FOR PUBLICATION

  UNITED STATES COURT OF APPEALS
       FOR THE NINTH CIRCUIT

BENJAMIN KOHN,                            No. 20-17316

             Plaintiff-Appellant,       D.C. No. 4:20-cv-
                                           04827-PJH
 v.

STATE BAR OF CALIFORNIA;                    OPINION
CALIFORNIA COMMITTEE OF
BAR EXAMINERS, and Their Agents
in Their Official Capacity,

             Defendants-Appellees.

      Appeal from the United States District Court
         for the Northern District of California
      Phyllis J. Hamilton, District Judge, Presiding

  Argued and Submitted En Banc September 20, 2023
              San Francisco, California

                Filed December 6, 2023

 Before: Mary H. Murguia, Chief Judge, and Johnnie B.
 Rawlinson, Sandra S. Ikuta, John B. Owens, Daniel A.
 Bress, Danielle J. Forrest, Patrick J. Bumatay, Jennifer
Sung, Gabriel P. Sanchez, Holly A. Thomas and Salvador
              Mendoza, Jr., Circuit Judges.
2                KOHN V. STATE BAR OF CALIFORNIA


                     Opinion by Judge Owens;
             Partial Concurrence by Judge Mendoza;
    Partial Concurrence and Partial Dissent by Judge Bumatay


                          SUMMARY *


               Eleventh Amendment Immunity

    The en banc court (1) affirmed in part the district court’s
dismissal of attorney Benjamin Kohn’s action against the
State Bar of California and the California Committee of Bar
Examiners under Title II of the Americans with Disabilities
Act, Section 504 of the Rehabilitation Act, and California
law; and (2) remanded to the original three-judge panel for
consideration of the remaining issues.
    In the State Bar’s role in the admission of attorneys, it
acts under the authority and at the direction of the California
Supreme Court. Kohn sought monetary damages and other
relief based on the State Bar’s refusal to provide him with
certain test-taking accommodations for the bar exam. The
district court dismissed the action on the basis of Eleventh
Amendment immunity.
    The en banc court reaffirmed that the California State
Bar enjoys Eleventh Amendment immunity from suit in
federal court. The en banc court held that Eleventh
Amendment immunity extends not only to suits in which a
state itself is a named party, but also to suits against an “arm


*
 This summary constitutes no part of the opinion of the court. It has
been prepared by court staff for the convenience of the reader.
               KOHN V. STATE BAR OF CALIFORNIA                 3


of the state.” The Ninth Circuit’s version of the test for
determining whether an entity is an arm of the state applied
the so-called Mitchell factors. The en banc court concluded
that the Mitchell factors test should be reshaped in light of
developments in Supreme Court doctrine and the Ninth
Circuit’s      experience      applying       the     Mitchell
factors. Accordingly, the en banc court adopted the D.C.
Circuit’s three-factor test, which considers: (1) the state’s
intent as to the status of the entity, including the functions
performed by the entity; (2) the state’s control over the
entity; and (3) the entity’s overall effects on the state
treasury.
    Applying this updated three-factor test, the en banc court
held that the California State Bar is an arm of the state and
entitled to sovereign immunity. The en banc court
concluded that the first factor, California’s intent as to the
State Bar, strongly favored the conclusion that it is an arm of
the state, as did the second factor, the state’s control over the
State Bar. The en banc court concluded that the third factor,
the State Bar’s effects on the state treasury, presented a
closer call but was not dispositive.
   Concurring in part, Judge Mendoza agreed with the
majority that the Mitchell factors were out of step with the
Supreme Court’s jurisprudence and that the California State
Bar is an arm of the state for sovereign immunity
purposes. He wrote separately to caution against adopting
the D.C. Circuit’s approach to weighing the sovereign
immunity factors, and he disagreed with the majority’s
wholesale embrace of the D.C. Circuit’s entity-based
approach to sovereign immunity.
  Concurring in part and dissenting in part, Judge
Bumatay, joined by Judge Sung, wrote that he agreed with
4             KOHN V. STATE BAR OF CALIFORNIA


the majority’s abandonment of the Mitchell factors in favor
of the D.C. Circuit’s more streamlined approach, looking at
intent, control, and overall effects on a state’s treasury to
determine whether an entity is an arm of the state. Judge
Bumatay, however, disagreed with the majority’s
application of this new approach, and he would hold that
each of its factors cuts against finding sovereign immunity
for the California State Bar.



                        COUNSEL

Gregory R. Michael (argued) and Dorothy C. Yamamoto,
Michael Yamamoto LLP, Berkeley, California, for Plaintiff-
Appellant.
Brady R. Dewar (argued), Ellin Davtyan, Robert G. Retana,
Rita K. Himes, and Jean R. Krasilnikoff, Office of General
Counsel, The State Bar of California, San Francisco,
California; for Defendants-Appellees.
Julian Sarkar, SarkarLaw, San Francisco, California, for
Amicus Curiae SarkarLaw.
Claudia Center, Disability Rights Education and Defense
Fund, Berkeley, California; Jinny Kim, Disability Rights
Advocates, Berkeley, California; Laura A. Scalia, Legal Aid
At Work, San Francisco, California; for Amici Curiae
Disability Rights Education and Defense Fund, Inc., Legal
Aid Work, et. al.
               KOHN V. STATE BAR OF CALIFORNIA               5

OPINION

OWENS, Circuit Judge:

    For nearly forty years, the California State Bar has
enjoyed Eleventh Amendment immunity in federal court.
See, e.g., Lupert v. Cal. State Bar, 

, 1327 (9th
Cir. 1985); Hirsh v. Justs. of the Sup. Ct. of Cal., 

, 715 (9th Cir. 1995) (per curiam). Appellant Benjamin
Kohn, a licensed California attorney, seeks to change that.
He contends that the State Bar is not an “arm of the state,”
and he can sue it without restriction. Consistent with every
other circuit, we reaffirm that the State Bar enjoys Eleventh
Amendment protection in federal court and update our arm
of the state jurisprudence to better reflect the Supreme
Court’s most recent guidance.
 I. FACTUAL AND PROCEDURAL BACKGROUND
    The California State Bar is the “administrative arm” of
the California Supreme Court “for the purpose of assisting
in matters of admission and discipline of attorneys.” In re
Rose, 

 (Cal. 2000) (quoting In re Att’y
Discipline Sys., 

 (Cal. 1998)); see also Cal.
R. Ct. 9.3 (“The State Bar serves as the administrative arm
of the Supreme Court for admissions matters.”). Under the
California Constitution, “[e]very person admitted and
licensed to practice law in [the] [s]tate is and shall be a
member of the State Bar . . . .” Cal. Const. art. VI, § 9. The
State Bar “acts under the authority and at the direction of the
Supreme Court[,]” which has “inherent jurisdiction over the
practice of law” in the state. Cal. R. Ct. 9.3. As part of its
role in the admission of attorneys, the State Bar examines
candidates’ qualifications, administers the bar exam, and
6                 KOHN V. STATE BAR OF CALIFORNIA


certifies candidates to the California Supreme Court. Id.;

, 6060(g).
    The claims in this case stem from the State Bar’s
admission function. Kohn filed a federal complaint against
the State Bar seeking monetary damages and other relief. He
alleged that its refusal to provide him with certain test-taking
accommodations violated Title II of the Americans with
Disabilities Act, Section 504 of the Rehabilitation Act,
sections of the California Government Code, and
California’s Unruh Civil Rights Act. 1 The State Bar moved
to dismiss the lawsuit on several grounds, including that the
Eleventh Amendment prohibited the action from going
forward. 2
    The district court agreed with the State Bar. It granted
the motion to dismiss and quoted Hirsh’s clear holding for
support: “The Eleventh Amendment’s grant of sovereign
immunity bars monetary relief from state agencies such as
California’s Bar Association and Bar Court.” Hirsh, 67 F.3d
at 715. Hirsh relied exclusively on Lupert for its holding.
Id; see also Lupert, 761 F.2d at 1327 (“The Eleventh


1
 

 et seq. (Title II of the Americans with Disabilities
Act); 

 (Section 504 of the Rehabilitation Act); Cal. Gov’t
Code §§ 11135 et seq., 12944 et seq.; 

(f) (Unruh
Civil Rights Act).
2
  In this opinion, we reach only the issue of whether the State Bar is an
arm of the state for purposes of sovereign immunity. We remand the
case to the original three-judge panel for consideration of the remaining
issues. See United States v. Dreyer, 

, 1270 n.1 (9th Cir.
2015) (en banc) (“If the Court votes to hear or rehear a case en banc, the
en banc court may, in its discretion, choose to limit the issues it
considers.” (quoting Kyocera Corp. v. Prudential-Bache Trade Servs.,
Inc., 

, 995 (9th Cir. 2003) (en banc))).
                 KOHN V. STATE BAR OF CALIFORNIA                       7


Amendment bars this suit against the named agencies as the
state did not consent to being sued.”).
    Normally that would be the end of the story. A nearly
forty-year-old precedent that largely has gone unchallenged 3
would control the panel’s decision, and en bancs are quite
rare. But this story is only getting started.
    Lupert and Hirsh were largely silent as to why the State
Bar enjoyed Eleventh Amendment immunity, and Hirsh
ignored a long line of caselaw setting out our test (often
called the Mitchell factors) for determining whether a state
agency, like the State Bar, is an arm of the state entitled to
such protection. See, e.g., Jackson v. Hayakawa, 

, 1349–50 (9th Cir. 1982); Mitchell v. L.A. Cmty. Coll.
Dist., 

, 201–02 (9th Cir. 1988). We have
applied the Mitchell factors and the Eleventh Amendment to


3
 See, e.g., Viriyapanthu v. State Bar of Cal., 

, 313 (9th
Cir. 2020) (The “State Bar of California . . . [is] entitled to sovereign
immunity.”); Vartanian v. State Bar of Cal., 

, 600 (9th
Cir. 2019) (same); Kinney v. State Bar of Cal., 

(9th Cir. 2017) (same); Haroonian v. Comm. of Bar Exam’rs, 

 (9th Cir. 2017) (same); Kinney v. State Bar of Cal., 

 (9th Cir. 2017) (same); Tanasescu v. State Bar of Cal.,

 (9th Cir. 2014) (same); Joseph v. State Bar of
Cal., 

 (9th Cir. 2014) (same); Khanna v. State Bar
of Cal., 

, 177 (9th Cir. 2009) (same); Torres v. State
Bar of Cal., 

, 14–15 (9th Cir. 2005) (same); Taggart v.
State Bar of Cal., 

 (9th Cir. 2003) (same).
  We generally have taken the same approach with respect to other state
bars. See, e.g., O’Connor v. Nevada, 

 (9th Cir. 1982)
(State Bar of Nevada); Strojnik v. State Bar of Ariz., 

,
776 (9th Cir. 2020); Block v. Wash. State Bar Ass’n, 

,
731 (9th Cir. 2019). But see Crowe v. Or. State Bar, 

(9th Cir. 2021) (per curiam); infra pp. 32–33.
8                KOHN V. STATE BAR OF CALIFORNIA


a wide range of state entities. 4 Yet we have spent little time
over these decades considering whether our law accurately
captures the latest Supreme Court thinking.
    We sua sponte took this case en banc to decide whether
(1) the Mitchell factors, described infra pp. 11–17, remain
the optimal means to conduct an arm of the state analysis;
and (2) the California State Bar enjoys Eleventh Amendment
protection under a more rigorous scrutiny than it received in
Lupert and Hirsh.




4
  See, e.g., Ray v. County of Los Angeles, 

, 709–11 (9th Cir.
2019) (county); Sato v. Orange Cnty. Dep’t of Educ., 

, 928–
34 (9th Cir. 2017) (school districts and county offices of education);
Beentjes v. Placer Cnty. Air Pollution Control Dist., 

, 778–
86 (9th Cir. 2005) (air pollution control district); Aguon v.
Commonwealth Ports Auth., 

, 901–04 (9th Cir. 2003)
(public corporation created to manage ports); In re Lazar, 

,
982–84 (9th Cir. 2001) (reimbursement program run by state water-
resources control board); Streit v. County of Los Angeles, 

,
566–67 (9th Cir. 2001) (county sheriff’s department); Hale v. Arizona,

 (9th Cir. 1993) (en banc) (operator of prison labor
program), abrogated on other grounds as recognized by Nwauzor v.
GEO Grp., Inc., 

, 514–15 (9th Cir. 2023); Alaska Cargo
Transp., Inc. v. Alaska R.R. Corp., 

, 380–82 (9th Cir. 1993)
(state-created railroad corporation); ITSI T.V. Prods., Inc. v. Agric.
Ass’ns, 

, 1292–93 (9th Cir. 1993) (state fair and exposition);
Brooks v. Sulphur Springs Valley Elec. Coop., 

 (9th
Cir. 1991) (public utilities commission and nonprofit public service
corporation); Durning v. Citibank, N.A., 

, 1423–28 (9th
Cir. 1991) (community development authority); Austin v. State Indus.
Ins. Sys., 

, 678–79 (9th Cir. 1991) (state-run workers’
compensation program).
               KOHN V. STATE BAR OF CALIFORNIA               9


  II. DISCUSSION
     A. Standard of Review
   “We review de novo a dismissal on the basis of sovereign
immunity or for failure to state a claim upon which relief can
be granted.” Ariz. Students’ Ass’n v. Ariz. Bd. of Regents,

 (9th Cir. 2016). Whether an entity is an
arm of the state within the meaning of the Eleventh
Amendment is a question of federal law. Regents of the
Univ. of Cal. v. Doe, 

, 429 n.5 (1997)
(“Regents”).
     B. The Arm of the State Doctrine and the Mitchell
        Factors
    The Eleventh Amendment provides that “[t]he Judicial
power of the United States shall not be construed to extend
to any suit in law or equity, commenced or prosecuted
against one of the United States by Citizens of another State,
or by Citizens or Subjects of any Foreign State.” U.S. Const.
amend. XI. Longstanding Supreme Court precedent has
interpreted this Amendment to immunize states from suit in
federal court by citizens and noncitizens alike. See, e.g.,
Seminole Tribe of Fla. v. Florida, 

 (1996)
(“Seminole Tribe”); Blatchford v. Native Vill. of Noatak, 

 (1991); Hans v. Louisiana, 

(1890). This immunity extends not just to suits in which the
state itself is a named party but also to those against an “arm
of the [s]tate.” Mt. Healthy City Sch. Dist. Bd. of Educ. v.
Doyle, 

 (1977); accord Hess v. Port Auth.
Trans-Hudson Corp., 

 (1994); Regents, 

.
    There is no standard test for determining whether an
entity is an arm of the state for purposes of sovereign
10                KOHN V. STATE BAR OF CALIFORNIA


immunity. See Fresenius Med. Care Cardiovascular Res.,
Inc. v. P.R. & Caribbean Cardiovascular Ctr. Corp., 

 (1st Cir. 2003) (“The arm of the state analytical
doctrine has moved freely . . . applying common
principles.”).   The circuits have developed different
approaches to this question based on considerations the
Supreme Court has identified as relevant, including “the
nature of the entity created by state law,” Mt. Healthy City
Sch. Dist. Bd. of Educ., 

, whether the state
“structured” the entity to “enjoy the special constitutional
protection of the [s]tate[] [itself],” Hess, 513 U.S. at 43–44
(citation omitted), and the state’s legal liability for
judgments against the entity, Lake Country Ests., Inc. v.
Tahoe Reg’l Plan. Agency, 

, 401–02 (1979). 5

5
  See, e.g., Waskul v. Washtenaw Cnty. Cmty. Mental Health, 

 (6th Cir. 2020) (“(1) the [s]tate’s potential liability for a
judgment against the entity; (2) the language by which state statutes and
state courts refer to the entity and the degree of state control and veto
power over the entity’s actions; (3) whether state or local officials
appoint the board members of the entity; and (4) whether the entity’s
functions fall within the traditional purview of state or local government”
(citation omitted)); Leitner v. Westchester Cmty. Coll., 

(2d Cir. 2015) (alternating between (1) a six-factor test evaluating
entity’s structure and treatment under state law and (2) a two-factor test
considering extent of state responsibility for judgment and state
supervision of the entity); Walker v. Jefferson Cnty. Bd. of Educ., 

 (11th Cir. 2014) (“(1) how the state law defines the entity;
(2) the degree of state control over the entity; and (3) the entity’s fiscal
autonomy” (citation omitted)); P.R. Ports Auth. v. Fed. Mar. Comm’n,

 (D.C. Cir. 2008) (analyzing state intent, state control,
and effects on state treasury); Bowers v. Nat’l Coll. Athletic Ass’n, 

 (3d Cir. 2007) (“(1) whether the payment of the judgment
would come from the state; (2) what status the entity has under state law;
and (3) what degree of autonomy the entity has”); Thomas v. St. Louis
Bd. of Police Comm’rs, 

 (8th Cir. 2006) (looking to
effect of judgment on state treasury and to degree of entity’s autonomy
                  KOHN V. STATE BAR OF CALIFORNIA                         11


The Supreme Court has directed that “[w]hen indicators of
immunity point in different directions, the Eleventh
Amendment’s twin reasons for being remain our prime
guide”: the states’ dignity and their financial solvency.
Hess, 

.
    Our version of the arm of the state test, the so-called
Mitchell factors, arose from a grab bag of Supreme Court
and Ninth Circuit precedent and is normally reduced to the
following:

         [1] whether a money judgment would be
         satisfied out of state funds, [2] whether the
         entity performs central governmental
         functions, [3] whether the entity may sue or
         be sued, [4] whether the entity has the power


and control over its own affairs); Takle v. Univ. of Wis. Hosp. & Clinics
Auth., 

, 769–71 (7th Cir. 2005) (considering effect of
judgment on state treasury, nature of entity’s function, and treatment
under state law); Md. Stadium Auth. v. Ellerbe Becket Inc., 

,
260–61 (4th Cir. 2005) (considering effect on state treasury; entity’s
independence from state; entity’s involvement in statewide concerns;
and state-law treatment of entity); Fresenius Med. Care Cardiovascular
Res., Inc., 

 (asking (1) whether state has “clearly structured
the entity to share its sovereignty” and, if the first stage is inconclusive,
(2) whether “damages will be paid from the public treasury”); Sw. Bell
Tel. Co. v. City of El Paso, 

 (5th Cir. 2001) (“the
relationship between the state and the entity . . . the essential nature of
the proceeding, the nature of the entity created by state law, and whether
a money judgment against the instrumentality would be enforceable
against the state”); Duke v. Grady Mun. Schs., 

 (10th
Cir. 1997) (“degree of autonomy . . . as determined by the
characterization of the agency by state law and the extent of guidance
and control exercised by the state” and “extent of financing the agency
receives independent of the state treasury and its ability to provide for its
own financing” (citation omitted)).
12             KOHN V. STATE BAR OF CALIFORNIA


       to take property in its own name or only the
       name of the state, and [5] the corporate status
       of the entity.

Belanger v. Madera Unified Sch. Dist., 

, 250–
51 (9th Cir. 1992) (quoting Mitchell, 

).
    This case presents the question of whether we ought to
reshape the Mitchell factors in light of developments in
Supreme Court doctrine and our experience applying them.
We conclude that we should.
    First, the Mitchell factors are out of step with current
Supreme Court jurisprudence. Under Mitchell, we have
placed the greatest weight on the first factor—whether a
money judgment would be satisfied out of state funds. See,
e.g., Durning v. Citibank, N.A., 

 (9th
Cir. 1991) (“[T]he source from which the sums sought by the
plaintiff must come is the most important single factor in
determining whether the Eleventh Amendment bars federal
jurisdiction.” (citations omitted)). Our decision to prioritize
the first factor was a “recognition of Edelman [v. Jordan],”
which held that the Eleventh Amendment bars suits that seek
to impose liability that “would have to be satisfied out of
public funds from the state treasury.” 

 (citing Edelman v.
Jordan, 

 (1974)).
    But, since Edelman and Mitchell, the Supreme Court has
clarified that “[t]he Eleventh Amendment does not exist
solely in order to ‘preven[t] federal-court judgments that
must be paid out of a [s]tate’s treasury.’” Seminole Tribe,

 (second alteration in original) (quoting Hess,

). “[I]t also serves to avoid ‘the indignity of
subjecting a [s]tate to the coercive process of judicial
tribunals at the instance of private parties.’” 

 (quoting
                  KOHN V. STATE BAR OF CALIFORNIA                       13


P.R. Aqueduct & Sewer Auth. v. Metcalf & Eddy, Inc., 

 (1993)).
    Consequently, the inquiry into whether a state is legally
liable for judgments against an entity is important not as “a
formalistic question of ultimate financial liability,” but
because it is “an indicator of the relationship between the
[s]tate and its creation.” Regents, 

. Taking
heed of this doctrinal development, our sister circuits have
moved away from an excessive emphasis on the treasury
factor. 6 We, however, have never considered what the
Supreme Court’s more recent cases require, instead
maintaining a primary focus on the treasury factor.
Consequently, we have underemphasized the dignity
interests of the states, one of the Eleventh Amendment’s
“twin reasons for being.” Hess, 

.
    Kohn asserts that we should continue to prioritize the
treasury factor. He anchors his argument in Hess’s statement
that the “vast majority of Circuits . . . have concluded that
the state treasury factor is the most important.” 

 (ellipsis in original) (citation omitted). But he ignores
that Hess attached the same level of importance to state
dignity. See 

. Hess involved the potential
immunity of a bistate entity created under the Compact

6
  See, e.g., Bowers, 

 (“[W]e can no longer ascribe primacy
to the first factor” of “whether payment comes from the state treasury.”
(citation omitted)); Fresenius Med. Care Cardiovascular Res., Inc., 322
F.3d at 65–66 (holding that, “[i]n the aftermath of Hess,” “potential
payment from the state treasury is the most critical factor” only if “there
is an ambiguity about the direction in which the structural analysis
points” (emphasis added)); cf. Duke, 

 (“[E]ven after Hess
and [Regents], which emphasized the primacy of the impact on the state
treasury as a factor in determining immunity, other factors remain
relevant.”).
14             KOHN V. STATE BAR OF CALIFORNIA


Clause. Id. at 35. The Supreme Court recognized that,
because bistate entities “occupy a significantly different
position in our federal system than do the [s]tates
themselves,” “[s]uit in federal court is not an affront to the
dignity of a Compact Clause entity” or to “the integrity of
the compacting States.” Id. at 40–41.
    But this acknowledgment—that a suit against a bistate
entity does not threaten its parent state’s dignity interest in
the same way that a suit against that state itself would—does
not mean that the state’s dignity interest is less important in
determining whether a suit against an entity is a suit against
the state itself. Therefore, we read Hess for what it says: that
state dignity and solvency are the Eleventh Amendment’s
“twin reasons for being” and entitled to equal weight. Id. at
47, 52; see also P.R. Ports Auth. v. Fed. Mar. Comm’n, 

 (D.C. Cir. 2008) (“Hess does not require a
focus solely on the financial impact of the entity on the
[s]tate. Rather, Hess ‘pays considerable deference to the
dignity interests of the state, focusing on both explicit and
implicit indications that the state sought to cloak an entity in
its Eleventh Amendment immunity.’” (quoting Fresenius
Med. Care Cardiovascular Res., Inc., 

)). As
a result, our continued elevation of state solvency under the
Mitchell factors conflicts with the Supreme Court’s
guidance.
    The Mitchell factors are not only inconsistent with
Supreme Court arm of the state doctrine—they also generate
a muddled arm of the state analysis within our Circuit. For
example, some of the Mitchell factors are of questionable
relevance. Consider the third factor, “whether the entity may
sue or be sued.” Belanger, 

. Under
California law, a variety of state entities, including the State
Bar, 

, may “sue and be sued.”
                  KOHN V. STATE BAR OF CALIFORNIA                      15


See, e.g., Belanger, 

 (school district);
Beentjes v. Placer Cnty. Air Pollution Control Dist., 

 (9th Cir. 2005) (air pollution control district);
Sato v. Orange Cnty. Dep’t of Educ., 

 (9th
Cir. 2017) (county office of education). But this provision
has limited relevance for purposes of federal immunity.
    The Supreme Court has explicitly held that a state does
not “consent to suit in federal court merely by stating its
intention to ‘sue and be sued . . . .’” Coll. Sav. Bank v. Fla.
Prepaid Postsecondary Educ. Expense Bd., 

 (1999) (citing Fla. Dep’t of Health & Rehab. Servs. v.
Fla. Nursing Home Ass’n, 

, 149–50 (1981) (per
curiam)). 7 Thus, “[a] mere statutory grant of the power to
sue or be sued . . . is not enough to waive immunity from
suits brought in federal court if it may fairly be construed as
limited to a waiver of immunity in the state’s own courts.”
Durning, 

 n.4 (citing Welch v. Tex. Dep’t of
Highways & Pub. Transp., 

, 473–74 (1987)).
As a result, we have said that the sue or be sued factor “is
entitled to less weight.” Belanger, 

(“California school districts can sue and be sued in their own
name,” but “[i]f a school district is a state agency for
purposes of the Eleventh Amendment, suits against the



7
  See also Atascadero State Hosp. v. Scanlon, 

 (1985)
(declining to find waiver “[i]n the absence of an unequivocal waiver
specifically applicable to federal-court jurisdiction”), superseded by
statute as recognized by United States ex rel. Cain v. Salish Kootenai
Coll., Inc., 

 (9th Cir. 2017); cf. Biden v. Nebraska, 

 (2023) (“Every government corporation . . . is a
corporation, after all, with the power[] . . . to sue and be sued,” but it
“nonetheless remains (for many purposes at least) part of the
[g]overnment itself.” (internal citations and quotations omitted)).
16               KOHN V. STATE BAR OF CALIFORNIA


district in its own name are subject to the same Eleventh
Amendment constraints as suits against the state.”).
    Similar problems arise under factor four—whether the
entity has the power to take property in its own name or only
the name of the state. Even where an entity can hold
property in its own name and thus satisfies factor four, we
have said “the property ownership analysis is a close
question” if “[state] law . . . treats such property as state
property.” 

 As a result, this factor is also “entitled to little
weight.” 

 Therefore, at least two of the Mitchell factors
do not do much, if any, work. The caselaw bears this out:
While factors three and four are almost invariably satisfied,
they do not have a predictable effect on the outcome of any
individual case. 8
    We also have wavered as to whether we evaluate the
second Mitchell factor—whether the entity performs central
governmental functions—at the entity or activity level. If
the Mitchell analysis is entity based, then an entity is either
immune or not. But, if the Mitchell analysis is activity based,
then an entity’s immunity from suit may vary depending on
the function it performs. In applying Mitchell, we have said
both that we cannot “hold that [an entity] is immune from
suit with respect to some of its activities . . . but not others”

8
  Compare Crowe, 

 (entity could sue and be sued and
take property in its own name and was not an arm of the state); Ray, 

 (same); Beentjes, 397 F.3d at 784–86 (same); Holz v. Nenana
City Pub. Sch. Dist., 

, 1187–89 (9th Cir. 2003) (same);
Savage v. Glendale Union High Sch., 

, 1049–51 (9th Cir.
2003) (same), and Durning, 950 F.2d at 1427–28 (same), with Sato, 861
F.3d at 933–34 (entity could sue and be sued and take property in its own
name but was an arm of the state); Aguon, 

 (same);
Belanger, 

 (same), and Alaska Cargo Transp., Inc., 5
F.3d at 381–82 (same).
                KOHN V. STATE BAR OF CALIFORNIA                17


because “[t]o do so would impermissibly qualify sovereign
immunity, which by its nature is absolute,” Durning, 

, and that “we look to whether the [entity], in
performing the particular function at issue, performs a
central government function,” rather than “whether the
[entity] performs central government functions in general,”
Ray, 

. Scholarly criticism has focused on this
inconsistency and argued that it allows lower courts in our
Circuit to “twist” the arm of the state doctrine depending on
the defendant. See Kelsey Joyce Dayton, Tangled Arms:
Modernizing and Unifying the Arm-of-the-State Doctrine, 

, 1633 (2019) (arguing that lower courts
in the Ninth Circuit “brush aside aspects of” the Mitchell
factors “in cases that prove especially troublesome”).
    Notably, we did not always apply the Mitchell factors
mechanically as a five-part test. For example, in Franceschi
v. Schwartz, 

 (9th Cir. 1995) (per curiam), we
cited Mitchell but analyzed only how “state law treats the
entity . . . in an effort to assess the extent to which the entity
‘derives its power from the [s]tate and is ultimately regulated
by the [s]tate.’” 

 (citations omitted). Thus, our
determination that a municipal court was an arm of the state
turned on “the extensive control exercised by the state over
the municipal courts.” 

 Likewise, in Rounds v. Or. State
Bd. of Higher Educ., 

 (9th Cir. 1999), we cited
Mitchell solely for the proposition that we must look to an
entity’s “nature as created by state law” and “whether [it]
performs central governmental functions” to conduct the
arm of the state analysis. Id. at 1035.
    These cases suggest an earlier recognition that the best
arm of the state test is not a multi-factor checklist involving
potentially irrelevant factors but an analysis that drills down
on whether the state “structured” the entity to enjoy
18              KOHN V. STATE BAR OF CALIFORNIA


immunity from suit. Hess, 

 (citation omitted).
The D.C. Circuit’s test fits the bill. In an opinion by then-
Judge Kavanaugh, the D.C. Circuit distilled the
developments in the Supreme Court’s more recent caselaw
into a three-factor test: “(1) the [s]tate’s intent as to the status
of the entity, including the functions performed by the entity;
(2) the [s]tate’s control over the entity; and (3) the entity’s
overall effects on the state treasury.” P.R. Ports Auth., 

 (citing Hess, 513 U.S. at 44–46; Lake Country
Ests., Inc., 440 U.S. at 401–02; Mt. Healthy City Sch. Dist.
Bd. of Educ., 429 U.S. at 280–81).
    The first factor of intent turns on whether state law
expressly characterizes the entity as a governmental
instrumentality rather than as a local governmental or non-
governmental entity; whether the entity performs state
governmental functions; whether the entity is treated as a
governmental instrumentality for purposes of other state
law; and state representations about the entity’s status. 

. The second factor depends on how members of the
governing body of the entity are appointed and removed, as
well as whether the state can “directly supervise and control
[the entity’s] ongoing operations.” 

. And, the third
factor, though relevant, is not dispositive. While Kohn
argues that this factor is the most important, we agree with
the D.C. Circuit that the Eleventh Amendment “does not
require a focus solely on the financial impact of the entity on
the [s]tate” because the Eleventh Amendment is equally
concerned with “the dignity interests of the state.” 

(citation omitted) (interpreting Hess).
    We have not updated the Mitchell factors since we first
articulated them in 1982 in Jackson, despite the Supreme
Court’s intervening decisions in seminal sovereign
immunity cases such as Regents, Hess, and Seminole Tribe.
                 KOHN V. STATE BAR OF CALIFORNIA                      19


Had we revisited our arm of the state jurisprudence after
these cases, as several of our sister circuits have, see supra
note 6, we would have realized that the Mitchell factors had
failed to keep up.
    By contrast, the D.C. Circuit test is consistent with
current Supreme Court precedent. The intent and control
factors advance the states’ dignity interests, and the treasury
factor protects the states’ financial solvency. As a result, this
test best promotes the Eleventh Amendment’s “twin reasons
for being.” Hess, 

. Since the D.C. Circuit’s
three-factor test better encapsulates the current state of the
law than the Mitchell factors and avoids their problems, we
adopt it here and no longer endorse the Mitchell factors.
    We likewise adopt the D.C. Circuit’s rule that “[u]nder
the three-factor test, an entity either is or is not an arm of the
[s]tate: The status of an entity does not change from one case
to the next based on the nature of the suit, the [s]tate’s
financial responsibility in one case as compared to another,
or other variable factors.” P.R. Ports Auth., 

.
The Supreme Court declined to resolve this question in
Regents. 

 n.10 (“Nor is it necessary to
determine whether there may be some state instrumentalities
that qualify as ‘arms of the [s]tate’ for some purposes but not
others.”). However, an entity-based approach complies
better with the D.C. Circuit’s test, which builds an analysis
of an entity’s functions into its intent prong but does not so
narrowly limit its overall scope. See P.R. Ports Auth., 

. And, like the D.C. Circuit, other circuits
evaluate immunity at the level of the entity. 9


9
 See, e.g., Lowe v. Hamilton Cnty. Dep’t of Job & Fam. Servs., 

 (6th Cir. 2010) (declining to extend immunity to an entity after
20                KOHN V. STATE BAR OF CALIFORNIA


     The entity-based approach also makes sense as a matter
of principle. We agree with our statement in Durning that
“sovereign immunity . . . by its nature is absolute.” 

. The possibility that immunity may be waived or
abrogated does not diminish this point. Waiver and
abrogation are second-stage inquiries as to whether, if an
entity is immune, that immunity may be overcome. See
Torres v. Tex. Dep’t of Pub. Safety, 

(2022); cf. Fin. Oversight & Mgmt. Bd. v. Centro de
Periodismo Investigativo, Inc., 

, 346 (2023)
(“[W]e assume without deciding that Puerto Rico is immune
from suit in federal district court, and that the Board partakes
of that immunity. We address only whether, accepting those
premises, [the statute] effects an abrogation.”). But waiver
and abrogation do not undermine the absolute nature of the
first-stage question of whether immunity exists.
    An entity-based approach also better promotes
consistency, predictability, and finality because it settles an
entity’s immunity “unless and until there are relevant
changes in the state law governing the entity.” P.R. Ports
Auth., 

. By contrast, an activity-based
approach would allow parties to relitigate an entity’s
immunity simply by articulating the challenged activity at a
different level of generality. Thus, even once an entity was
deemed immune, it still could be “subject[ed] . . . to the
coercive process of judicial tribunals at the instance of
private parties,” undermining the very purpose of immunity.
Seminole Tribe, 

 (quoting P.R. Aqueduct &

considering all its functions, not just the function at issue, in reliance on
Regents permitting such an entity-based approach); Hudson v. City of
New Orleans, 

, 682 n.1 (5th Cir. 1999) (rejecting the view
that “we look at the [specific] function of the [entity] being sued . . . in
our Eleventh Amendment analysis” (citation omitted)).
               KOHN V. STATE BAR OF CALIFORNIA               21


Sewer Auth., 

); cf. Maliandi v. Montclair
State Univ., 

, 92–93 (3d. Cir. 2016) (rejecting the
approach of parsing claim-specific Eleventh Amendment
immunity as “untenable—both practically and in principle”).
    Though our decision to implement the D.C. Circuit’s test
represents a change in our jurisprudence, this new
framework is unlikely to lead to different results in cases that
previously applied the Mitchell factors and held an entity
entitled to immunity. The D.C. Circuit test does not
overemphasize the treasury factor or rely on considerations
that are minimally relevant to the immunity analysis, aspects
of Mitchell that could erroneously lead to a conclusion of no
immunity. Although each case will be decided on its own
facts, we have no reason to believe that our decision today
will substantially destabilize past decisions granting
sovereign immunity to state entities within the Ninth Circuit.
Indeed, that is the case here as to the California State Bar, as
we now explain.
     C. Applying the Updated Three-Factor Test
        Confirms that the California State Bar is an
        Arm of the State
    Though we update our test, the California State Bar’s
status remains the same: It is an arm of the state and entitled
to sovereign immunity.
           i. Intent
   First, California’s intent with respect to the State Bar
supports immunity. California law “characterizes” the State
Bar as a “governmental instrumentality.” P.R. Ports Auth.,

 (citing Hess, 513 U.S. at 44–45 (considering
whether legislation “type[d]” the entity “as a state agency”)).
The State Bar is codified in the California Constitution,
22             KOHN V. STATE BAR OF CALIFORNIA


which prescribes that “[e]very person admitted and licensed
to practice law in this State is and shall be a member of the
State Bar . . . .” Cal. Const. art. VI, § 9; cf. In re New York,

 (1921) (in determining that a suit against
a defendant in his official capacity was a suit against the
state, the Court noted that the defendant’s office was
“established and its duties prescribed by the Constitution of
the state”). Further, “[a]ll property of the State Bar is . . .
held for essential public and governmental purposes in the
judicial branch of the government,” 

 § 6008, and “[b]onds,
notes, debentures, and other evidences of indebtedness of the
State Bar are . . . issued for essential public and
governmental purposes in the judicial branch of
government,” id. § 6008.2.
    The state legislature’s characterization of an entity is not
the only important metric for the intent factor—state court
treatment is also relevant. See, e.g., Hess, 

(“State courts . . . repeatedly have typed the Port Authority
an agency of the [s]tates . . . .”). The California Supreme
Court’s description of the State Bar as its “administrative
arm” for attorney discipline and admission purposes cuts
decisively in favor of the State Bar’s immunity. E.g., In re
Rose, 

; In re Att’y Discipline Sys., 

. As does the California Supreme Court’s
reference to the State Bar as “a constitutional entity within
the judicial article of the California Constitution.” Obrien v.
Jones, 

 (Cal. 2000).
    The dissent focuses on the State Bar’s status as a “public
corporation” under the California Constitution to argue that
“California law treats the State Bar the same way as it treats
independent municipalities,” which “cuts strongly against
sovereign immunity.” But, as the various definitions of
“public corporation” cited by the dissent indicate, the term
               KOHN V. STATE BAR OF CALIFORNIA               23


“public corporation” can mean different things in different
places. For example, as the dissent acknowledges, while
certain provisions of California law define “public
corporation[s]” as “municipal corporation[s]” or “political
subdivision[s],” see, e.g., Cal. Gov’t Code § 67510, others
use the term “public corporation” to refer to the state of
California, see, e.g., id. §§ 6300, 12100.50, or even the
United States, see, e.g., 

.
     These varied definitions indicate that the designation
“public corporation” merely means that something is not
private. See, e.g., 

 (“As used in this
part, ‘person’ means any person, firm, association,
organization, partnership, business trust, corporation, or
company, but not including any public corporation or other
public entity.”). Beyond that, context matters. As a result,
labeling the State Bar as a “public corporation” begs the
question of whether it is an arm of the state, which is why
we apply the three-factor test.           While California’s
designation of the State Bar as a “public corporation” may
be inconclusive regarding its intent with respect to the State
Bar, the State Bar’s codification in the California
Constitution and treatment by the California Supreme Court
clarifies any ambiguity as to whether California law
“characterizes” the State Bar as a “governmental
instrumentality.” P.R. Ports Auth., 

 (citing
Hess, 513 U.S. at 44–45); see also Hagman v. Meher Mount
Corp., 

 (Cal. Ct. App. 2013)
(citing State Bar’s status as a “public corporation” under the
California Constitution for the proposition that “‘public
corporation’ is a term of art used to designate certain entities
that exercise governmental functions”).
   The dissent’s reliance on Keller v. State Bar of
California, 

 (1990), is likewise misplaced because
24             KOHN V. STATE BAR OF CALIFORNIA


the California state legislature has since restructured the
State Bar in ways that indicate a stronger intent to treat it as
an arm of the state. After Keller, the state legislature
converted the State Bar’s conference of delegates into a
private entity. 

, 2356–58 (amending

). It moved the functions
and activities of the State Bar’s sixteen specialty law
sections to a new voluntary private corporation, the
California Lawyers Association, which explicitly “shall not
be considered a state, local, or public body for any purpose.”
See The Nonprofit Association Act, 

,
3357–58 (codified at 

,
6056.3). It took the power to appoint members to the State
Bar’s governing body away from the Bar’s members,
granting total control over appointment to the three branches
of the state government. 2017 Cal. Stat. at 3353–54
(amending 

, 6013.1, 6013.3,
6013.5). Finally, today, the Bar regulates “licensees” rather
than “members” who pay “fees” rather than “dues.” 

, 4357 (amending 

). This separation of the State Bar’s associational and
regulatory functions evinces California’s intent for the State
Bar to be an arm of the state.
    Moreover, the State Bar “performs functions typically
performed by state governments.” P.R. Ports Auth., 

 (citing Hess, 

). “Since the
founding of the Republic, the licensing and regulation of
lawyers has been left exclusively to the [s]tates . . . . The
[s]tates prescribe the qualifications for admission to practice
and the standards of professional conduct. They also are
responsible for the discipline of lawyers.” Leis v. Flynt, 

 (1979). “The interest of the [s]tates in
regulating lawyers is especially great since lawyers are
               KOHN V. STATE BAR OF CALIFORNIA               25


essential to the primary governmental function of
administering justice . . . .” Goldfarb v. Va. State Bar, 

 (1975).
    The state legislature has tasked the State Bar with the
“[p]rotection of the public” in its “licensing, regulatory, and
disciplinary functions.” 

.
The State Bar carries out “the core functions of admission
and discipline of attorneys,” which go to the heart of
California’s interest in regulating lawyers. Obrien, 

 (citation omitted).          It examines candidates’
qualifications for admission and administers the bar exam,
which is a prerequisite to practicing law in the state; it also
certifies candidates for admission to the California Supreme
Court. 

, 6060(g); Cal. R. Ct.
9.3. The California Supreme Court has explicitly stated that
the State Bar’s “assistance . . . in the disciplinary process is
an integral part of the judicial function.” Obrien, 

. Consequently, the State Bar performs governmental
functions. See In re Wade, 

 (9th Cir.
1991) (per curiam) (“In conducting disciplinary
proceedings, the [Arizona State] Bar is enforcing its police
or regulatory power.”).
     We also consider whether the State Bar is “treated as a
governmental instrumentality for purposes of other
[California] laws.” P.R. Ports Auth., 

, 876
(citing Hess, 513 U.S. at 44–45). The State Bar is subject to
California public-records and open-meeting laws. 

. Its property is tax-exempt. 

 § 6008.
   Though California law authorizes the State Bar to “sue
and be sued,” id. § 6001, this provision “may fairly be
construed as limited to a waiver of immunity in the state’s
own courts,” Durning, 

 n.4 (citing Welch,
26               KOHN V. STATE BAR OF CALIFORNIA


483 U.S. at 473–74); see also supra pp. 14–16. Individuals
can challenge “[d]eterminations and recommendations of
the bar in matters of discipline and admission” in the
California Supreme Court. Saleeby v. State Bar of Cal., 

 (Cal. 1985). 10 The California Court of Appeal
has reasoned that these matters remain within the California
Supreme Court’s exclusive “original jurisdiction over the
admissions process,” as the “enactment of a comprehensive
statutory scheme . . . established a public agency . . . without
diminishing the court’s authority over admissions.” Smith v.
Cal. State Bar, 

 (Cal. Ct. App. 1989).
Thus, the most reasonable construction of the “sue and be
sued” provision is as part of this statutory scheme to
establish the conditions under which the State Bar may be
sued in state court, not as a waiver of federal sovereign
immunity.
    The dissent asserts that California law treats the State
Bar “as distinct from state agencies” because it provides that
“[n]o law of this state, restricting, or prescribing a mode of
procedure for the exercise of powers of state public bodies
or state agencies . . . shall be applicable to the State Bar,
unless the [l]egislature expressly so declares.” 

. But the California Supreme Court has
interpreted this provision in the exact opposite way: “That
the legislature considered the State Bar as at least akin to a
state public body or agency . . . is illustrated by the last
paragraph of section 6001, where it appears the [l]egislature
felt the necessity of providing that laws prescribing

10
   See also 

 (authorizing California
Supreme Court review of Bar certification decisions); 

 § 6082
(authorizing California Supreme Court review of Bar reinstatement
decisions); Cal. R. Ct. 9.13 (setting forth procedures for review of Bar
decisions).
               KOHN V. STATE BAR OF CALIFORNIA                27


procedures for state bodies . . . did not apply to the State Bar,
thus indicating that the [l]egislature considered the State Bar
in their category.” Chron. Publ’g Co. v. Superior Ct., 

 (Cal. 1960) (in bank). In sum, the State Bar
“is treated as a governmental instrumentality for purposes of
other state laws.” P.R. Ports Auth., 

, 876
(citing Hess, 513 U.S. at 44–45).
    Accordingly, the first factor—California’s intent as to
the State Bar—favors the conclusion that the State Bar is an
arm of the state and entitled to its immunity.
          ii. Control
    The second factor of control considers, first, “how the
directors and officers” of the entity “are appointed.” P.R.
Ports Auth., 

 (citing Hess, 

).
Again, this factor cuts in favor of immunity. The California
Supreme Court, the state legislature, and the governor
appoint the State Bar’s Board of Trustees. 

, 6013.1, 6013.3, 6013.5. Officials within the
three branches of the state government also appoint the
Committee of Bar Examiners, a body that oversees the bar
exam and admission. 

 §§ 6046, 6046.5; Cal. R. Ct. 9.4.
Thus, the power to appoint the State Bar’s governing
structure is housed wholly within the state government.
    Beyond appointment, the California Supreme Court
exercises significant control over the State Bar’s
functioning. The California Supreme Court has “inherent
jurisdiction over the practice of law” in the state, so the State
Bar “acts under the authority and at the direction of the
Supreme Court.” Cal. R. Ct. 9.3. Admission rules adopted
by the State Bar are subject to California Supreme Court
review and approval, id. at 9.5, and the State Bar must report
to the California Supreme Court on each administration of
28             KOHN V. STATE BAR OF CALIFORNIA


the bar exam, id. at 9.6(c). Relatedly, the State Bar’s
admission and disciplinary decisions are subject to
California Supreme Court review. 

, 6082; Cal. R. Ct. 9.13. The California Supreme
Court has explained that “in matters of discipline and
disbarment, the State Bar is but an arm of th[e] court,” which
“retains its power to control any such disciplinary
proceeding at any step,” In re Att’y Discipline Sys., 

 (citation omitted), including by, for example, imposing
procedural standards on such proceedings, Emslie v. State
Bar of Cal., 

 (Cal. 1974).
    Presented with a comparably close relationship between
the Arizona Supreme Court and the Arizona State Bar, the
Supreme Court concluded that the Arizona State Bar’s
actions were state action and therefore exempt from antitrust
law. Bates v. State Bar of Ariz., 

, 361 (1977).
The Supreme Court reasoned that where the Arizona State
Bar’s role was “completely defined” by the Arizona
Supreme Court, and it acted “as the agent of the court under
its continuous supervision,” claims arising from that role
were “against the State. The Arizona Supreme Court [was]
the real party in interest.” Id. Likewise, the California
Supreme Court’s significant degree of control over the State
Bar strongly suggests that the State Bar is an arm of the
judicial branch of California.
    The dissent recognizes that the State Bar is subject to the
Supreme Court’s “supervision.” However, it contends that
“supervision is not control” because the Supreme Court of
California “does not veto the decisions of the State Bar” but
“merely chooses whether to adopt the State Bar’s
recommendations as to admission and discipline.” This is a
distinction without a difference, as the Supreme Court of
California need not have the power to veto the decisions of
                KOHN V. STATE BAR OF CALIFORNIA                29


the State Bar when it has total control over which of those
decisions will be adopted in the first place.
    The California state legislature also controls the State
Bar’s ability to raise revenue. Though the legislature has
authorized the State Bar to raise its own funds, which are
“paid into the treasury of the State Bar,” id. § 6063, the
legislature sets an annual cap on the amount the State Bar
can charge in licensee fees, id. § 6140, and requires the State
Bar to submit an annual budget for the legislature’s review
and approval in conjunction with any bill that would
authorize the State Bar to collect such fees, id. § 6140.1. The
legislature’s power over the State Bar’s fundraising ability
and annual budget further illustrates the state’s control over
the State Bar.
   Thus, the second factor of control also favors the State
Bar’s immunity.
          iii. Treasury
     Finally, we consider the State Bar’s financial
relationship to California and its overall effects on
California’s treasury. P.R. Ports Auth., 

(citing Hess, 513 U.S. at 43–44). “In analyzing this third
factor . . . the relevant issue is a [s]tate’s overall
responsibility for funding the entity or paying the entity’s
debts or judgments, not whether the [s]tate would be
responsible to pay a judgment in the particular case at
issue.” 

 (emphasis in original) (citations omitted).
    There is no dispute that California law makes the State
Bar responsible for its own debts and liabilities, so California
would not be liable for a judgment against the State Bar. 

. The State Bar, however, posits
that, because “[a]ll property of the State Bar . . . [is] held for
30             KOHN V. STATE BAR OF CALIFORNIA


essential public and governmental purposes in the judicial
branch of government,” 

 § 6008, the State Bar’s funds are
state funds. The State Bar also points to the control the
California state legislature exercises over its ability to raise
revenue. See supra p. 29. These aspects of the State Bar’s
financial administration do not prove that California is
responsible for funding the State Bar or paying its debts or
judgments. But they are indicia of California’s intent with
respect to the State Bar and control over it, and they
undermine Kohn’s portrait of the State Bar as a financially
self-sustaining, independent entity.
    The State Bar also relies on cases where we considered
whether the State, even if not directly liable for a judgment
against an entity under state law, would be the “real,
substantial party in interest,” Regents, 

(citation omitted), because the entity performed essential
governmental functions that the state could not do without.
For example, in Alaska Cargo Transport, Inc. v. Alaska R.R.
Corp., 

 (9th Cir. 1993), Alaska would not have
been liable for a judgment against a state-created railroad.

 at 380–81. But the railroad was “a unique and essential
fixture in the lives of thousands of widely dispersed
Alaskans” and “perform[ed] a vital governmental function.”

 Therefore, we concluded that, in the face of a large
judgment, the railroad “would be compelled to turn to
legislative appropriation in order to remain in business, and
the legislature would have to respond favorably so that the
‘essential’ transportation function would continue to be
performed . . . .” 

 (citation omitted).
    Similarly, in Aguon v. Commonwealth Ports Auth., 

 (9th Cir. 2003), we determined that the
Commonwealth Ports Authority of the Northern Mariana
Islands, a public corporation created to operate and manage
                KOHN V. STATE BAR OF CALIFORNIA                31


the Northern Mariana Islands’ ports, “perform[ed] a central
governmental function,” so that if it “were to be faced with
a large money judgment which it could not pay, the
Commonwealth would be compelled to protect its island
economy by responding with an appropriation to provide the
citizens of the Commonwealth with essential seaport and
airport services.” 

 at 902–03.
    The State Bar presses a similar argument. As a surrogate
for the California Supreme Court, the State Bar performs “a
vital governmental function” in the regulation of lawyers.
Alaska Cargo Transp., Inc., 

. Therefore, a
“structure of compulsion” might force California “into the
role of real, substantial party in interest” if the State Bar were
unable to satisfy a money judgment against it to ensure that
the State Bar could continue to serve this role. Holz v.
Nenana City Pub. Sch. Dist., 

 (9th Cir.
2003).
    This factor presents a closer call. The State Bar’s
functions are “essential to the primary governmental
function of administering justice . . . .” Goldfarb, 

. However, we also have said that “in the absence of
a showing that money used to pay a judgment will
necessarily be replaced with state funds, ‘we adhere to our
basic proposition that the fact that the state may ultimately
volunteer to pay the judgment . . . is immaterial . . . .’”
Beentjes, 

 (quoting Holz, 

).
     Either way, despite Kohn’s arguments to the contrary,
this third factor is not dispositive. See supra pp. 13–14.
Given that the intent and control factors strongly favor the
conclusion that California “structured” the State Bar to
“enjoy the special constitutional protection of the [s]tate[]
[itself],” the third factor, placed in its proper context, cannot
32                KOHN V. STATE BAR OF CALIFORNIA


overcome the first two. Hess, 513 U.S. at 43–44 (citation
omitted). We see no reason to disturb our nearly forty-year-
old determination that the California State Bar is an arm of
the state and entitled to immunity in federal court.
                                   ***
    This conclusion puts us in good company. In the years
since we last considered the State Bar’s immunity in Lupert
and Hirsh, all the other federal circuits to have considered
the question have agreed: State bars are arms of the state and
enjoy sovereign immunity under the Eleventh
Amendment. 11
    The one circuit court decision that bucks this trend is our
own. In Crowe v. Or. State Bar, 

 (9th Cir. 2021)
(per curiam), we applied the Mitchell factors to conclude that
the Oregon State Bar is not an arm of the state. 

 at 731–
33. Although there may be some differences between the
California and Oregon State Bars, whether the Oregon State

11
  See, e.g., T.W. v. N.Y. State Bd. of L. Exam’rs, 

 (2d Cir.
2021) (“The Board of Law Examiners, as an ‘arm[]’ of the State of New
York, ‘share[s] in [its] immunity . . . .” (first and second alterations in
original) (citation omitted)); Nichols v. Ala. State Bar, 

(11th Cir. 2016) (per curiam) (holding Alabama State Bar immune given
that its powers were “public in nature and would otherwise be exercised
by the Alabama Supreme Court”); Dubuc v. Mich. Bd. of L. Exam’rs,

 (6th Cir. 2003) (holding Michigan Board and Bar
immune in the absence of evidence as to whether Michigan would be
responsible for judgment against the Bar because “the Board and the Bar
are merely extensions of the Michigan Supreme Court”); Thiel v. State
Bar of Wis., 

 (7th Cir. 1996) (“The State Bar is immune
from suit under the Eleventh Amendment.”), overruled on other grounds
by Kingstad v. State Bar of Wis., 

 (7th Cir. 2010);
Green v. State Bar of Tex., 

, 1087–88 (5th Cir. 1994)
(dismissing § 1983 claims against Texas State Bar committee as barred
by the Eleventh Amendment).
               KOHN V. STATE BAR OF CALIFORNIA                33


Bar would be an arm of the state under the three-factor test
we now employ, rather than the Mitchell factors, is not
before us today. Any future case brought against the Oregon
State Bar will need to be analyzed under the new test we
articulate in this decision.
  III. CONCLUSION
    In sum, we update our arm of the state jurisprudence to
better reflect the Supreme Court’s latest guidance and affirm
our precedent that the California State Bar is entitled to
immunity from suit in federal court. We remand to the
original three-judge panel for consideration of the remaining
issues consistent with this opinion.
   AFFIRMED IN PART and REMANDED to the
three-judge panel.


MENDOZA, Circuit Judge, concurring in part:

    I agree with the majority that the factors set out in
Mitchell v. Los Angeles Community College District are out
of step with the Supreme Court’s jurisprudence. I also agree
that, in this case, the California Bar is an arm of the state for
sovereign immunity purposes. I write separately for two
reasons. First, I probe the panel’s adoption of the D.C.
Circuit’s sovereign immunity test and its reading of Hess v.
Port Authority Trans-Hudson Corp. The D.C. Circuit’s
three-factor test makes good legal and practical sense.
But that circuit’s approach to weighing the sovereign
immunity factors hews far more closely to Justice
O’Connor’s dissenting opinion in Hess than Justice
Ginsburg’s majority. I do not read our majority opinion as
adopting that aspect of the D.C. Circuit’s reasoning, and I
34             KOHN V. STATE BAR OF CALIFORNIA


caution against doing so. Second, I disagree with the
majority’s wholesale embrace of the D.C. Circuit’s entity-
based approach to sovereign immunity. This case was a
close call, and I urge my colleagues to be wary of deeming
certain state instrumentalities—which often perform
functions unrelated to the express delegation of state
power—categorically immune from every federal suit.
Doing so lacks good cause in either precedent or fact.
                              I
    The scope of state sovereign immunity extends more
broadly than the Eleventh Amendment’s text. See Alden v.
Maine, 

 (1999); Seminole Tribe of Fla. v.
Florida, 

 (1996) (“[W]e have understood the
Eleventh Amendment to stand not so much for what it says,
but for the presupposition . . . which it confirms.” (quoting
Blatchford v. Native Vill. of Noatak, 

(1991))). The Eleventh Amendment, passed by Congress in
1794 and ratified by the states in 1795, accomplishes more
than nullifying Chisholm v. Georgia to protect states’ purses
and restricting the Article III diversity jurisdiction of the
federal courts. See Seminole Tribe, 

, 68 (citing
Chisholm v. Georgia, 

 (1793)). Its “object and
purpose” is “to prevent the indignity of subjecting a state to
the coercive process of judicial tribunals at the instance of
private parties.” Ex parte Ayers, 

 (1887);
see also Alden, 

. The Eleventh Amendment
thus confirms the centuries-old presupposition that “each
State is a sovereign entity in our federal system,” Seminole
Tribe, 

 (citing Hans v. Louisiana, 

 (1890)), and that “courts may not ordinarily hear a suit
brought by any person against a nonconsenting state,”
Torres v. Tex. Dep’t of Pub. Safety, 597 U.S. --, 

, 2461–62 (2022). At its core, Eleventh Amendment
              KOHN V. STATE BAR OF CALIFORNIA             35


immunity is rooted in the “respect owed [the states] as
members of the federation.” P.R. Aqueduct & Sewer Auth.
v. Metcalf & Eddy, Inc., 

 (1993).
    The Eleventh Amendment sometimes extends sovereign
immunity to state instrumentalities that operate as arms of
the state, barring valid abrogation or waiver. See P.R.
Aqueduct & Sewer Auth., 

 (“Absent waiver,
neither a State nor agencies acting under its control may ‘be
subject to suit in federal court.’”); Alden, 

.
The Supreme Court, however, has offered limited guidance
as to when respect for a state’s dignity compels us to
immunize a state instrumentality. Treasury concerns have
consistently and historically spurred the extension of
sovereign immunity to state-created entities. See, e.g., Lake
Country Ests., Inc. v. Tahoe Reg’l Plan. Agency, 

 (1979) (“[S]ome agencies exercising state power
have been permitted to invoke the Amendment in order to
protect the state treasury from liability[.]”); see also
Edelman v. Jordan, 

 (1974) (“[A] suit by
private parties seeking to impose a liability which must be
paid from public funds in the state treasury is barred by the
Eleventh Amendment.”). The legal “structure” of the entity
affects any sovereign immunity analysis, as well. Lake
Country Ests., 

; see also Mt. Healthy City
Sch. Dist. Bd. of Educ. v. Doyle, 

 (1977)
(considering the “nature of the entity created by state law”
in determining whether the entity was an arm of the state).
The Court has also directed us to examine various factors,
including whether the entity receives guidance and extensive
funds from the state, the entity’s ability to raise revenue,
whether the state appoints its board, the function of the
entity, and the entity’s power to issue bonds or levy taxes.
Mt. Healthy, 

; Lake Country Ests., 

             KOHN V. STATE BAR OF CALIFORNIA


at 400–02. But these cases do not lay out a concrete arm-of-
the state test for the extension of Eleventh Amendment
sovereign immunity.
                              A
    Hess v. Port Authority Trans-Hudson Corp. is a notable
exception. 

 (1994). In Hess, the Court addressed
whether the Eleventh Amendment immunized a bistate Port
Authority, charged with coordinating transport through a
port on the New York-New Jersey border, from a personal
injury suit brought by railroad workers under federal law. 

. It answered “no.” 

 at 32–33. Guided by Lake
Country Estates’ examination of a bistate entity, the Hess
Court first focused on the entity’s “structure,” examining
various “indicators” of immunity, including: (1) the entity’s
“function”; (2) the nature of the entity’s governing body;
(3) the entity’s implementing legislation; and (4) whether
the states have “financial responsibility” for the entity,
including “legal liability.” 

 at 43–46. Unlike the bistate
entity in Lake Country Estates, where the factors disfavored
sovereign immunity, the Hess Court concluded that those
factors “point[ed] in different directions” for the Port
Authority’s immunity from suit. 

. The Court thus
needed a tiebreaker.
    So it invoked the Eleventh Amendment’s “twin reasons
for being” to guide its analysis: “solvency” and “dignity.”
Hess, 

. Those twin reasons, however, have
a critical factor in common: the practical and legal effect of
a judgment on a state’s treasury, which trumped concerns
over “control” of the entity. See 

; 

 (“[T]he
impetus for the Eleventh Amendment[ is] the prevention of
federal-court judgments that must be paid out of a State’s
treasury.”). Indeed, the Hess Court explicitly declined to
               KOHN V. STATE BAR OF CALIFORNIA               37


“render[] control dispositive,” reasoning that control was a
“perilous,” “uncertain,” and “unreliable” indicator of a
state’s intention to make an agency immune. 

 at 47–48
(citations omitted). By contrast, addressing whether a
judgment against the entity would put the state’s purse on
the line was certain; and the Hess Court thus affirmed that
treasury concerns are “the most salient factor in Eleventh
Amendment determinations.” 

. Or, as Justice
O’Connor’s dissent succinctly puts it: “for determining arm-
of-the-state status, we may now substitute a single
overriding criterion, vulnerability of the state treasury.” 

 (O’Connor, J., dissenting). After all, “[t]he Court
dismisses consideration of control altogether.” 

(O’Connor, J., dissenting).
                               B
     Three years later, the Court issued its decision in Regents
of the University of California v. Doe. 

 (1997).
Like Hess, the Court in Regents addressed when to apply
Eleventh Amendment immunity to a state instrumentality,
zeroing in on “the relationship between the State and the
entity in question.” 

. To determine whether this
relationship gives rise to immunity, the Regents Court
examined (1) “the essential nature and effect of the
proceeding”; (2) “the nature of the entity created by state
law”; and, as in Lake Country Estates and Hess, (3) “whether
a money judgment against a state instrumentality or official
would be enforceable against the State . . . .” 

 at 429–30
(citing, among other cases, Hess, 

; Mt. Healthy,

; Kennecott Copper Corp. v. State Tax
Comm’n, 

 (1946); Ford Motor Co. v. Dep’t of
Treasury of State of Ind., 

 (1945)). And the
Regents Court reaffirmed the “considerable importance” of
that last consideration. 

.
38             KOHN V. STATE BAR OF CALIFORNIA


                              II
     Despite the Court’s guidance in Regents and Hess, the
circuit courts have struggled to translate that precedent into
a workable, arm-of-the-state standard for Eleventh
Amendment sovereign immunity. Most have revised or
expanded upon existing, factor-based tests to assess whether
subjecting a state agency to suit in federal court would
infringe on the “dignity,” “solvency,” and “respect”
concerns underpinning the Eleventh Amendment. Some
follow Hess quite faithfully. They examine the sovereign
nature of the entity using factor-based tests; if those factors
point in different directions, they determine whether
“solvency” and dignity” concerns compel an answer, while
recognizing that a judgment’s impact on the state’s treasury
is the most important factor. See, e.g., Fresenius Med. Care
Cardiovascular Res., Inc. v. P.R. & Caribbean
Cardiovascular Ctr. Corp., 

 (1st Cir. 2003).
Other circuits employ balancing tests—or in the Second
Circuit’s case, nesting, factor-based, balancing tests—to
assess whether the Eleventh Amendment shields a state
instrumentality from suit. See, e.g., Leitner v. Westchester
Cmty. Coll., 

, 135–36 (2d Cir. 2015) (citing
Clissuras v. City Univ. of N.Y., 

 (2d Cir. 2004)
and Mancuso v. N.Y. State Thruway Auth., 

 (2d
Cir. 1996)); Sw. Bell Tel. Co. v. City of El Paso, 

 (5th Cir. 2001). Some abandon a test altogether,
see Takle v. Univ. of Wis. Hosp. & Clinics Auth., 

, 769–81 (7th Cir. 2005), or, following Regents’ lead,
focus primarily on the relationship between the entity and
the state, Duke v. Grady Mun. Schs., 

 (10th
Cir. 1997).
   Our decision in Mitchell, which preceded Regents and
Hess, was just such a factor-based test—an imperfect
               KOHN V. STATE BAR OF CALIFORNIA              39


vehicle, perhaps, but one that we repeatedly defended as
compatible with Supreme Court precedent. See, e.g., Ray v.
Cnty. of L.A., 

 (9th Cir. 2019); see also
Crowe v. Or. State Bar, 

 (9th Cir. 2021),
cert denied 

 (Mem). Today, we reverse course
and align ourselves with the D.C. Circuit’s approach in
Puerto Rico Ports Authority v. Federal Maritime
Commission, 

 (D.C. Cir. 2008). That decision
directs courts to examine: “(1) the State’s intent as to the
status of the entity, including the functions performed by the
entity; (2) the State’s control over the entity; and (3) the
entity’s overall effects on the state treasury.” 

. I
concur in that choice—the D.C. Circuit’s test is a fair one.
But I caution us from taking the D.C. Circuit’s decision too
far by adopting (1) its approach to weighing the sovereign
immunity factors, which curiously departs from Hess’s
majority in favor of its dissent; or (2) its entity-based
approach to sovereign immunity.
                              A
    The Eleventh Amendment’s “twin reasons for being”—
“state dignity” and “solvency”—ought to be afforded equal
weight in our sovereign immunity analysis. See Hess, 

; see also Auer v. Robbins, 

, 456
n.1 (1997) (affording equal weight to treasury and dignity
considerations). I like the majority’s analytical framework
for these concerns, which neatly parses the “intent” and
“control” factors as advancing the state’s dignity interest,
while the “treasury” factor protects the states’ financial
solvency. But I am concerned by the equal weight that the
D.C. Circuit and, by implication, the majority, appear to give
the intent, control, and treasury factors, respectively. I read
the D.C. Circuit’s decision as putting its thumb on the scale
for “state dignity” (which, adding together its attendant
40             KOHN V. STATE BAR OF CALIFORNIA


factors, would get approximately two-thirds of the immunity
vote), veering quite far from Hess and Regents. In my view,
control and intent should count for half the vote, and the
treasury factor should be weighed equally against them. To
be consistent with the Court’s guidance, I urge my
colleagues not to read our decision today as de-emphasizing
solvency concerns in favor of those that implicate dignity.
    In my and most circuit courts’ view, Supreme Court
precedent is clear: whether the state legally or practically
pays a money damages judgment against the entity is central
to the sovereign immunity analysis. See supra Section I; see
also Fresenius Med. Care, 

; Leitner, 

 (“The first factor, ‘the vulnerability of the State’s
purse,’ is ‘the most salient factor in Eleventh Amendment
determinations.’” (quoting Hess, 513 U.S. at 47–48)); Md.
Stadium Auth. v. Ellerbe Becket Inc., 

, 261 (4th
Cir. 2005) (“[T]he most important consideration is whether
the state treasury will be responsible for paying any
judgment that might be awarded.” (citation omitted)); Sw.
Bell Tele. Co., 

 (“[C]ourts must review . . .
whether a money judgment against the instrumentality
would be enforceable against the state.”); Waskul v.
Washtenaw Cnty. Cmty. Mental Health, 

(6th Cir. 2020) (reciting that the “state’s potential legal
liability for a judgment against the defendant ‘is the foremost
factor’ to consider in our sovereign immunity analysis”);
Takle, 

 (noting that an entity would be
immune if it “were financed by the state . . . so that any
judgment against it would be paid out of state funds”);
Thomas v. St. Louis Bd. of Police Comm’rs, 

 (8th Cir. 2006) (noting that courts assess autonomy and
control and, “more importantly, whether a money judgment
against the agency will be paid with state funds”); Duke, 127
               KOHN V. STATE BAR OF CALIFORNIA             41


F.3d at 975 (“The Supreme Court has indicated more
recently that ‘the vulnerability of the State’s purse [i]s the
most     salient    factor    in    Eleventh       Amendment
determinations.’”) quoting Hess, 

)) (alteration
in Duke). We said the same two years ago. See Crowe, 

. And while solvency is not the sole concern that
we address when considering an entity’s sovereign
immunity, see Seminole Tribe, 

 (considering
what protections a state is owed by the Eleventh
Amendment, and not who is entitled to share that
protection), that does not mean it is lesser than its twin.
    To read otherwise, as the D.C. Circuit does, draws far
closer to Hess’s dissent than its majority. In dismissing a
party’s outsized reliance on treasury considerations, the D.C.
Circuit reasoned that predominantly focusing on a state’s
financial liability in the lawsuit “would inappropriately
convert a sufficient condition for sovereign immunity into
the single necessary condition for arm-of-the-state status.”
P.R. Ports Auth., 

 (emphasis in original).
Citing Hess, the D.C. Circuit further stated: “That is not the
law[.]” 

. But under a plain reading of Hess’s
majority, that was the law; it was Justice O’Connor’s dissent
that sought to limit such an approach. Hess, 

(O’Connor, J., dissenting). Criticizing the majority’s
“conclusion that the vulnerability of the state treasury is
determinative,” she wrote that its treasury-focused analysis
“takes a sufficient condition for Eleventh Amendment
immunity, and erroneously transforms it into a necessary
condition.” 

 (emphasis in original). Like the D.C.
Circuit in Puerto Rico Ports Authority, Justice O’Connor’s
dissent in Hess thus championed state “control” as either co-
extensive with, or more important than, “treasury” concerns
when assessing sovereign immunity. Compare Hess, 513
42             KOHN V. STATE BAR OF CALIFORNIA


U.S. at 61 (O’Connor, J. dissenting) (reasoning that “control
can exist even where the State assumes no liability for the
entity’s debts”) with P.R. Ports Auth., 

(rejecting the argument “that there is no sovereign immunity
if the State is not obligated to pay a judgment in a particular
case”).
    But the Supreme Court has expressly declined to
embrace Justice O’Connor’s attempts in Hess to “look
beyond the potential impact of an adverse judgment on the
state treasury, and examine the extent to which the elected
state government exercises ‘real, immediate control and
oversight’ over the [entity.]” Regents, 519 U.S. at 431–32
(quoting Hess, 

 (O’Connor, J., dissenting)).
And the Court reaffirmed the critical role played by the
treasury consideration to Eleventh Amendment immunity
for state instrumentalities. Id. at 431. It might be true that
an equally weighted, intent-control-treasury test would
avoid Hess’s “counterproductive” and “objectionable”
aspects, including its “nearly exclusive focus on the
vulnerability of the state’s treasury.” Héctor G. Bladuell,
Twins or Triplets?: Protecting the Eleventh Amendment
through a Three-Prong Arm-of-the-State Test, 

, 842 (2007) (proposing, in large part, the test
adopted by the D.C. Circuit the following year). But I agree
with the First Circuit—“Hess binds us,” Fresenius Med.
Care, 322 F.3d at 67–68, and we can neither sidestep it nor
recharacterize it.
    Thankfully, our decision today does not reach this issue
and the D.C. Circuit’s decision does not foreclose our
approach to weighing these factors. Here, the first two
factors weigh strongly in favor of immunity, and the third
factor, though “mixed,” indicates a significant financial
relationship between the State Bar and California. So we
               KOHN V. STATE BAR OF CALIFORNIA             43


need not and, as I read the majority, do not determine the
exact weight that we should ascribe to each factor. To
remain consistent with Hess and Regents, I encourage us not
to neglect solvency for dignity by indexing too heavily on
control and intent.
                              B
    Unlike the majority, I hesitate to embrace the D.C.
Circuit’s conclusion that “once an entity is determined to be
an arm of the State under the three-factor test, that
conclusion applies unless and until there are relevant
changes in the state law governing the entity.” P.R. Ports
Auth., 

. In my view, we need not reach this
issue today, given that it was neither briefed nor argued. And
while this categorical approach to sovereign immunity may
make our job easier as judges, it lacks consistent support in
our precedent or practice and would lead to anomalous
results.
    The majority is right that our Mitchell-based precedent
on the question of “entity versus activity” immunity is
mixed. Compare Durning v. Citibank, N.A., 

 (9th Cir. 1991), with Ray, 

. In 1995,
however, we held that state entities that function in various
capacities are “not entitled to Eleventh Amendment
immunity” with respect to every function; instead, the
indicia of sovereign immunity “must be examined closely to
ascertain that the [entity] is indeed functioning as an arm of
the state.” Doe v. Lawrence Livermore Nat’l Lab’y, 

, 775 (9th Cir. 1995), rev’d sub nom. on other grounds
by Regents of the Univ. of Cal. v. Doe, 

 (1997).
When the Supreme Court reversed our decision in Doe, it
declined to disturb that holding. See Regents, 

 n.2. This might well be because the “function” of an
44             KOHN V. STATE BAR OF CALIFORNIA


entity, including the function that “g[ives] rise to the specific
controversy at issue in [the] litigation,” sheds significant
light on the sovereign immunity analysis. See Lake Country
Ests., 440 U.S. at 402; Hess, 513 U.S. at 43–46; c.f., N. Ins.
Co. of N.Y. v. Chatham Cnty., GA., 

 (2006)
(phrasing the relevant immunity question as whether “the
County . . . was acting as an arm of the State . . . in operating
the drawbridge”); Regents, 519 U.S. at 431–32 (leaving open
whether it is appropriate to consider, for sovereign immunity
purposes, “the character of the function that gave rise to the
litigation”). See also Walker v. Jefferson Cnty. Bd. of Educ.,

 (11th Cir. 2014) (tethering the sovereign
immunity analysis to “the particular function in which the
[entity] was engaged when taking the actions out of which
liability is asserted to arise”) (alteration in original). After
all, contrary to our reasoning in Durning, sovereign
immunity is not always absolute; it is subject to waiver and
valid abrogation. See Lapides v. Bd. of Regents of Univ. Sys.
of Ga., 

 (2002). And while a state’s
immunity may well be absolute as a first-stage inquiry, I do
not see why a state instrumentality’s immunity necessarily
follows suit.
    Preserving a function-based approach instead of the D.C.
Circuit’s entity-based approach serves the Eleventh
Amendment’s twin purposes. See Seminole Tribe, 

 (noting that the Eleventh Amendment prevents
“federal-court judgments that must be paid out of a State’s
treasury” and “serves to avoid ‘the indignity of subjecting a
State to the coercive process of judicial tribunals at the
instance of private parties’”) (cleaned up). First, the
function-based approach narrows our focus to the entity’s
conduct in the dispute at hand, avoiding an overbroad view
of sovereign immunity. An arm of the state is not the state,
                KOHN V. STATE BAR OF CALIFORNIA                45


and its dignity interest is relevant insofar as it functions as
an arm of the state. After all, it is the entity’s conduct, and
not merely its legal status, that brings it to court, thus raising
Eleventh Amendment concerns in the first place. See Lake
Country Ests., 440 U.S. at 402; c.f., ex parte Ayers, 

; Regents, 

 (assessing the “nature and
effect of the proceeding” to determine the relationship
between the state instrumentality and the state). Indeed, our
very framing of the question—“whether a state
instrumentality may invoke the State’s immunity”—
reinforces this view. See Regents, 

 (emphasis
added). I see nothing in our or the Supreme Court’s Eleventh
Amendment jurisprudence that would require us to declare a
state instrumentality, like a state, sovereignly immune in
perpetuity.
    Second, a function-based approach avoids a far-ranging
inquiry into potentially irrelevant aspects of an entity’s legal
structure, which might be leveraged to immunize that entity
for conduct otherwise divorced from or beyond the state’s
mandate. State instrumentalities are, increasingly, sprawling
institutions with generalized state mandates, attending to a
myriad of far narrower state, local, and even private
interests. And it is not hard to imagine a scenario in which
a state instrumentality is deemed immune as an “arm of the
state” in one case, despite administering to a host of local
and private interests not at issue in that litigation. Our new
entity-based approach would broadly immunize that entity’s
conduct and any future conduct, categorically granting or
denying immunity regardless of that instrumentality’s
activity.     Indeed, the entity-based approach could,
ostensibly, immunize a state instrumentality even when it
acts contrary to or in excess of the direction and authority it
received from the state. On balance, a function-based
46             KOHN V. STATE BAR OF CALIFORNIA


approach thus avoids over-and under-categorizations of
sovereign immunity, aligning our doctrine with Hess’s and
Regents’ goals.
    These theoretical concerns have practical implications.
Take, for example, public universities, which have long been
afforded sovereign immunity under the Eleventh
Amendment. See, e.g., Lapides, 

 (assuming
the university system’s immunity for purposes of examining
waiver); Rounds v. Or. State Bd. of Higher Educ., 

, 1035 (9th Cir. 1999) (“[T]he University performs the
central governmental function of providing opportunities for
‘deserving and qualified citizens to realize their aspirations
for higher education’ . . . . [It] is an arm of the State of
Oregon for Eleventh Amendment immunity purposes.”)
(internal citation omitted). Yet this immunity, construed at
the entity level, protects universities from a variety of suits
seemingly divorced from their state mandate to provide
higher education, such as federal patent prosecution. Indeed,
state universities successfully invoke sovereign immunity to
block collaborative research partners from challenging the
inventorship of patents arising out of that research, see
Xechem Int’l, Inc. v. Univ. of Tex. M.D. Anderson Cancer
Ctr., 

 (Fed. Cir. 2004); dictate venue for their
own patent enforcement lawsuits, see Tegic Commc’ns
Corp. v. Bd. of Regents of Univ. of Tex. Sys., 

(Fed. Cir. 2006); and avoid being compelled to join a patent
infringement lawsuit brought by an exclusive licensee of the
university’s patent, see Gensetix, Inc. v. Bd. of Regents of
Univ. of Tex. Sys., 

 (Fed. Cir. 2020). Their
private counterparts have no such privilege. Accord
Christopher M. Holman, State Universities Push the Limits
of Eleventh Amendment Sovereign Immunity at the Federal
Circuit, 39 Biotech. L. Rep. 347, 360 (2020) (“[State
               KOHN V. STATE BAR OF CALIFORNIA              47


universities] exploit the patent system as a sword, while
largely insulating themselves from liability or judicial
intervention through the shield of Eleventh Amendment
state sovereign immunity.”) (citing Pennington Seed, Inc. v.
Produce Exch. No. 299, 

 (Fed. Cir.
2006)).
    Or consider the challenge posed by county sheriffs’
departments. They are often structured as quasi-local and
quasi-state entities, following mandates issued by both
governments. Yet, when they are categorically deemed arms
of the state, plaintiffs are “unable to sue” over “entirely
locally dictated policies.” See Kelsey Joyce Dayton,
Tangled Arms: Modernizing and Unifying the Arm-of-the-
State Doctrine, 86 U. Chi. L.R. 1604, 1650 (2019) (citing
Cromer v. Brown, 

 (4th Cir. 1996)).
Adopting an explicit, entity-based approach to sovereign
immunity only magnifies these concerns because it permits
a state instrumentality to avoid federal lawsuits in
perpetuity—regardless of the roles it chooses to play or the
actions it chooses to take.
    Today’s case illustrates the value of a function-based
approach that accounts for the conduct at issue in the dispute.
Consistent with the D.C. Circuit’s test, both the majority and
the dissent address the “intent” factor by devoting
considerable attention to the “functions” performed by the
State Bar. The dissent, relying on Keller v. State Bar of
California, 

 (1990) and Crowe, 

,
construes “function” broadly, and it discusses the State Bar’s
largely “advisory” role, which points away from immunity.
By contrast, the majority focuses on the State Bar’s role in
licensing and regulating lawyers, including its “core
functions of admission and discipline of attorneys.” In
finding the State Bar immune, the majority draws particular
48             KOHN V. STATE BAR OF CALIFORNIA


attention to the State Bar’s job of “examin[ing] candidates’
qualifications for admission and administer[ing] the bar
exam.” In my view, the majority’s approach is more
persuasive because it examines the State Bar’s function as it
relates to the conduct at issue here—namely, the State Bar’s
allegedly discriminatory administration of the bar exam. An
entity-based approach, by contrast, provides little means of
resolving the sovereignty dispute teed up by the majority and
dissent over the State Bar’s broad function. In essence, it
leaves parties to dredge up aspects of that entity’s legal status
to make their case for or against immunity, without reference
to the challenged conduct at hand.
    It seems logical to me that consideration of the function-
at-issue must remain relevant to the sovereign immunity
analysis. Nor does it seem particularly unworkable. We
managed to successfully grapple with “the function at issue”
analysis for many years—even finding it consistent with
Supreme Court precedent. See Ray, 

(reasoning that the “function at issue” test “fits with the
Court’s statement in Chatham”); see also Streit v. Cnty. of
L.A., 

 (9th Cir. 2001). So I would not
abandon it now.
                             ***
    In sum, I do not see our decision today as a Trojan horse,
carrying Hess’s dissent in its stomach. It is, instead, a
faithful translation of the Supreme Court’s arm-of-the-state
precedent, and a long overdue update to our sovereign
immunity case law. Although I depart from the majority’s
reasoning with respect to entity-based immunity, so long as
we continue to appropriately weigh the sovereign immunity
factors, I am pleased to concur in the outcome.
               KOHN V. STATE BAR OF CALIFORNIA             49


BUMATAY, Circuit Judge, joined by SUNG, Circuit
Judge, concurring in part and dissenting in part:

    Our court rightly abandons the multi-factor test from
Mitchell v. Los Angeles Community College District, 

 (9th Cir. 1988), in favor of the D.C. Circuit’s more
streamlined approach articulated in Puerto Rico Ports
Authority v. Federal Maritime Commission, 

(D.C. Cir. 2008). Looking at intent, control, and overall
effects on a State’s treasury to determine whether an entity
is an “arm of the State” more closely aligns with Supreme
Court precedent. Those factors better illuminate the “twin”
aims of Eleventh Amendment sovereign immunity—State
dignity and State solvency. Hess v. Port Auth. Trans-
Hudson Corp., 

 (1994). So I agree with the
majority’s retirement of the Mitchell test.
    But I part ways with the majority’s application of this
new approach to the facts before us. In my view, each of its
factors cuts against finding sovereign immunity for the State
Bar of California. First, California has made evident its
intent to treat the State Bar more like an independent state-
created entity, such as a municipality, rather than an “arm of
the State.” Second, California has relinquished nearly all
direct and immediate control over the Bar. And finally,
California is not on the hook for the Bar’s funding or its
debts. With these considerations in mind, we should have
recognized that the State Bar is not entitled to the sovereign
immunity reserved only for the State and its
instrumentalities.
   For these reasons, I join Parts I, II.A, and II.B of the
majority’s opinion and respectfully dissent from the rest.
50             KOHN V. STATE BAR OF CALIFORNIA


                              I.
    The Eleventh Amendment provides that “[t]he Judicial
power of the United States shall not be construed to extend
to any suit in law or equity, commenced or prosecuted
against one of the United States by Citizens of another State,
or by Citizens or Subjects of any Foreign State.” U.S. Const.
amend. XI. The Supreme Court has long held that actions
“‘against one of the United States’ encompasses not only
actions in which a State is actually named as the defendant,
but also certain actions against state agents and state
instrumentalities.” Regents of the Univ. of Cal. v. Doe, 

 (1997) (quoting U.S. Const. amend. XI). The
Constitution extends this grant of immunity to state agencies
and instrumentalities to preserve the States’ dignity and
financial solvency—“the Eleventh Amendment’s twin
reasons for being.” Hess, 

.
    We, along with several other federal courts, have long
struggled to formulate a consistent test for determining
whether a state-created entity should be afforded sovereign
immunity. But as the majority explains, today we adopt the
D.C. Circuit’s approach, which follows the Supreme Court’s
guidance in this difficult area of law. See Maj. Op. 18–19.
That approach requires us to look at (1) whether the State
has expressed its intent to treat the entity like an arm of the
State, (2) whether the State exercises significant control over
the entity, and (3) whether private suits against the entity
would impact the State’s treasury. P.R. Ports Auth., 

.
    The question here is whether the State Bar of California
enjoys the State of California’s sovereign immunity.
Following our newly adopted analysis, I would hold that it
does not.
                KOHN V. STATE BAR OF CALIFORNIA                51


                               A.
    Start with intent. We must look first at how state law
characterizes the “nature of the entity” and whether the State
treats the entity “more like a county or a city than . . . like an
arm of the State.” Mt. Healthy City Sch. Dist. Bd. of Educ.
v. Doyle, 

 (1977).             That involves
determining whether the entity is a legal entity that exists
“separate” from the State. Lake Country Ests., Inc. v. Tahoe
Reg’l Plan. Agency, 

 (1979). From there,
we can assess whether the State intended for the entity to
enjoy sovereign immunity. Here, California law classifies
the State Bar much like a municipality, the Bar operates
unlike a state agency, and even the California Supreme Court
has disclaimed the State Bar’s role in state governance. This
factor thus cuts against immunity here.
    State Bar as a Municipality-like Public Corporation
    California law treats the State Bar the same way as it
treats independent municipalities. California’s Constitution
establishes the State Bar as a “public corporation.” Cal.
Const. art. VI, § 9. That term has been used in California to
describe municipalities and the like for nearly a century and
a half. See, e.g., Ex parte Wall, 

 (1874)
(classifying “subordinate local governments,” like
“counties, towns and cities,” as “local public corporations”),
overruled on other grounds by Ex parte Beck, 

(1912); Martin v. Aston, 

 (1882) (observing
that the California Constitution prohibits imposing “taxes
upon counties, cities, towns, or other public or municipal
corporations”); In re Werner, 

 (1900)
(observing that “[a]ll municipal corporations are public
corporations” and that “public corporation” and “municipal
corporation,” while technically distinct, are often considered
52               KOHN V. STATE BAR OF CALIFORNIA


“synonymous”). The U.S. Supreme Court has had a similar
understanding of the term. See, e.g., Trs. of Dartmouth Coll.
v. Woodward, 

, 668–69 (1819)
(opinion of Story, J.) (stating that public corporations “exist
for public political purposes only, such as towns, cities,
parishes and counties”).
    And historically, “neither public corporations nor
political subdivisions [were] clothed with that immunity
from suit which belongs to the state alone by virtue of its
sovereignty.” Hopkins v. Clemson Agric. Coll. of S.C.,

 (1911); Lincoln County v. Luning, 

 (1890) (observing that even though counties
are “integral to the State,” they are still unprotected by
sovereign immunity); Mt. Healthy, 

 (stating
sovereign immunity does not extend to “municipal
corporation[s] or political subdivision[s]”). So classification
as a “public corporation” cuts strongly against sovereign
immunity.
    Here, the State Bar was established as a “public
corporation” in 1927—not to imbue it with State authority
but to recognize its importance to the public interest. See
State Bar of Cal. v. Superior Ct. of L.A. Cnty., 

 (1929). 1 Back then, a superior court judge challenged
the State Bar’s classification as a “public corporation.” 

at 328–30. Given “its membership,” “its function,” and “its
independence of public regulation and control,” amici for the
judge argued that it should be considered a “private

1
 When the California Legislature created the State Bar, state law defined
“public corporation” as “one formed or organized for the government of
a portion of the State”—i.e., a local government. Keller v. State Bar, 

 (1989) (emphasis added) (quoting 

 (repealed)), rev’d 

 (1990).
               KOHN V. STATE BAR OF CALIFORNIA              53


corporation.” Id. at 329. The Supreme Court of California
rejected that view. While California’s highest court
recognized that the “profession and practice of the law”
involves “in a limited sense a matter of private choice and
concern,” it is also “essentially and more largely a matter of
public interest and concern.” Id. at 330. That’s because of
the “integral and indispensable” role attorneys serve in “our
system of administering justice.” Id. Thus, the profession
and practice of law “is not such a matter of purely private
concern,” id. at 332, but its relation to the administration of
civil and criminal law make it the “proper subject of
legislative regulation and control,” id. at 331. So from the
beginning, the State Bar was not conceived of as an “arm of
the State,” but an entity subject to special legislative
oversight given its unique public-interest role. Certainly,
nothing in the State Bar’s classification as a “public
corporation” grants it more immunities or privileges than
municipalities, which have no sovereign immunity.
    And the same understanding of the State Bar’s
classification as a “public corporation” exists today. The
State Bar is statutorily established as a “public corporation”
in the Business and Professions Code—not the Government
Code. See 

. Although “public
corporation” isn’t defined in the Business and Professions
Code, it is elsewhere in California law:

       • The Government Code defines “public
           corporation” as “any county, city and
           county,      city,    town,     municipal
           corporation, district of any kind or class,
           authority, redevelopment agency or
           political subdivision of this state.” 

 (as codified in the
54         KOHN V. STATE BAR OF CALIFORNIA


        San Francisco Bay Area Transportation
        Terminal Authority Act).
     • That definition was carried over to the
        Financial Code. See 

(f) (“This division does not apply
        to any public corporation as defined in
        [§] 67510 of the Government Code[.]”).
     • The Government Code defines “local
        agency” to include municipalities and
        other “public corporation[s].” See, e.g.,
        

, 53200(a),
        53215, 53227.2(a), 53460(a), 53820,
        53850(a), 54307.
     • Definitions of “political subdivision”
        often equate “public corporation[s]” with
        municipalities. See, e.g., 

 (“‘Political subdivision’
        means any county, city, city and county,
        public corporation, district or other
        political entity or public corporation of
        this State.”).
     • As do definitions of “[l]ocal public
        entity,” which “includes [any] county,
        city district, public authority, public
        agency, and any other political
        subdivision or public corporation in the
        State, but does not include the State.”
        

; 

 § 970(c)
        (similar); see also id. § 5600 (defining
        “[p]ublic body” as “any county, city and
               KOHN V. STATE BAR OF CALIFORNIA             55


           county, city, public district, public
           authority or other public corporation”).
       • And the Labor Code distinguishes the
           “State” from “public corporations” for
           employment purposes. See 

 (separating “Employer” into
           distinct categories of “(a) [t]he State and
           every State agency” and “(b) [e]ach
           county, city, district, and all public and
           quasi public corporations”); see also 

 §
           9006 (same).

    To be sure, California law also defines “public
corporation” to include the State in some limited
circumstances. See, e.g., 

(b)
(defining “Public Corporation” to include “the state” and
“municipalit[ies]” for foreign trade zones); 

 (for the Metropolitan Water District, “public
corporation” includes both the “United States,” “any other
state,” or any state “subdivision”); 

(b)(1) (for the California Foreign Investment
Program, the term means “the state” or “any corporate
municipal instrumentality”).        But notice that these
definitions are significant outliers and are limited to only
those distinct areas of the law. And invariably these
definitions include municipalities, which no one believes are
entitled to sovereign immunity. So these outliers have no
relevance for our sovereign immunity inquiry.
    Thus, under California law, the State Bar’s classification
as a “public corporation” only signifies that it should be
treated like a municipality—not an arm of the State. So it’s
a red herring to rely on assumptions about the term “public
56             KOHN V. STATE BAR OF CALIFORNIA


corporation” to find immunity here. We’ve not made these
assumptions about public corporations before. See Crowe v.
Or. State Bar, 

 (9th Cir. 2021)
(acknowledging the Oregon State Bar’s status as a “public
corporation” but ruling that it was not entitled to immunity).
    The Supreme Court of California once relied on the State
Bar’s superficial classification as a “public corporation” to
consider it a “governmental agency”—only to be
unanimously reversed by the U.S. Supreme Court. See
Keller v. State Bar, 47 Cal. 3d at 1162–63. In that case,
members of the State Bar sued the Bar for forcing them to
pay dues to advance political ideas the members disagreed
with, in violation of their First and Fourteenth Amendment
right to free speech. Keller v. State Bar of Cal., 

 (1990). The California high court reasoned that the Bar’s
status as a public corporation made it a state agency, and thus
the Bar could use dues for any purpose within the scope of
its authority. 

 at 6–7. But the U.S. Supreme Court
rejected that reasoning. It noted that the State Bar “is a good
deal different from most other entities that would be
regarded in common parlance as ‘governmental agencies.’”

. That’s because “[t]he State Bar of California was
created, not to participate in the general government of the
State, but to provide specialized professional advice to those
with the ultimate responsibility of governing the legal
profession”—the Supreme Court of California. 

.
Indeed, while the State Bar performs “important and
valuable services” for the California’s court system, the Bar
itself plays only an “advisory” role. 

. The Supreme
Court thus overruled the California court’s ruling that the
State Bar was a government entity. We should not make the
same mistake. See Crowe, 

 (holding that
Keller’s “analysis is pertinent and analogous to the
                KOHN V. STATE BAR OF CALIFORNIA                57


[sovereign] immunity question”). While the State Bar may
have separated its associational and administrative functions
since then, the administrative half of the Bar remains a
largely autonomous and advisory public corporation. So
none of the changes identified by the majority undermine the
thrust of the Supreme Court’s holding—the Bar’s mere
advisory role means that it should not be treated as a
government entity.
    State Bar’s Statutory Functions
      Beyond California’s express designation of the State Bar
as something like a political subdivision not entitled to
sovereign immunity, the State Bar does not function like a
state agency. By statute, it’s treated as distinct from state
agencies. The California Legislature expressly withheld
“the exercise of powers of state bodies or state agencies”
from the Bar. 

 (“No law of
this state restricting, or prescribing a mode of procedure for
the exercise of powers of state public bodies or state agencies
. . . shall be applicable to the State Bar, unless the Legislature
expressly so declares.”).
    On the other hand, California law imbues the Bar with
other powers typically indicative of its separate corporate
status. For example, the Bar can issue bonds in its own
name, 

(b), which “strongly
suggests that it has a legal independence from the state for
Eleventh Amendment purposes.” Durning v. Citibank, N.A.,

 (9th Cir. 1991); see also Mt. Healthy,

 (noting that a school board’s authority to
issue bonds made it more like a county or city and thus not
entitled to sovereign immunity).
   California law also gives the State Bar the power to sue
and be sued, which again is “strongly suggestive of . . .
58             KOHN V. STATE BAR OF CALIFORNIA


autonomy and independence” from the State. Durning,

 n.4. In Durning, our court held that the
Wyoming Community Development Authority did not enjoy
sovereign immunity in part because the Wyoming
Legislature “unequivocally grant[ed] the Authority the
power to ‘[s]ue and be sued’ in its own right.” 

 at 1427
(quoting Wyo. Stat. § 9–7–105(a)(i) (1977)). We face the
same circumstances here. The California Legislature
unmistakably gave the State Bar the power to “sue and be
sued.” 

. The majority
attempts to wave away this point by relying on Durning’s
caveat that “[a] mere statutory grant of the power to sue or
be sued . . . is not enough to waive immunity from suits
brought in federal court if it may fairly be construed as
limited to a waiver of immunity in the state’s own courts.”
Maj. Op. 25 (quoting Durning, 

 n.4). But
the language in § 6001 is virtually identical to the language
the Durning court said weighed against immunity. Compare

 (“The State Bar . . . may sue
and be sued.”), with Wyo. Stat. § 9–7–105(a)(i) (1977)
(stating that the authority may “[s]ue and be sued”). What’s
more, the California Attorney General is statutorily
responsible for representing all state agencies in
California—with few exceptions—but is not so obligated
with respect to the State Bar. See 

(“The Attorney General shall . . . prosecute or defend all
causes to which the state . . . is a party[.]”). That suggests
that the State Bar’s ability to “sue and be sued” is another
designation of its independent legal status.
    Of course, California law subjects the State Bar to some
of the same government-only laws as state agencies. For
instance, the State Bar must comply with California public-
records and open-meetings laws. Cal. Bus. & Prof. Code
               KOHN V. STATE BAR OF CALIFORNIA                59


§ 6001. But this alone does not swing this factor decisively
toward finding immunity, especially when the Bar’s
“separate corporate status is clearly established.” Durning,

. The Oregon State Bar was also subject to
Oregon’s public records law, and yet there we found no
immunity for the Oregon State Bar. Crowe, 

.
Determining California’s intent for its State Bar requires a
holistic approach—not one that turns on incidental
similarities between corporations and state agencies.
    State Bar as an Administrative Assistant
    Likewise, the Supreme Court of California has
disclaimed the State Bar’s role as a government
decisionmaker. The California Supreme Court has said that
the “State Bar is not in the same class as state administrative
agencies placed within the executive branch.” In re Rose, 

 (2000) (simplified). That’s because the
“State Bar Court exercises no judicial power.” Id. at 436
(emphasis added).          Rather, the State Bar “makes
recommendations” to the California Supreme Court, “which
then undertakes an independent determination of the law and
the facts, exercises its inherent jurisdiction over attorney
discipline, and enters the first and only disciplinary order.”
Id. So even though the California court has described the
State Bar as its “administrative arm,” id. at 438 (simplified),
and “a constitutional entity within the judicial article of the
California Constitution,” Obrien v. Jones, 

(2000), this isn’t dispositive of the State’s intent. Indeed, the
California Supreme Court made these observations in the
context of establishing that the State Bar is merely an
“administrative       assistant”    with     no      independent
decisionmaking authority. In re Rose, 

.
And it reaffirmed that the State Bar lacked the “powers to
regulate and control the attorney admission and disciplinary
60             KOHN V. STATE BAR OF CALIFORNIA


system”—powers that are part of the inherent judicial
authority of the California Supreme Court. Obrien,

. So given this context, these superficial
descriptions of the State Bar reveal little about the State’s
intent.
                            ***
    Taken together, the State Bar’s classification as a
municipality-like public corporation, the State Bar’s
statutory functions separate from the State and its agencies,
and the California Supreme Court’s descriptions of the Bar
as merely advisory all weigh strongly against immunity.
                              B.
    Next, we look at the amount of control California
exercises over the State Bar. Admittedly, this factor is a
closer call. But ultimately, this factor also cuts against
immunity.
    While somewhat opaque, control can be assessed based
on whether the State may “appoint and . . . remove” the
entity’s officers, whether the State may “veto [the entity’s]
actions,” and whether “the State[’s] legislature[] can
determine the projects the [entity] undertakes.” Hess, 

. But, as the Court warned, “[g]auging actual
control . . . can be a perilous inquiry, an uncertain and
unreliable exercise.” 

 (simplified). And, of course,
“ultimate control of every state-created entity resides with
the State” given the State’s power to “destroy or reshape any
unit it creates.” 

 So “ultimate control” is not dispositive.
After all, “[p]olitical subdivisions exist solely at the whim
and behest of their State . . . yet cities and counties do not
enjoy Eleventh Amendment immunity.” 

 (simplified).
               KOHN V. STATE BAR OF CALIFORNIA               61


    So we’re not looking for just any kind of control; we’re
looking for the kind that demonstrates that “the State ‘clearly
structured the entity to share its sovereignty.’” P.R. Ports
Auth., 

 (simplified). It “should turn on real,
immediate control and oversight, rather than on the
potentiality of a State taking action to seize the reins.” Hess,

 (O’Connor, J., dissenting). Indeed, in Hess,
even with the States’ power to appoint and remove officers,
the States’ veto power, and the States’ determination of the
entity’s projects, that level of control wasn’t enough to
establish sovereign immunity. 

 at 48–53 (majority
opinion).
     Here, the State Bar’s Board of Trustees is appointed by
all three branches of California’s government—the Supreme
Court of California, the State Legislature, and the Governor.
See         

, 6013.1, 6013.3, 6013.5. But that appointment
power alone doesn’t demonstrate control sufficient to find
immunity. In fact, the Court has explicitly rejected such a
myopic view of control. See Auer v. Robbins, 

, 456 n.1 (1997) (“While the Governor appoints four of
the board’s five members . . . the city of St. Louis is
responsible for the board’s financial liabilities . . . and the
board is not subject to the State’s direction or control in any
other respect. It is therefore not an ‘arm of the State’ for
Eleventh Amendment purposes.” (simplified)).
    Looking beyond appointment power, California has far
less control over the Bar’s Board. For instance, unlike the
officers in Puerto Rico Ports Authority or in Hess, the
Board’s members and officers are not removable at will. So
once they’ve appointed members to the Board, California’s
state officials lose the power to “directly supervise and
control [the Bar’s] ongoing operations” by way of removal.
62             KOHN V. STATE BAR OF CALIFORNIA


P.R. Ports Auth., 

. And unlike in Puerto Rico
Ports Authority, no government official serves in the Bar’s
leadership. In fact, the Board consists of only attorneys and
members of the public—not judges—which strongly
suggests the California Legislature intended the Bar to be
advisory. See 

, 6013.3,
6013.5. Even if government officials were on the Board, this
fact on its own is not enough to establish control. Durning,

 (acknowledging that the governor and state
treasurer serve on the entity’s board but still denying
immunity). Thus, once the Board’s members and officers
take their positions, California lacks direct control over the
Bar’s day-to-day affairs.
    True, the Board is under the supervision of the Supreme
Court of California. But supervision is not control. For
instance, unlike the States in Hess, the Supreme Court of
California does not veto the decisions of the State Bar. See
Hess, 

; see also Lake Country Ests., 440 U.S.
at 402 (noting that the State’s lack of veto power over the
entity made the entity more like a municipality). Rather, the
California court merely chooses whether to adopt the State
Bar’s recommendations as to admission and discipline. So
although the Bar reports to California’s highest court, the
court does not exercise direct control over how the Bar
operates or what recommendations it may ultimately make.
Such an advisory role cuts against immunity here. As in
Oregon, the State Bar is “not the typical government official
or agency, but rather a professional association that provides
recommendations to the ultimate regulator of the legal
profession.” Crowe, 

 (simplified).
    Nor does the State Legislature’s regulation of the State
Bar change the calculus. It should be of no surprise that the
State has the authority to regulate the State Bar. As Justice
               KOHN V. STATE BAR OF CALIFORNIA                63


O’Connor observed, “[v]irtually every enterprise, municipal
or private, flourishes in some sense at the behest of the
State.” Id. at 62 (O’Connor, J., dissenting). So the mere fact
that the State Bar is subject to California legislation does not
automatically make it an instrumentality of the State. Far
from it. Indeed, the indirect nature of legislative action over
the State Bar underscores how little control the State has
over the Bar. Unlike with state agencies, the California
Legislature does not appropriate State Bar funds. At most,
the California Legislature can cap the amount the State Bar
collects in licensing fees. See Maj. Op. 29 (citing 

). The State Bar’s counsel conceded at
argument that the Legislature might negotiate with the State
Bar through fee caps and other legislative measures to
encourage the Bar to spend its money in a certain manner.
For instance, if the State Bar wished to sell one of its
buildings, the California Legislature could express its
disapproval and threaten to cap licensing fees, but it couldn’t
outright veto the sale. This is not the same kind of “legal
control” we would expect to see the State exert over a state
agency or other instrumentality. Cf. P.R. Ports Auth., 

 (explaining that the Governor’s authority to
direct the entity to demolish infrastructure illustrates that the
entity “operates subject to the control of the Governor”).
Instead, it is the State attempting to prod an independent
institution into choosing an action under threat of legislative
retaliation. This is not the “real” and “immediate” control
required to show, “not just on paper, but also in its
operation,” that the State and the State Bar are effectively the
same. 

   The Bar is thus not subject to a level of State control that
would cloak it in sovereign immunity.
64             KOHN V. STATE BAR OF CALIFORNIA


                              C.
    Finally, we analyze the Bar’s impact on the State’s
treasury. Here, we ask whether the entity “generates its own
revenues” and whether the State bears legal liability for the
entity’s debts. Hess, 

. This factor cuts
decisively against immunity.
     It is inescapable that this suit—or any other—against the
State Bar would not impact the State’s treasury. As the
majority admits, “California law makes the State Bar
responsible for its own debts and liabilities.” Maj. Op. 29
(citing 

). And the State Bar
is tasked with raising its own funding. That the California
Legislature may impose a cap on Bar dues does not alter the
State Bar’s financial independence from the State. Simply
put, the State is not responsible for the Bar’s funding, debts,
or liabilities. If Kohn were to ultimately prevail here, neither
California nor its citizens would bear the costs of any
judgment.        Thus, this factor lands squarely against
immunity.
     The impact on the State’s treasury is a big deal even if
it’s not dispositive. While the Eleventh Amendment may
have “twin reasons for being,” Hess, 

, the
State’s solvency was the “impetus for the Eleventh
Amendment,” 

. That did not change after Hess. In
Seminole Tribe of Florida v. Florida, 

 (1996),
the Court merely observed that the “[t]he Eleventh
Amendment does not exist solely in order to prevent federal-
court judgments that must be paid out of a State’s treasury.”

 (emphasis added) (simplified). That means we must
respect the State’s sovereign immunity even in cases of
“prospective injunctive relief”—when money judgments
against the States are not at issue. 

 But in the context of
               KOHN V. STATE BAR OF CALIFORNIA             65


a suit for damages, as here, the Court has never brushed off
or minimized the importance of the treasury factor in the
sovereign immunity analysis.
    With this framing, we should readily acknowledge the
treasury factor’s import—not downplay it—in assessing
whether the Bar is an arm of the State. Given that intent and
control together represent one half of sovereign immunity’s
purpose—the State’s dignity interest—the overall effects on
the State’s treasury make up the other half. See P.R. Ports
Auth., 

. So while perhaps not dispositive, the
treasury factor must at least be treated as equally important
to the intent and control factors when combined. Whenever
intent and control together cut only weakly in favor of
immunity, the twin concern for treasury should win the day.
Here, the answer should have been even more obvious
because all the factors point the same way: no immunity.
                             II.
    California law lays out a structure for the State Bar like
an independent municipality. By creating that structure,
California has shown an intent not to clothe the State Bar
with the immunity that California enjoys. More than that,
California has treated the State Bar as a separate entity by
allowing it to operate without significant control or
direction. And finally, the State Bar’s liabilities are
independent of the State. Each of these factors strongly
points to concluding no immunity for the State Bar.
    Unfortunately, our court failed to recognize the clear
signs that California has laid out before us and thus the
majority mistakenly affords the State Bar total immunity
from suit. I would have paid due respect to the sovereign’s
wishes.
66              KOHN V. STATE BAR OF CALIFORNIA


     I respectfully dissent.