City of Oakland v. Wells Fargo & Company

FOR PUBLICATION

UNITED STATES COURT OF APPEALS
FOR THE NINTH CIRCUIT

CITY OF OAKLAND, A Municipal No. 19-15169
Corporation,
Plaintiff-Appellee, D.C. No.
3:15-cv-04321-
v. EMC

WELLS FARGO & COMPANY; WELLS
FARGO BANK, N.A., OPINION
Defendants-Appellants.

Appeal from the United States District Court
for the Northern District of California
Edward M. Chen, District Judge, Presiding

Argued and Submitted En Banc June 23, 2021
Pasadena, California

Filed September 28, 2021

Before: Sidney R. Thomas, Chief Judge, and M. Margaret
McKeown, Kim McLane Wardlaw, Richard A. Paez,
Consuelo M. Callahan, Sandra S. Ikuta, Jacqueline H.
Nguyen, Andrew D. Hurwitz, Ryan D. Nelson, Bridget S.
Bade, and Lawrence VanDyke, Circuit Judges.

Opinion by Judge McKeown
2 CITY OF OAKLAND V. WELLS FARGO & CO.

SUMMARY *

Fair Housing Act

The en banc court affirmed in part and reversed in part
the district court’s partial grant and partial denial of Wells
Fargo’s motion to dismiss and remanded for dismissal of the
City of Oakland’s claims under the Fair Housing Act,
alleging that Wells Fargo’s discriminatory lending practices
caused higher default rates, which in turn triggered higher
foreclosure rates that drove down the assessed value of
properties, and which ultimately resulted in lost property tax
revenue and increased municipal expenditures.

The en banc court held that under Bank of America
Corp. v. City of Miami, 137 S. Ct. 1296 (2017),
foreseeability alone is not sufficient to establish proximate
cause under the Fair Housing Act, and there must be “some
direct relation between the injury asserted and the injurious
conduct alleged.” The en banc court held that the
downstream “ripples of harm” from Wells Fargo’s alleged
lending practices were too attenuated and traveled too far
beyond Wells Fargo’s alleged misconduct to establish
proximate cause.

The en banc court affirmed the district court’s dismissal
of the City’s damages claim related to increased municipal
expenditures and reversed the district court’s denial of Wells
Fargo’s motion to dismiss the damages claim related to lost

*
This summary constitutes no part of the opinion of the court. It
has been prepared by court staff for the convenience of the reader.
 CITY OF OAKLAND V. WELLS FARGO & CO. 3

property tax revenue and claims for injunctive and
declaratory relief.

The en banc court held that the City of Oakland did not
sufficiently plead proximate cause for its reduced tax
revenue claim because its theory of harm went beyond the
first step of the causal chain, which was the harm to minority
buyers who received predatory loans. The en banc court
concluded that the Fair Housing Act is not a statute that
supports proximate cause for injuries further downstream,
and the extension of proximate cause beyond the first step
was not administratively possible and convenient. For the
same reasons, the City also failed sufficiently to plead
proximate cause for its increased municipal expenses claim.

The en banc court held that, in addition to claims for
damages, the proximate-cause requirement in Miami also
applies to injunctive and declaratory relief. It therefore
reversed the district court’s judgment to the contrary.

COUNSEL

Neal Kumar Katyal (argued), Colleen Roh Sinzdak,
Benjamin A. Field, and Sean Marotta, Hogan Lovells US
LLP, Washington, D.C.; Paul F. Hancock and Olivia
Kelman, K&L Gates LLP, Miami, Florida; Edward P.
Sangster and Daniel W. Fox, K&L Gates LLP, San
Francisco, California; Terry E. Sanchez, Munger Tolles &
Olson LLP, Los Angeles, California; Bart H. Williams and
Manuel F. Cachan, Proskauer Rose LLP, Los Angeles,
California; for Defendants-Appellants.
4 CITY OF OAKLAND V. WELLS FARGO & CO.

Robert S. Peck (argued), Center for Constitutional Litigation
P.C., Washington, D.C.; Barbara J. Parker, Oakland City
Attorney; Maria Bee, Chief Assistant City Attorney; Office
of the City Attorney, Oakland, California; Joel Liberson,
Trial & Appellate Resources P.C., Torrance, California;
Yosef Peretz and Ruth Israely, Peretz & Associates, San
Francisco, California; for Plaintiff-Appellee.

D. Scott Change, Housing Rights Center, Los Angeles,
California; Jamie Crook, American Civil Liberties Union
Foundation of Northern California, San Francisco,
California; David Loy, American Civil Liberties Union of
San Diego & Imperial Counties, San Diego, California; Julia
Devanthéry, American Civil Liberties Union of Southern
California, Los Angeles, California; Sandra S. Park and
Alejandro Ortiz, American Civil Liberties Union
Foundation, New York, New York; Morgan Williams,
National Fair Housing Alliance, Washington, D.C.; Ajmel
Quereshi, NAACP Legal Defense & Education Fund Inc.,
Washington, D.C.; for Amici Curiae American Civil
Liberties Union Foundation, American Civil Liberties Union
Foundation of Northern California, American Civil Liberties
Union Foundation of Southern California, American Civil
Liberties Union of San Diego & Imperial Counties, AARP,
NAACP Legal Defense & Educational Fund Inc., National
Fair Housing Alliance Inc., Poverty & Race Research Action
Council, and Twelve Local Fair Housing Centers in the
Ninth Circuit.

Dennis J. Herrera, City Attorney; Aileen M. McGrath, Co-
Chief of Appellate Litigation; City Attorney’s Office, San
Francisco, California; for Amicus Curiae City and County of
San Francisco.
 CITY OF OAKLAND V. WELLS FARGO & CO. 5

Michael L. Newman, Senior Assistant Attorney General;
Christine Chuang, Supervising Deputy Attorney General;
Shubhra Shivpuri and Srividya Panchalam; California
Department of Justice, Oakland, California; for Amicus
Curiae State of California.

Daniel P. Kearney Jr. and Matthew E. Vigeant, Wilmer
Cutler Pickering Hale & Dorr LLP, Washington, D.C.;
Steven P. Lehotsky and Emily J. Kennedy, U.S. Chamber
Litigation Center, Washington, D.C.; for Amicus Curiae
Chamber of Commerce of the United States of America.

Micha Star Liberty, Liberty Law, Oakland, California;
Marcus J. Jackson and David M. Arbogast, Jackson
Litigation, Carlsbad, California; for Amicus Curiae
California Black Chamber of Commerce.

William Michael Cunningham, Washington, D.C., pro se
Amicus Curiae.

OPINION

McKEOWN, Circuit Judge:

Only a few years ago, the Supreme Court addressed the
proximate-cause standard of the Fair Housing Act (“FHA”),
42 U.S.C. §§ 3601–3619, 3631, in Bank of America Corp. v.
City of Miami (“Miami”), 137 S. Ct. 1296 (2017).
Emphasizing that “foreseeability alone” is not sufficient to
establish proximate cause, the Court required “some direct
relation between the injury asserted and the injurious
conduct alleged.” Id. at 1305–06 (quoting Holmes v. Sec.
Inv. Prot. Corp., 503 U.S. 258, 268 (1992)). In
acknowledging that “[t]he housing market is interconnected
6 CITY OF OAKLAND V. WELLS FARGO & CO.

with economic and social life,” the Court observed that “[a]
violation of the FHA may, therefore, ‘be expected to cause
ripples of harm to flow’ far beyond the defendant’s
misconduct.” Id. at 1306 (quoting Associated Gen.
Contractors of Cal., Inc. v. Carpenters, 459 U.S. 519, 534
(1983)). Nonetheless, the Court limited the legal
consequences of those ripples: “Nothing in the statute
suggests that Congress intended to provide a remedy
wherever those ripples travel.” Id.

The City of Oakland (“Oakland”) claims that Wells
Fargo’s discriminatory lending practices caused higher
default rates, which in turn triggered higher foreclosure rates
that drove down the assessed value of properties, and which
ultimately resulted in lost property tax revenue and increased
municipal expenditures. These downstream “ripples of
harm” are too attenuated and travel too “far beyond” Wells
Fargo’s alleged misconduct to establish proximate cause. Id.
In this interlocutory appeal under 28 U.S.C. § 1292(b), we
therefore reverse the district court’s partial denial of Wells
Fargo’s motion to dismiss and remand for dismissal of the
FHA claims.

I. BACKGROUND

A. FACTUAL BACKGROUND

According to Oakland’s First Amended Complaint (the
“Complaint”), Wells Fargo violated the FHA by engaging in
mortgage-lending practices that discriminated against
African-American and Latino borrowers. Oakland alleges
that Wells Fargo had a longstanding “policy and practice of
steering minority borrowers” into mortgage loans with
“terms that have higher costs and risk features than more
favorable and less expensive loans for which the borrower
was eligible and which are regularly issued to similarly
 CITY OF OAKLAND V. WELLS FARGO & CO. 7

situated white borrowers.” Specifically, Oakland claims that
Wells Fargo’s practices resulted in giving a higher
proportion of riskier “adjustable rate loans to minority
borrowers than white borrowers” and giving “very few . . .
conventional 30-year fixed rate mortgages” to minority
borrowers.

According to Oakland, the discriminatory loans to
minority borrowers increased default and foreclosure rates
and decreased property values, which resulted in two
economic harms: a decrease in property tax revenue and the
simultaneous need for increased municipal expenditures to
address public health and safety issues. Oakland also alleges
that the discriminatory lending caused it non-economic
injury by undermining its racial-integration goals.

To support its allegations that the discriminatory lending
caused these harms, Oakland conducted a series of
regression analyses. As Oakland notes, a regression analysis
is a statistical method that examines “the relationship that
exists in a set of data between a variable to be explained—
called the ‘dependent variable’—and one or more
‘explanatory variables.’” Oakland “controll[ed] for
borrower race and objective risk characteristics,” to ensure
that borrowers being compared were similarly situated—that
is, that they “posses[ed] similar underwriting and borrower
characteristics.”

Based on Wells Fargo’s own lending history data,
Oakland found that, between 2004 and 2013, African-
American and Latino borrowers were 2.583 and 3.312 times
more likely, respectively, to receive loans with
discriminatory terms than similarly situated white
borrowers. Again controlling for “objective risk
characteristics,” Oakland found that the discriminatory loans
were 1.753 times more likely to result in foreclosure than
8 CITY OF OAKLAND V. WELLS FARGO & CO.

non-discriminatory loans. These differences, according to
Oakland, were statistically significant, meaning that there
was less than a one percent chance that the observed
differences would have occurred by chance. The Complaint
alleges that the risky, expensive loans led to foreclosure at
higher rates because “(1) the borrowers are required to make
higher loan payments; and (2) as foreclosures begin to occur
in a neighborhood, refinancing out of high-cost and high-risk
loans becomes increasingly difficult due to suppressed loan-
to-value ratios.”

The higher default and foreclosure rates then allegedly
decreased property values. The Complaint asserts that
“[h]omes in foreclosure tend to experience a substantial
decline in value,” which in turn reduces Oakland’s tax
revenue.

The foreclosures also allegedly required Oakland to
spend and divert resources to, among others, the police and
fire departments, the Oakland Building Services Division
and Code Enforcement, and the Oakland City Attorney’s
Office, to “remediate blighted conditions.”

B. PROCEDURAL BACKGROUND

Oakland sued Wells Fargo for damages as well as
declaratory and injunctive relief. While the case was
pending in the district court, the Supreme Court decided
Miami and clarified the requirements for proximate cause
under the FHA. 137 S. Ct. at 1305–06. The district court
accordingly instructed Oakland to amend its complaint in
light of Miami. Oakland did so, and Wells Fargo moved to
dismiss.

The district court dismissed Oakland’s claims as to
increased municipal expenditures but allowed Oakland’s
 CITY OF OAKLAND V. WELLS FARGO & CO. 9

claims as to decreased property tax revenue to proceed.
With respect to non-economic injuries, namely that
discriminatory lending practices undermined Oakland’s
racial-integration goals, the district court dismissed
Oakland’s claim on standing grounds. Finally, the court
allowed all claims for declaratory and injunctive relief to
proceed, reasoning that Miami’s directness requirement
“does not appear to extend” to these claims.

The district court certified two issues for interlocutory
appeal under 28 U.S.C. § 1292(b): (1) whether Oakland’s
claims for damages satisfy the FHA’s proximate-cause
requirement, and (2) whether that proximate-cause
requirement applies to claims for injunctive and declaratory
relief.

A panel of this court affirmed the district court’s
determination that Oakland sufficiently pleaded proximate
cause for the decreased property tax revenue claim; affirmed
the district court’s determination that Oakland failed to plead
proximate cause for the increased municipal expenditures
claim; and reversed the district court’s determination that
Miami’s proximate-cause requirement did not apply to
injunctive and declaratory relief. City of Oakland v. Wells
Fargo & Co., 972 F.3d 1112, 1137 (9th Cir. 2020), vacated,
993 F.3d 1077 (9th Cir. 2021). We voted to rehear the case
en banc. City of Oakland v. Wells Fargo & Co., 993 F.3d
1077 (9th Cir. 2021).

II. ANALYSIS

The FHA forbids “discriminat[ing] against any person in
the terms, conditions, or privileges of sale or rental of a
dwelling, or in the provision of services or facilities in
connection therewith, because of race.” 42 U.S.C.
§ 3604(b). Apropos of the lending practices at issue here,
10 CITY OF OAKLAND V. WELLS FARGO & CO.

the statute also makes it unlawful for “any person or other
entity whose business includes engaging in residential real
estate-related transactions to discriminate against any person
in making available such a transaction, or in the terms or
conditions of such a transaction, because of race.” Id.
§ 3605(a). Under the FHA, any “aggrieved person” may file
a civil action seeking damages for a violation of the statute.
Id. §§ 3613(a)(1)(A), (c)(1). An “aggrieved person” is
defined to include “any person who . . . claims to have been
injured by a discriminatory housing practice.” Id.
§ 3602(i)(1).

Benchmarking Oakland’s allegations against the
requirements of the FHA, we review de novo whether
Oakland adequately alleged proximate cause to survive a
motion to dismiss under Federal Rule of Civil Procedure
12(b)(6). See Pakootas v. Teck Cominco Metals, Ltd.,
830 F.3d 975, 980 (9th Cir. 2016). To answer this question,
we credit Oakland’s well-pleaded allegations as true and
look to the Court’s guidance in Miami, along with its related
proximate cause analysis in Lexmark International, Inc. v.
Static Control Components, Inc., 572 U.S. 118 (2014), and
Holmes v. Securities Investor Protection Corp., 503 U.S.
258 (1992), among other cases.

A. UNDER MIAMI, PROXIMATE CAUSE UNDER THE
FHA REQUIRES A DIRECT RELATION BETWEEN
WELLS FARGO’S CHALLENGED PRACTICES AND
OAKLAND’S ASSERTED INJURIES

In Miami, the Supreme Court considered allegations
almost identical to those made here. 137 S. Ct. at 1300–01.
The City of Miami alleged a series of predatory
discriminatory practices by Bank of America and Wells
Fargo, which resulted in higher default and foreclosure rates
for minority borrowers than for similarly situated white
 CITY OF OAKLAND V. WELLS FARGO & CO. 11

borrowers. Id. In turn, “[h]igher foreclosure rates lowered
property values and diminished property-tax revenue.” Id.
at 1301–02. The “[h]igher foreclosure rates—especially
when accompanied by vacancies—also increased demand
for municipal services” necessary “to remedy blight and
unsafe and dangerous conditions.” Id. at 1302.

The Eleventh Circuit held that Miami had adequately
pleaded proximate cause because the harm to the city was a
foreseeable result of the discriminatory lending. City of
Miami v. Bank of Am. Corp., 800 F.3d 1262, 1282 (11th Cir.
2015), vacated, 137 S. Ct. 1296 (2017). The Supreme Court
squarely rejected the Eleventh Circuit’s foreseeability
standard, explaining that “[i]n the context of the FHA,
foreseeability alone does not ensure the close connection that
proximate cause requires.” 137 S. Ct. at 1306. Instead, the
Court instructed that the proper standard was the more
stringent “direct relation” standard, which requires “some
direct relation between the injury asserted and the injurious
conduct alleged.” Id. (quoting Holmes, 503 U.S. at 268).
This direct-relation standard previously had been applied by
the Court to a number of common-law based statutes. See,
e.g., Lexmark, 572 U.S. at 132–40 (applying the direct-
relation standard to a claim brought under the Lanham Act);
Holmes, 503 U.S. at 268–74 (same, under the Racketeer
Influenced and Corrupt Organizations Act (“RICO”));
Associated Gen. Contractors of Cal., Inc., 459 U.S. at 540–
46 (same, under the Clayton Act). Citing to these cases,
Miami held that the FHA “is no exception,” because a
damages claim under the FHA is “akin to a ‘tort action.’”
137 S. Ct. at 1305 (quoting Meyer v. Holley, 537 U.S. 280,
285 (2003)).

In explaining the mechanics of the direct-relation
standard, the Court began by reaffirming “[t]he general
12 CITY OF OAKLAND V. WELLS FARGO & CO.

tendency . . . not to go beyond the first step” of the causal
chain. Id. at 1306 (internal quotation marks omitted)
(quoting Hemi Grp., LLC v. City of New York, 559 U.S. 1,
10 (2010)). The Court then noted, however, that “[w]hat
falls within that ‘first step’ depends in part on the ‘nature of
the statutory cause of action’ and an assessment ‘of what is
administratively possible and convenient.’” Id. (first
quoting Lexmark, 572 U.S. at 133; and then Holmes,
503 U.S. at 268).

Although the parties urged the Court to delineate
“precise boundaries” and to determine whether Miami’s
financial injuries met the direct-relation standard, the Court
declined to do so and left it to “the lower courts [to] define,
in the first instance, the contours of proximate cause under
the FHA.” Id. On remand, the Eleventh Circuit revisited the
case and issued an opinion. City of Miami v. Wells Fargo &
Co., 923 F.3d 1260 (11th Cir. 2019), vacated as moot, 140 S.
Ct. 1259 (2020). However, while a petition for writ of
certiorari was pending, Miami voluntarily dismissed its
lawsuit against Bank of America and other financial
institutions. See City of Miami v. Bank of Am. Corp., No.
13-cv-24506, slip op. at 1 (S.D. Fla. Jan. 30, 2020) (order
granting plaintiff’s unopposed motion for dismissal with
prejudice); City of Miami v. Wells Fargo & Co., No. 13-cv-
24508, slip op. at 1 (S.D. Fla. Jan. 30, 2020) (order granting
plaintiff’s unopposed motion for dismissal with prejudice).
The Court granted the petition and vacated the Eleventh
Circuit’s judgment as moot. Wells Fargo & Co. v. City of
Miami, 140 S. Ct. 1259 (2020). Although there is no circuit
precedent on the proximate-cause standard under the FHA,
the Supreme Court’s binding directives in Miami and its
earlier proximate-cause jurisprudence drive our analysis.
 CITY OF OAKLAND V. WELLS FARGO & CO. 13

B. OAKLAND DID NOT SUFFICIENTLY PLEAD
PROXIMATE CAUSE FOR ITS REDUCED TAX
REVENUE CLAIM

We begin where Miami began, with “[t]he general
tendency . . . not to go beyond the first step.” 137 S. Ct.
at 1306 (internal quotation marks and citation omitted).
There is no question that Oakland’s theory of harm goes
beyond the first step—the harm to minority borrowers who
receive predatory loans. Oakland’s theory of harm runs far
beyond that—to depressed housing values, and ultimately to
reduced tax revenue and increased municipal expenditures.
Oakland thus fails “a strict application of the ‘general
tendency’ not to stretch proximate causation ‘beyond the
first step.’” Lexmark, 572 U.S. at 139 (quoting Holmes,
503 U.S. at 271).

Oakland’s hope in this case—which turns out to be
misplaced—is that there is some basis not to “conform[] . . .
to the general tendency” not to go beyond the first step.
Holmes, 503 U.S. at 272. Indeed, there is some give in the
joints as to “[w]hat falls within that ‘first step.’” 137 S. Ct.
at 1306 (quoting Lexmark, 572 U.S. at 139). That is, there
are times when a proximate-cause analysis may
appropriately diverge from the general tendency.
Historically, the Court has framed this as going “beyond the
first step,” rather than expanding “what falls within” it. See,
e.g., Lexmark, 572 U.S. at 139 (“beyond the first step”
(citation omitted)); Hemi Grp., 559 U.S. at 10 (same);
Holmes, 503 U.S. at 271 (same). We see no meaningful
distinction between the historical framing of going “beyond”
the first step and Miami’s framing of expanding what is
considered to be “within” the first step. It is clear that both
of these exceptions fall outside the norm and hinge on a
statute-specific textual analysis of the “conduct the statute
14 CITY OF OAKLAND V. WELLS FARGO & CO.

prohibits.” Lexmark, 572 U.S. at 133. Whether conceived
as an expansive first step or an extension beyond the first
step, what matters doctrinally is that some direct relation is
required and that, under certain limited circumstances,
courts need not “conform[]” to the general first-step
tendency. Holmes, 503 U.S. at 271–72.

The Court in Miami articulated that these circumstances
rest in part on two considerations: “the ‘nature of the
statutory cause of action’ and an assessment ‘of what is
administratively possible and convenient.’” 137 S. Ct.
at 1306 (first quoting Lexmark, 572 U.S. at 133; and then
Holmes, 503 U.S. at 268). We thus turn to these
considerations.

1. The Nature of the Statutory Cause of Action

The nature of a particular statutory cause of action
implicates whether proximate cause can extend beyond the
first step because some statutes support proximate cause for
injuries further downstream. While the Supreme Court in
Miami did not directly answer whether the FHA is such a
statute, in our view the principles in Miami require us to
conclude that it is not.

To begin, we examine the “injurious conduct”
encompassed by the FHA. Id. (quoting Holmes, 503 U.S.
at 268). According to Miami, the FHA prohibits “lending to
minority borrowers on worse terms than equally
creditworthy nonminority borrowers and inducing defaults
by failing to extend refinancing and loan modifications to
minority borrowers on fair terms.” Id. at 1305. The harm
that the statute guards against—issuing discriminatory loans
that result in a default because of failure to refinance or
modify the loans on fair terms—is thus situated at the first
step: the issuance of the discriminatory loan. The harm to
 CITY OF OAKLAND V. WELLS FARGO & CO. 15

the borrower has a clear direct relation to conduct prohibited
by the FHA.

By contrast, the situations in which the Court has
countenanced a finding of proximate cause beyond the first
step arise from statutes that themselves encompass harm
beyond the first step. Two key cases illustrate this principle:
Bridge v. Phoenix Bond & Indemnity Co., 553 U.S. 639
(2008) and Lexmark, 572 U.S. 118.

In Bridge, the Court held that proximate cause extended
beyond the first step for a RICO claim predicated on mail
fraud. 553 U.S. at 661. The case arose from a county auction
of tax liens. Id. at 642–44. Whenever there was a tie
between the highest bidders, the county awarded the lien
based on a fair rotation of which bidder received the last lien
from a tie bid. Id. at 642–43. To game the system, bidders
began using multiple agents. Id. at 643. The county banned
that practice, but certain bidders allegedly continued using
multiple agents and filed false attestations of compliance.
Id. at 643–44. Upon learning of this scheme, competing
bidders sued, alleging that the fraud to the county harmed
their chances of being awarded liens. Id. The wrinkle from
a directness standpoint was that the first step was the false
attestation to the county. The offending bidders challenged
both standing and proximate cause on the ground that the
competing bidders had not relied on the false attestation. Id.
at 645–46. The harm to the competing bidders did not come
until later in the sequence. Id.

Nonetheless, based on the nature of the mail fraud
statute, the Court held that reliance on the misrepresentations
was not “a prerequisite to establishing proximate causation.”
Id. at 661. Key to the Court’s analysis was a parsing of the
statute and its conclusion that the statutory offense of mail
fraud “does not require proof of reliance.” Id. at 656. In this
16 CITY OF OAKLAND V. WELLS FARGO & CO.

way, the mail fraud statute permits proximate cause to
extend beyond the first step to reach the harmed party—the
competing bidder. That extension was particularly
necessary in Bridge, because the first-step party (the county)
suffered no injury. Id. at 658 (explaining that “respondents
and other losing bidders were the only parties injured by
petitioners’ misrepresentations”). Contrasting this situation
with Holmes and Anza v. Ideal Steel Supply Corp., 547 U.S.
451 (2006), the Court underscored that the competing
bidders’ “alleged injury—the loss of valuable liens—is the
direct result of [the] fraud” and “there are no independent
factors that account for [the] injury.” Bridge, 553 U.S.
at 658.

Just two years later, in Hemi Group, the Court took the
opportunity to highlight the principles from Bridge.
559 U.S. at 1. Under RICO, New York City sued the Hemi
Group, an online cigarette seller, for failure to file certain
reports of sales. Id. at 4. New York’s causal theory was that
without those reports, it was unable to go after direct
cigarette purchasers to collect tax revenue. Id. This theory
was rejected because the city could not show that the failure
to file reports “led directly to its injuries.” Id. at 14. The
city’s effort to rely on Bridge fell flat. Id. at 14–15. Unlike
the theory in Bridge, “the [c]ity’s theory . . . [was] anything
but straightforward: Multiple steps . . . separate[d] the
alleged fraud from the asserted injury.” Id. at 15. The
Court’s explanation that New York’s theory “rest[ed] on the
independent actions of third and even fourth parties,” id.,
echoes the attenuated, multi-step causal chain proffered by
Oakland.

Consistent with Bridge, the Court in Lexmark considered
another statute that permits proximate cause to extend
beyond the first step: the false-advertising provisions of the
 CITY OF OAKLAND V. WELLS FARGO & CO. 17

Lanham Act. 572 U.S 118. Lexmark, a manufacturer and
seller of laser printers and toner cartridges, dominated the
market for cartridges compatible with its printers. Id.
at 120–21. The question was whether Lexmark’s alleged
false advertising, which directly harmed certain printer
cartridge “remanufacturing” companies, also proximately
caused harm to Static Control, a company that made
microchips exclusively for the remanufacturing companies. 1
Id. at 132–40. Reasoning that proximate cause was satisfied,
the Court rejected the view that consumer deception is an
intervening step that breaks the proximate cause link. Id.
at 133, 140.

In coming to this conclusion, the Court examined the
common-law origins of the Lanham Act’s prohibition on
false advertising, which justified a more flexible approach.
Id. at 133. Typically, the first step in a false advertising
claim results from injuries “suffered by consumers who are
deceived by the advertising.” Id. But, as the Court put it,
because the Lanham Act “authorizes suit only for
commercial injuries,” and because deceived consumers do
not suffer commercial injuries, “the intervening step of
consumer deception is not fatal to the showing of proximate
causation required by the statute.” Id. (citation omitted).
Importantly, the Court did not jettison the directness
requirement; rather, it noted that “a plaintiff can be directly
injured by a misrepresentation even where ‘a third party, and
not the plaintiff, . . . relied on’ it.” Id. (quoting Bridge,
553 U.S. at 656). Even so, “the harm alleged” still must have

1
Remanufacturers “acquire used Lexmark toner cartridges,
refurbish them, and sell them in competition with new and refurbished
cartridges sold by Lexmark.” Id. at 121.
18 CITY OF OAKLAND V. WELLS FARGO & CO.

“a sufficiently close connection to the conduct the statute
prohibits.” Id.

Lexmark further held that proximate cause could be
extended beyond the direct competitor to another injured
party, in that case Static Control, because the harm flowed
automatically from the direct competitor remanufacturing
companies to Static Control. Id. at 139–40. “[I]f the
remanufacturers sold 10,000 fewer refurbished cartridges
because of Lexmark’s false advertising, then it would follow
more or less automatically that Static Control sold 10,000
fewer microchips for the same reason . . . .” Id. at 140
(emphasis added). Because there was “something very close
to a 1:1 relationship” between the number of cartridges sold
(or not sold) by the remanufacturer and the number of
microchips sold (or not sold) by the third-party chip
manufacturer, the intervening step did not cut off proximate
causation. Id. at 139. Emphasizing the “relatively unique
circumstances,” the Court held that the remanufacturers
were not “more immediate victim[s]” than Static Control.
Id. at 140 (quoting Bridge, 553 U.S. at 658).

Unlike the statutes in Bridge and Lexmark, the FHA
provides a direct link between the prohibited conduct and the
borrower but does not support stretching proximate cause
principles beyond the first step.

Still, Oakland urges that Miami’s broad interpretation of
the FHA for standing purposes is a reason to embrace a
capacious proximate-cause standard. Before the district
court, Oakland urged that “[i]t would be illogical for
Oakland to have standing under the FHA to pursue lost
property taxes and increased municipal expenses, but still be
unable to state a claim for those very same injuries under the
FHA’s causation standard.” The district court rejected that
 CITY OF OAKLAND V. WELLS FARGO & CO. 19

argument, pointing out that statutory “[s]tanding is a
separate issue from proximate cause.” We agree.

Indeed, it is critical to separate Miami’s holdings on two
distinct and independent questions—statutory standing and
proximate cause. The Court first considered whether, for
purposes of statutory standing, Miami was an “aggrieved
person” under the FHA. 137 S. Ct. at 1302–05. Giving that
term a broad reading, the Court concluded that Miami
alleged “economic injuries that arguably fall within the
FHA’s zone of interests.” Id. at 1305. But that conclusion
in no way controls the separate inquiry into proximate cause.
Put simply, “[p]roximate causation is not a requirement of
Article III standing,” and they are not coextensive. Lexmark,
572 U.S. at 134 n.6. And injury must have “a sufficiently
close connection to the conduct the statute prohibits”—not
simply any harm that Congress sought to target in enacting
the statute. Miami, 137 S. Ct. at 1305 (quoting Lexmark,
572 U.S. at 133).

2. Administrability

Having determined that the nature of the statute does not
warrant the extension of proximate cause beyond the first
step, we turn to Miami’s second consideration: “what is
administratively possible and convenient.” 137 S. Ct.
at 1306. In articulating this inquiry, Miami cited to Holmes,
where the Court laid out three factors relevant to
administrability: (1) the ability to distinguish the “damages
attributable to the violation, as distinct from other,
independent, factors”; (2) “the risk of multiple recoveries”;
and (3) whether more direct plaintiffs could “be counted on
to vindicate the law as private attorneys general.” Holmes,
503 U.S. at 269–70 (citation omitted).
20 CITY OF OAKLAND V. WELLS FARGO & CO.

Before addressing the Holmes factors, we pause to
clarify their role in the directness analysis. Holmes used
these factors to determine whether the direct-relation
standard applied to a particular statute, RICO. Id. We
recently used the factors in the same way when assessing
whether the direct-relation standard applied to the Anti-
Terrorism Act. Fields v. Twitter, Inc., 881 F.3d 739, 746
(9th Cir. 2018) (“The ATA presents precisely the risks with
which the Court was concerned in Holmes . . . .”). In
addition to using the factors to determine whether the direct-
relation standard applies, the Court—as well as our court—
have also used the factors in applying the direct-relation
standard. See Anza, 547 U.S. at 456–60 (invoking the factors
to “illustrate” an indirect injury); see also Hemi Grp.,
559 U.S. at 11–12; Bridge, 553 U.S. at 657–58; Canyon
County v. Syngenta Seeds, Inc., 519 F.3d 969, 982–83 (9th
Cir. 2008) (using the factors to support failure of proximate
cause). By citing to Holmes in its description of the
administrability component of the direct-relation standard,
Miami appears to endorse the use of the Holmes factors
within the application of the direct-relation standard. Miami,
137 S. Ct. at 1306. In light of this history, we view the
Holmes factors as instructive, though not mandatory.

Here, the Holmes factors reinforce our view that Oakland
has not met the directness requirement of the proximate-
cause standard. The first factor is the ability to distinguish
the “damages attributable to the violation, as distinct from
other, independent, factors.” Holmes, 503 U.S. at 269
(citation omitted). Oakland’s theory of harm fails this test.

To begin, Oakland does not allege that an increase in
foreclosures is “surely attributable” to the discriminatory
lending. Lexmark, 572 U.S. at 140. Oakland’s long and
winding causal chain begins with the claim that Wells Fargo
 CITY OF OAKLAND V. WELLS FARGO & CO. 21

initiated predatory loans to minority borrowers. Then, those
borrowers were more likely to default on the loans. To
trigger default, the borrower must quit making loan
payments or violate some other term of the loan, such as
maintaining mandatory insurance. The reason for default
could be attributable to many independent factors, such as
job loss, a medical hardship, a death in the family, a divorce,
a fire or other catastrophe, Covid-19, broader economic
trends, or any number of other unpredictable causes not
present when the loan was made. And once default occurs,
Oakland’s chain of events then requires the act of
foreclosure. According to the Complaint, Wells Fargo may
have “sold the loan or servicing rights to a third party,”
which presumably initiated the foreclosures. Even in the
face of default, whether to initiate foreclosure, renegotiate
the loan, sell the loan, or even let it ride, is a decision that
extends beyond Wells Fargo. (And even if Wells Fargo
retained the loan, the same foreclosure decisions would
inure.) The chain becomes even more attenuated when
variables of property value (which could turn not only on
foreclosure but neglect, criminal activity, changing
demographics, and macroeconomic trends) and reduced tax
revenues are piled on top of a cascading number of
independent variables. Thus, Oakland’s “theory of liability
rests not just on separate actions, but separate actions carried
out by separate parties,” in some cases third, fourth, or fifth
parties. Hemi Grp., 559 U.S. at 11.

The difficulties in attributing damages here are a far cry
from the situation in Lexmark, where the Court held that the
harm flowed so “automatically” that there were no concerns
about attributing damages. 572 U.S. at 140. Because
Oakland only alleges that the discriminatory loans make
foreclosure and decreased tax revenue more likely, there is
not a 1:1 relationship between the discriminatory loan—the
22 CITY OF OAKLAND V. WELLS FARGO & CO.

conduct forbidden by the FHA—and decreased tax revenue.
Oakland’s alleged injuries are more similar to those of an
unpaid “landlord” or “electric company” whose misfortune
stems from a third party’s “inability to meet [its] financial
obligations.” Id. at 134 (quoting Anza, 547 U.S. at 458).
Any assessment of the actual relationship would require the
“‘speculative . . . proceedings’ or ‘intricate, uncertain
inquiries’” that Lexmark cautioned against. Id. at 140
(quoting Anza, 547 U.S. at 459–60); see also Anza, 547 U.S.
at 460 (“The element of proximate causation recognized in
Holmes is meant to prevent these types of intricate, uncertain
inquiries from overrunning . . . litigation.”).

Oakland’s efforts to fill these gaps in the causal chain
through regression analyses fall short. By their own terms,
the regression analyses only “show whether the fact that a
loan had discriminatory terms made that loan more likely to
result in foreclosure.” They do not purport to show that
discriminatory loans automatically result in decreased
property values and then in decreased tax revenue. Thus,
even accepting the results of the regression analyses as true,
a court would be left with the unacceptable challenge of
isolating the “damages attributable to the violation, as
distinct from other, independent, factors.” Holmes, 503 U.S.
at 269 (citation omitted). Leaving aside whether statistical
modeling could be used to buttress causation in the
appropriate case—an issue we do not decide—Oakland’s
multiple but disconnected analyses here cannot be glued
together to satisfy the directness requirement.

The second Holmes factor is whether allowing proximate
cause to extend beyond the first step would require the court
to “adopt complicated rules apportioning damages among
plaintiffs removed at different levels of injury from the
violative acts, to obviate the risk of multiple recoveries.” Id.
 CITY OF OAKLAND V. WELLS FARGO & CO. 23

(citations omitted). That risk is not present here because
only Oakland (or a related administrative authority) could
recover lost property tax revenue. But while the presence of
this risk can indicate the need to rigorously adhere to the first
step analysis, nothing suggests that the absence of a risk of
duplicative recoveries warrants extending beyond the first
step. Anza, 547 U.S. at 459.

The third Holmes factor is whether directly injured
victims “can generally be counted on to vindicate the law as
private attorneys general.” Holmes, 503 U.S. at 269–70.
Here, the answer is yes. Directly harmed borrowers can sue
individually and are incentivized to do so through the
availability of punitive damages, attorneys’ fees, and
equitable relief. See 42 U.S.C. § 3613(c)(1)–(2) (describing
the relief which may be granted under the FHA). Harmed
borrowers can also sue as a class. See, e.g., Havens Realty
Corp. v. Coleman, 455 U.S. 363, 366–67 & n.3 (1982).
Organizations can sue under the FHA and do so. See, e.g.,
Tex. Dep’t of Hous. & Cmty. Affs. v. Inclusive Cmtys.
Project, 576 U.S. 519, 526 (2015); Havens, 455 U.S. at 367.
It also bears noting that the Department of Justice (“DOJ”)
can sue to enforce the FHA, 42 U.S.C. § 3614(a), and that
the Department of Housing and Urban Development can
refer certain FHA enforcement matters to the DOJ with a
recommendation to sue, 42 U.S.C. § 3610(c). According to
the Complaint, the DOJ in fact sued Wells Fargo, which paid
$175 million to resolve FHA fair lending claims based on
discrimination in residential mortgage lending. And, in a
broader suit brought by the DOJ, Wells Fargo paid
$1.2 billion for improper lending practices. In short,
Oakland does not stand in the shoes of a party that cannot
vindicate violations under the FHA, and nothing in this case
counsels broadening the universe of actionable harms. See
Anza, 547 U.S. at 460.
24 CITY OF OAKLAND V. WELLS FARGO & CO.

Having followed Miami to consider the nature of the
statute and what is administratively possible and convenient,
we conclude that Oakland’s claimed harm of reduced tax
revenue is too remote from the cause of action and that
nothing counsels going “beyond the first step” of proximate
causation.

C. OAKLAND DID NOT SUFFICIENTLY PLEAD
PROXIMATE CAUSE FOR ITS INCREASED
MUNICIPAL EXPENSES CLAIM

The district court dismissed Oakland’s claim stemming
from increased municipal expenditures. Although under
28 U.S.C. § 1292(b) the court certified Oakland’s “claims
for damages” in the plural, and without specificity as to
which theory, in briefing the parties focused almost
exclusively on the lost revenue claim. Nonetheless, because
of the broad scope of certification, like the panel, we address
this claim.

Despite Oakland’s opportunity to amend the Complaint,
its allegations are conclusory. The increased municipal
expenditure claim is similar to the tax revenue theory, except
that it introduces even more independent factors to the causal
chain. For example, the theory relies not only on the fact
that a home will be foreclosed upon and the other variables,
but also that individual actors will commit civil and criminal
violations, thus necessitating more city resources to avoid
and remedy those harms. This claim, which lacks even a
scintilla of directness between the FHA violation and the
purported harm, is founded on speculation based on
conjecture. Because this claim is even further afield from
the alleged wrongdoing than the reduced tax revenue claim,
it fails the proximate cause test for the same reasons.
 CITY OF OAKLAND V. WELLS FARGO & CO. 25

D. THE PROXIMATE-CAUSE REQUIREMENT IN MIAMI
APPLIES TO INJUNCTIVE AND DECLARATORY
RELIEF

The district court held that Oakland did not need to
satisfy the proximate-cause requirement for the injunctive
and declaratory relief claims. The district court erred in this
regard—a point that, to its credit, Oakland does not contest
on appeal. The Court in Miami held that proximate cause is
required under the FHA, and in doing so, did not distinguish
between claims for damages and those for declaratory and
injunctive relief. 137 S. Ct. at 1305–06. We read Miami to
require a showing of proximate cause for all claims arising
under the FHA. This conclusion is buttressed by the Court’s
holding in Lexmark that proximate cause “is an element of
the cause of action,” 572 U.S. at 134 n.6, that must be
established “in every case,” id. at 135. Critically, Lexmark
uniformly applied the proximate cause test without making
any distinction between the damages and injunctive relief
claims. Id. at 132–40. We reverse the district court’s
judgment to the contrary.

CONCLUSION

We affirm the district court’s dismissal of the damages
claim related to increased municipal expenditures and
reverse the district court’s denial of Wells Fargo’s motion to
dismiss the damages claim related to lost property tax
revenue and the claims for injunctive and declaratory relief.

AFFIRMED in PART, REVERSED in PART, and
REMANDED for dismissal of the FHA claims and
proceedings consistent with this opinion. Costs shall be
awarded to Wells Fargo.