments for which just compensation must be paid. If the
City of Goleta wishes to attempt to increase the availability of
affordable housing by transferring the value of renting land
within its jurisdiction from the Park Owners to the incumbent
tenants, there is no constitutional impediment to doing so.30
The Fifth Amendment of the U.S. Constitution, however,
requires that the City compensate the Park Owners for taking
their property by regulation.
  30
    The Park Owners have not claimed that the government action is
impermissible because it fails to meet the “public use” requirement. See

.
13868           GUGGENHEIM v. CITY OF GOLETA
                              IV

   The Park Owners also ask us to strike down the RCO as a
violation of the Due Process Clause. In Lingle, the Supreme
Court clarified the difference between a challenge to a rent
control ordinance as a takings claim and as a substantive due
process claim, and affirmed the independent vitality of both
theories. “[The Takings Clause] is designed not to limit the
governmental interference with property rights per se, but
rather to secure compensation in the event of otherwise proper
interference.” 

. As we have explained:

    Due process violations cannot be remedied under the
    Takings Clause, because if a government action is
    found to be impermissible—for instance because it
    fails to meet the ‘public use’ requirement or is so
    arbitrary as to violate due process—that is the end of
    the inquiry. No amount of compensation can autho-
    rize such action.

Crown 

(quoting Equity Lifestyle Prop-

n.16 (citation omitted)). The Park
Owners have raised two different theories to support their due
process claim, which we address in turn.

                               A

   [25] The Park Owners’ more “traditional” due process the-
ory is foreclosed by precedent. The Supreme Court and we
have upheld rent control laws as rationally related to a legiti-
mate public purpose. See Pennell v. City of San Jose, 

, 13 (1988); Equity 

; Carson Har-
bor 

. In fact, we have already held that
a mobile home park rent control ordinance similar to the one
at issue survives a due process challenge. See Carson Harbor

. As in this case, the law challenged in
Carson Harbor Village was ineffective at preserving low-
                GUGGENHEIM v. CITY OF GOLETA              13869
income housing and merely caused a wealth transfer from the
park owners to incumbent tenants. We held:

    A generally applicable rent-control ordinance will
    survive a substantive due process challenge if it is
    designed to accomplish an objective within the gov-
    ernment’s police power, and if a rational relationship
    existed between the provisions and the purpose of
    the ordinances . . . . This deferential inquiry does not
    focus on the ultimate effectiveness of the law, but on
    whether the enacting body could have rationally
    believed at the time of enactment that the law would
    promote its objective.

marks and citations omitted).

                               B

   The Park Owners also raise a second due process theory:
that the RCO is a denial of substantive due process because
it fails to guarantee that they will earn a “fair and reasonable
return” on their investment. For this claim, the Park Owners
attack specifically RCO §§ 11A-5 and 11A-6. These sections
detail the “automatic increase” of 75 percent of the CPI, and
the procedures for requesting a “discretionary increase” where
increased operating costs, capital expenses, and capital
improvements are greater than the amount of the automatic
increase. See RCO §§ 11A-5, 11A-6. The Park Owners argue
that these provisions will necessarily lead to a time when the
Park Owners are denied rent increases that permit a reason-
able return on their investment. They also claim that the pro-
visions are constitutionally infirm because they provide no
mechanism by which Park Owners can challenge such rent
caps and secure a reasonable return.
13870               GUGGENHEIM v. CITY OF GOLETA
                                     1

   The Park Owners rely on a thin, but viable line of cases. In
Sierra Lake Reserve v. City of Rocklin, we considered another
mobile home park rent control ordinance that, like the one at
issue, imposed rent caps and caused a wealth transfer to
incumbent tenants. 

, 954 (9th Cir. 1991),
vacated, 

(1992).31 The plaintiffs argued that the
provisions of the ordinance permitted only a passing through
of increased costs, without allowing for a reasonable profit.

n.9. Reversing the district court’s dismissal on a
12(b)(6) motion, we held that the park owner had alleged a
viable substantive due process claim to the extent that the rent
control ordinance deprived it of a “fair and reasonable” return
on its investment by prohibiting rent increases designed to
capture a return on investment in capital improvements:

       Under Guaranty National [Ins. Co. v. Gates, 

(9th Cir. 1990)], every dollar the landlord
       puts into the property by way of capital improve-
       ments constitutes an investment in the property for
       which a ‘fair and reasonable’ return must be
       allowed. Breaking even is not enough; the law must
       provide for a profit on one’s investment. . . . To the
       extent plaintiff alleges that the rent increases allowed
       on account of capital improvements merely offset
       the cost of those improvements (or less), it has stated
       a claim for a violation of substantive due process
       under Guaranty National.

(internal citations omitted).

   [26] The Park Owners’ claim fails because they have only
  31
    Subsequent to a partial reversal by the Supreme Court on other
grounds, we vacated a portion of Sierra Lake, but retained the portions rel-
evant to this discussion. See Sierra Lake Reserve v. City of Rocklin, 

, 663 (9th Cir. 1993).
                 GUGGENHEIM v. CITY OF GOLETA               13871
brought a facial challenge to the RCO. A facial challenge to
a law’s constitutionality is “the most difficult challenge to
mount successfully, since the challenger must establish that
no set of circumstances exists under which the [law] would be
valid.” United States v. Salerno, 

, 745 (1987).
“The fact that [a challenged law] might operate unconstitu-
tionally under some conceivable set of circumstances is insuf-
ficient to render it wholly invalid . . . .” Id; see also 

(1987) (holding that, in a facial challenge,
the issue is whether the “mere enactment of the [regulation]
constitutes a taking”).

   [27] The Park Owners argue that the RCO necessarily
denies them a just and reasonable return on their capital
investments. This contention is belied by the text of the provi-
sion. Section 11A-5 provides that annual rent may be
increased by 75 percent of the CPI (the “automatic increase”).
The Park Owners, who are not satisfied with their RCO-
prescribed rent increase may seek arbitration for more rent to
cover actual expenses. Such rents are in addition to the auto-
matic increase as a “just and reasonable return on invest-
ment.” It is plain enough from this scheme that the RCO
makes some allowance “for a profit on one’s investment” and
not “merely [an] offset [for] the cost of those improvements.”
Sierra 

. Although the RCO may not pro-
vide a full return on investment in every case, we are satisfied
that the RCO provides for a reasonable profit in at least some
circumstances. We recognize that there may be some impreci-
sion between what the RCO will provide as a return and what
the Park Owners might consider a reasonable return, but the
Due Process Clause does not demand a perfect fit between the
economic regulatory scheme and its purpose. See Reno v. Flo-
res, 

, 305 (1993) (holding that the Due Process
clause demands no more than a “reasonable fit” between gov-
ernmental purpose and the means chosen to advance that pur-
pose); Williamson v. Lee Optical of Okla., Inc., 

,
487-89 (1955); Nat’l Ass’n for the Advancement of Psycho-
analysis v. Cal. Bd. of Psychology, 

, 1050-51
13872               GUGGENHEIM v. CITY OF GOLETA
(9th Cir. 2000); Carson Harbor 

; Mor-
seburg v. Balyon, 

, 979-80 (9th Cir. 1980).
Because there are circumstances under which the law would
be valid, the Park Owners’ facial challenge must fail. See

.32

                                     2

   [28] The Park Owners also argue that the RCO violates due
process because the provision provides no procedural “mech-
anism” by which they can file a grievance if they are not earn-
ing a just and reasonable return, such as a “discretionary
application or provision.” See RCO § 11A-5(i)(1) (“The arbi-
trator shall have no discretion to award additional amounts as
a just and reasonable return on investment.”). This argument,
while creative, is an end-run around the Park Owners’ previ-
ous argument. It fails for the same reason we rejected the
prior claim. Although the RCO may lack a process for adjust-
ing the reasonable rate of return similar to the arbitration pro-
cess for adjusting the discretionary increase to cover operating
costs and capital expenses, the absence of a process is only
relevant when the Park Owners can demonstrate they have
actually been denied a reasonable return. This claim must be
addressed as an as-applied challenge, not a facial challenge.

  As an alternative argument, the Park Owners attempt to
ease the “uphill battle” they face on their facial challenge,

, by arguing that they should be
excused from having to go through the hearing provisions set
out in the RCO under the “futility doctrine.” Under the futility
  32
    Because the Park Owners’ facial challenge fails we do not address
whether the district court erred in holding that, as a threshold requirement
to raise a “just and reasonable return” claim, the Park Owners must first
prove that the government’s actions were “arbitrary and unreasonable,
having no substantial relation to the public health, safety, morals, or gen-
eral welfare.” See Sierra 

; see also Fed. Power
Comm’n v. Natural Gas Pipeline Co., 

, 581-82, 585-86
(1942); Guar. Nat’l 

(9th Cir. 1990).
                 GUGGENHEIM v. CITY OF GOLETA              13873
doctrine, a claimant may bypass the procedures for relief
included in the challenged law if such procedures are shown
to be “unvailable or inadequate.” Equity 

(quoting 

).

   [29] The futility exception applies only if the challenger
has already attempted to use the state procedures “and has
shown pursuit of such remedies would be futile.” Equity Life-

. The record contains no evidence that
the Park Owners have attempted to use the RCO procedures,
much less proven them constitutionally inadequate. It would
be mere speculation for us to accept the Park Owners’ unsub-
stantiated claims that a request for a rent increase sufficient to
secure a reasonable return would be denied. See 

(“Because petitioners do not claim to have run that
gauntlet, . . . this case provides no occasion to consider how
the procedure has been applied to petitioners’ property, and
we accordingly confine ourselves to the face of the statute.”);
see also Equity 

.

                                V

   [30] Finally, the Park Owners argue that the RCO violates
the Equal Protection Clause because it singles out mobile
home park owners, as opposed to other sorts of housing pro-
viders, to bear the burden of an affordable housing program.
This argument is governed by our decision in Equity Lifestyle.
In that case, we held that a mobile home rent control ordi-
nance does not violate the Equal Protection Clause because it
is rationally related to the legitimate public interest of promot-
ing affordable housing. Equity 

.
This is true even if the statute singles out mobile home park
owners, does not increase the amount of available affordable
housing, and “serve[s] the sole purpose of transferring the
value of [the park owners’] property to a select private group
of tenants.” 

13874            GUGGENHEIM v. CITY OF GOLETA
                               VI

   [31] State and local governments have a legitimate interest
in increasing the availability of affordable housing for their
citizens. Translating that interest into effective public policy,
however, has proven difficult. The Supreme Court and our
court have addressed regulations like the City’s RCO with
some regularity; we have consistently questioned their inef-
fectiveness at increasing the availability of affordable hous-
ing, and we have commented on their pernicious side effects.
See, e.g., 

; Sierra 

; Carson Harbor 

-73; cf. Rich-
ardson, 

(reviewing a condominium rent con-
trol ordinance with similar effects). Nevertheless, so long as
these rent control ordinances are “designed to accomplish an
objective within the government’s police power, and if a
rational relationship existed between the provisions and the
purpose of the ordinances,” the Constitution affords state and
local governments the flexibility to experiment to find a
workable approach to the problem. Carson Harbor 

. We therefore affirm the district court’s find-
ings that the City’s RCO does not, on its face, violate the Due
Process Clause or the Equal Protection Clause.

   When such ordinances “go[ ] too far,” however, and require
some property owners to support policies that “in all fairness
and justice, should be borne by the public as a whole,” the
Constitution requires that the government provide just com-
pensation. 

(citation omitted). The Wil-
liamson prudential ripeness requirements have, for the most
part, forced us to close the courthouse door to aggrieved prop-
erty owners like the Park Owners, and to close our eyes to the
extreme effects of laws like the City’s RCO. The Park Own-
ers, however, have managed to pry these doors open a bit by
developing their case through three rounds of litigation in
state and federal court, and the City has forfeited any objec-
tion that the case is not fit for review. We will not, therefore,
throw these property owners back out and slam the court-
                GUGGENHEIM v. CITY OF GOLETA                13875
house door shut behind them. Today, our eyes are open. We
have weighed the Penn Central factors, and we find that the
RCO has effected a regulatory taking. Just compensation is
due.

   [32] We therefore reverse the district court’s judgment on
the takings claim and remand to the district court for further
proceedings. On remand, the district court may of course con-
sider any materials presented by either party that are relevant
to determining the total amount of just compensation due to
the Park Owners. See, e.g., Cienega 

. As noted in Part III.A.1, the district court here did not
consider the detailed figures included in either of the expert
reports presented before it, possibly because it found that such
evidence was precluded under a facial takings challenge under
Penn Central. We have now held that a facial challenge under
Penn Central exists as a viable legal claim, 

pp.
13839-40, and affirmed that this court’s precedents and the
nature of a takings inquiry allow for some evidence outside
the text of the statute to be admissible. 

The
district court may therefore properly consider such “detailed
figures,” in addition to any other evidence it deems relevant,
in conducting its analysis to ascertain the precise amount of
just compensation owed to the Park Owners. See, e.g., Rich-

n.2 (noting that an example using
exact dollar amounts is “illustrative” of the economic impact
of the regulation in a facial challenge).

  Costs shall be awarded to the Appellants.

  AFFIRMED        in   part,   REVERSED        in   part,    and
REMANDED.



KLEINFELD, Circuit Judge, dissenting:

  I respectfully dissent.
13876               GUGGENHEIM v. CITY OF GOLETA
   I agree with the majority that the prudential ripeness
requirement of Williamson County Regional Planning Com-
mission v. Hamilton Bank of Johnson City1 does not preclude
a decision on the merits, and I agree with the majority that the
rent control ordinance would amount to a regulatory taking
under Penn Central Transportation Co. v. New York City,2
were it not a re-enactment of one already in effect when the
Guggenheims purchased the trailer park. But I cannot agree
that there was a taking of anything for which the Guggen-
heims would be entitled to compensation, because they pur-
chased the park after the regulatory takings that mattered.

   The challenged rent control ordinance was first passed by
Santa Barbara County (the “County”) in 1979, and revised in
1987. The Guggenheims bought the trailer park in 1997,
which was, at the time, located in an unincorporated part of
the County. When the Guggenheims bought the trailer park,
the County had long since taken away much of the rising
value of the fee from the landlord and given it to the tenants
who then owned trailers at the park. By 1997 the purchase
price of the trailer park reflected the lower value of the trailer
park to the landlord under the ordinance.3 All the Guggen-
heims paid for was a trailer park burdened by the rent control
ordinance. And when they bought, the statute of limitations
had long since run on any takings claims arising from the
County’s 1979 and 1987 rent control ordinances.

   The parties have stipulated that there was a short period
during the day of February 1, 2002 — the day the City of
Goleta (the “City”) was incorporated — when the County rent
control ordinance did not apply. Later that same day, pursuant
to California statute, the City re-adopted the rent control ordi-
nance.4 Accordingly, the Guggenheims’ lawsuit is not barred
  1
    

(1985).
  2
    

(1978).
  3
    Majority op. at 13850-51.
  4
    California code requires a newly incorporated city to adopt “prior to
performing any other official act, [ ] an ordinance providing that all county
                   GUGGENHEIM v. CITY OF GOLETA                    13877
by the statute of limitations because they challenge the City’s
adoption, as part of its incorporation, of the County rent con-
trol ordinance, which followed the brief period when the ordi-
nance was not in effect. Because the ordinance amounts to a
regulatory taking, and not a physical taking,5 the Guggen-
heims’ challenge must be analyzed as regulatory taking.

   We disagree on how to apply the controlling Supreme
Court decision, Palazzolo v. Rhode Island.6 In that case, a sin-
gle shareholder owned a corporation that held land burdened
by regulations.7 When the corporation was dissolved, title to
the burdened land passed to the sole shareholder by operation
of law.8 The government claimed that because the sole share-
holder, the plaintiff, did not own the land when its use was
restricted, he had no claim for compensation on account of
any taking that had occurred when his corporation held title.9
In this factual circumstance, the Court held that the plaintiff
could pursue a claim for compensation, 5 to 4. Five justices
wrote for the Court that a regulatory takings claim “is not
barred by the mere fact that title was acquired after the effec-
tive date of the state-imposed restriction.”10 Justice
O’Connor’s concurrence states that the claim was not barred
in the circumstances presented in that case.11 Her stated rule

ordinances previously applicable shall remain in full force and effect as
city ordinances for a period of 120 days after incorporation, or until the
city council has enacted ordinances superseding the county ordinances,
whichever occurs first.” Cal. Gov’t Code § 57376. The City did this on
February 1, 2002. On April 22, 2002, the City re-adopted the entire
County Code, including the rent control ordinance, for an indefinite
period, subject to the City’s power to amend, repeal, or modify the Code.
   5
     Yee v. City of Escondido, 

(1992).
   6
     

(2001).
   7
     

   8
     

   9
     

   10
      

   11
      

(O’Connor, J., concurring).
13878               GUGGENHEIM v. CITY OF GOLETA
rejects treating a change of ownership before or after the
enactment of the regulation as per se barring or not barring a
takings claim.12 Instead, courts “must attend to those circum-
stances which are probative of what fairness requires in a
given case.”13 The four dissenters and Justice O’Connor
agreed that acquiring title after the taking could bar a takings
claim.14

   In Palazzolo, the transfer of title (by operation of law) had
no effect on the wealth of the plaintiff. He merely gained per-
sonal title to what he previously owned as 100% shareholder
of the corporation which held title.15 By contrast, in this case,
the Guggenheims bought the trailer park at a price presum-
ably reflecting the impact of the rent control on the prior own-
ers, a price lower than what they would have had to pay
without the rent control ordinance. The Guggenheims pur-
chased at arms length a trailer park already devalued by rent
control. The land in Palazzolo was devalued by the chal-
lenged regulations while the plaintiff owned the impacted
economic interest as a 100% shareholder in the corporation
holding the land.16

   We have two decisions on point. Daniel v. County of Santa
Barbara holds that although Palazzolo rejects a rule that a
purchaser who is aware of existing land-use regulations may
never pursue a takings claim, it “did not adopt the converse
of that rule,” that the successor could always recover.17 Daniel
distinguishes Palazzolo on two grounds. One is that Palazzolo
was a regulatory taking (as is the case at bar), while Daniel
  12
      

636.
  13
      

   14
      

(Stevens, J., concurring in part and dissenting in part); 

n.3 (Ginsburg, J., dissenting); 

-55 (Breyer, J., dissenting).
   15
      

(majority opinion).
   16
      Id.
   17
      

, 384 (9th Cir. 2002).
                  GUGGENHEIM v. CITY OF GOLETA                13879
was a physical taking.18 The second is that “the full value of
the [taking] had already been taken from the Daniels’ pre-
decessors, it took nothing of value from the Daniels.”19 This
second ground applies to the case at bar.

   Equity Lifestyle Properties, Inc. v. County of San Luis
Obispo involved another trailer park rent control ordinance.20
Our decision in that case discusses, but explicitly declines to
decide whether a plaintiff who had purchased after the ordi-
nance went into effect had a claim under Palazzolo.21 Instead,
it rejects the takings claim as barred by the statute of limita-
tions.22 Since the takings claim failed on an another ground,
Equity Lifestyle did not need to distinguish Daniel.

   Our case fits the second Daniel distinction from Palazzolo,
“[b]ecause the full value of the [taking] had already been
taken from the Daniels’ predecessors, it took nothing of value
from the Daniels.”23 It also fits the limitation Justice
O’Connor imposed in Palazzolo. Her opinion, and Daniel,
would both have us “attend to those circumstances which are
probative of what fairness requires in a given case.”24 Since
the Guggenheims benefitted from a lower purchase price
reflecting the burden of the rent control ordinance when they
bought the trailer park, fairness does not require that they be
compensated. Taking from Peter does not require giving com-
pensation to Paul.

  If this were a new rent control ordinance and a previous
owner had transferred the trailer park to the Guggenheims
  18
     

  20
     

(9th Cir. 2008).
  21
     

n.11.
  22
     

& n.15.
  23
     

.
  24
     

(O’Connor, J., concurring).
13880              GUGGENHEIM v. CITY OF GOLETA
before the statute of limitations had barred the seller’s claim,
then this might be an actionable case. But it is not. The naked
wealth transfer was in the 1970’s. The 2002 re-adoption was
merely a ministerial re-enactment that did not transfer wealth
from the Guggenheims.

   Or, if there had been a substantial period of time, instead
of less than one day, when the rent control ordinance had not
been in effect, and if the Guggenheims had bought at a price
reflecting freedom to charge market rents, they might have
suffered an impairment of their investment-backed expecta-
tions or a negative economic impact.25 But that is a hypotheti-
cal circumstance neither argued nor, on the facts of this case,
arguable. There is nothing in the record to support the notion
that the Guggenheims’ interest in the trailer park was worth
more before than after the City reenacted the County ordi-
nance.26 The reenactment had no economic impact on the Gug-
genheims.27

    The Guggenheims cannot demonstrate any investment
backed expectations that were harmed by the 2002 reenact-
ment of the ordinance unless they breathe life into the takings
claims that prior owners never brought. When the prior own-
ers let the statute of limitations run without challenging the
1970’s ordinance and the 1987 reenactment, their claim
expired. Palazzolo does not undermine the rule that “a takings
claim must [ ] comply with timeliness requirements.”28 The
time-barred claims could not establish investment backed
expectations. Palazzolo does not suggest that a transfer of title
revives dead claims. Instead, Palazzolo holds that transfer of
title will not necessarily bar a takings claim, a quite different
  25
     

note 4 and accompanying text.
  26
     Garneau v. City of Seattle, 

, 807-08 (9th Cir. 1998).
  27
     See Lingle v. Chevron U.S.A. Inc., 

, 538-39 (2005) (quot-
ing Penn 

).
  28
     Equity Lifestyle Props., Inc. v. County of San Luis Obispo, 

, 1190 (9th Cir. 2008).
                  GUGGENHEIM v. CITY OF GOLETA                    13881
proposition. In one case, there is no longer a claim and in the
other, there is a claim that has changed hands.

   The Guggenheims purchase of the trailer park in 1997 did
not breathe life into the dry bones of the takings claim that
had died years before.29 As the majority opinion concedes, the
Guggenheims “got exactly what they bargained for when they
purchased the Park—a mobile-home park subject to a detailed
rent control ordinance.”30 The City took nothing from what
they bought.

   The third factor analyzed by the majority, the “character of
the government action,”31 is a continuation of the old ordi-
nance, the same one that applied when the Guggenheims
bought the trailer park. The brief gap and readoption did not
reapportion public burdens, as did the 1987 and 1979 ordi-
nances.32

   Because the rent control ordinance did not harm the Gug-
genheims, they do not have a regulatory takings claim under
Penn Central Transportation Co. v. New York City,33 or the
test Justice O’Connor set forth in Palazzolo, which forces us
to consider “what fairness requires in a given case.”34 Fairness
cuts the other way.

   Unfairness arises in this case in quite another quarter
because of the market distortions created by the rent control
ordinance during the years after its enactment in 1979. As the
majority explains, the rent control ordinance has had the
effect of raising the average price of a trailer in the park by
  29
     Ezekiel 37:1-14 (King James).
  30
     See 

(quoting Penn 

).
  31
     

).
  32
     Armstrong v. United States, 

, 49 (1960)
  33
     

(1978).
  34
     

(O’Connor, J., concurring).
13882              GUGGENHEIM v. CITY OF GOLETA
$105,054, 88% of the sale price.35 But for the rent control
ordinance, the average trailer would be worth only $14,037.
The people who really do have investment backed expecta-
tions in this circumstance are those who have bought trailers
since rent control went into effect. Tenants come and go, and
even though rent control transfers wealth to “the tenants,”
after a while, it is likely to affect different tenants from those
who benefitted from the transfer. The present tenants lost
nothing on account of the City’s reinstitution of the County
ordinance. But they would lose, on average, over $100,000
each, if the rent control ordinance were repealed. They have
no legal protection against repeal, and have invested, essen-
tially, in reliance on the stability of government decisions that
create market distortions.36 Repeal would not amount to a tak-
ing, but continuation of the ordinance deprives no one, not the
plaintiffs and not the tenants, of any compensable value.




  35
      Majority Op. at 13837 n.11.
  36
      Madera Irrigation Dist. v. Hancock, 

, 1403 (9th Cir.
1993) (holding that “[r]easonable expectations arising out of past policy
but without a basis in cognizable property rights may be honored by pru-
dent politicians, because to do otherwise might be unfair, or because vola-
tility in government policy will reduce its effectiveness in inducing long
term changes in behavior. But violation of such expectations cannot give
rise to a Fifth Amendment claim.”); see also Nat’l Union Fire Ins. v.
United States, 

, 1422 (9th Cir. 1997); Peterson v. U.S. Dep’t
of Interior, 

, 812-13 (9th Cir. 1990).