Barnett v. Hall, Estill, Hardwick, Gable

                                      PUBLISH                          FILED
                                                           United States Court of Appeals
                     UNITED STATES COURT OF APPEALS                Tenth Circuit

                           FOR THE TENTH CIRCUIT                   April 20, 2020
                       _________________________________
                                                              Christopher M. Wolpert
                                                                  Clerk of Court


CHRISTOPHER BARNETT,

      Plaintiff - Appellant/Cross-Appellee,

v.                                                   No. 18-5090

HALL, ESTILL, HARDWICK, GABLE,
GOLDEN & NELSON, P.C.; J. KEVIN
HAYES, Special Administrator of the
Estate of John Patrick Cremin;
JOHNATHAN L. ROGERS;
UNIVERSITY OF TULSA, a private
university,

      Defendants - Appellees/Cross-
      Appellants.

–––––––––––––––––––––––––––––––––––

CHRISTOPHER BARNETT,

      Plaintiff - Appellee,                          No. 18-5091

v.

HALL, ESTILL, HARDWICK, GABLE,
GOLDEN & NELSON, P.C.; J. KEVIN
HAYES, Special Administrator of the
Estate of John Patrick Cremin;
JOHNATHAN L. ROGERS,

      Defendants - Appellants,

and
 UNIVERSITY OF TULSA,

       Defendant.

 –––––––––––––––––––––––––––––––––––

 CHRISTOPHER BARNETT,
                                                             No. 18-5092
       Plaintiff - Appellee,

 v.

 UNIVERSITY OF TULSA,

       Defendant - Appellant,

 and

 HALL, ESTILL, HARDWICK, GABLE,
 GOLDEN & NELSON, P.C.; J. KEVIN
 HAYES, Special Administrator of the
 Estate of John Patrick Cremin;
 JONATHAN L. ROGERS,

       Defendants.
                        _________________________________

                     Appeal from the United States District Court
                       for the Northern District of Oklahoma
                       (D.C. No. 4:18-CV-00064-TCK-FHM)
                       _________________________________

Brendan M. McHugh, Route 66 Attorneys, L.L.C., Claremore, OK (Dana Jim, Route 66
Attorneys, L.L.C., Claremore, OK, on the briefs) on behalf of Appellant/Cross-Appellee.

John David Lackey, Tulsa, OK (Phil R. Richards, Tulsa, OK, on the briefs) on behalf of
Appellees/Cross-Appellants.
                       ________________________________

Before HARTZ, BALDOCK, and EID, Circuit Judges.
                  _________________________________


                                             2
HARTZ, Circuit Judge.
                          _________________________________

       Plaintiff Christopher Barnett appeals the judgment of the United States District Court

for the Northern District of Oklahoma dismissing with prejudice his federal civil-rights claims

for failure to state a claim and dismissing with prejudice his state-law claims because they do

not survive the restrictions imposed by the Oklahoma Citizens Participation Act (OCPA),

Okla. Stat. tit. 12, §§ 1430–40 . Defendants cross-appeal the district court’s denial of attorney

fees under the OCPA, contending that an award of attorney fees is mandatory. Exercising

jurisdiction under 

28 U.S.C. § 1291

, we affirm in part and reverse in part.

        We affirm the dismissal of the federal-law claims, agreeing with the district court that

the complaint does not adequately allege that any of the Defendants acted under color of state

law. But we reverse the judgment on the state-law claims and remand to the district court with

instructions to dismiss the claims without prejudice or remand them to the state court. Our

reversal is in keeping with the regular practice in this circuit of dismissing without prejudice

state-law claims for which the district court has only supplemental, rather than original,

jurisdiction when the federal-law claims to which they are supplemental are dismissed early in

the litigation. Because of the reversal, we have no occasion to consider the merits of

Defendants’ cross-appeal on attorney fees.

       Barnett’s complaint bases his claims on an incident on January 4, 2018, related to a

hearing in Oklahoma state court on an open-records case he had brought against Tulsa

Community College. According to Barnett, two lawyers in the firm of Hall, Estill,

Hardwick, Gable, Golden & Nelson, P.C., (Hall Estill)— namely, J. Patrick Cremin and


                                                   3
Jonathan Rogers— falsely reported to the office of the state attorney general (AG) that

Barnett had made a threat. (When, where, or to whom it was allegedly directed are not

disclosed in the complaint.) The AG’s office then relayed this report to the county sheriff.

When Barnett arrived at the courtroom for the hearing, the state-court judge instructed him

to speak with a deputy sheriff. After Barnett denied making any threat, the deputy told him

to stay inside the courtroom until he received permission to leave. At some point the AG’s

office arrived with its own security detail. When the proceedings began, the state-court

judge discussed the report in open court.

       Barnett filed suit in state court the next day against Cremin, Rogers, Hall Estill, and

Tulsa University (TU), alleging federal civil-rights claims under 

42 U.S.C. § 1983

 and state

tort claims because he had been unlawfully seized when he was forbidden to leave the

courtroom, had been cast in a false light by the public airing of the alleged threat, and had

been retaliated against by Defendants for his exercise of his rights to free speech and access

to the courts. (TU is not mentioned in the complaint except to say that Hall Estill is

representing TU in a state-court lawsuit brought against it by Barnett’s husband and that

Barnett has criticized TU in “numerous public forums.” Aplt. App. at 2.) Defendants

removed the case to federal court under 

28 U.S.C. § 1441

(a) based on the federal district

court’s original jurisdiction over claims under § 1983. The district court then had

jurisdiction over the state-law claims under the court’s supplemental jurisdiction. See 

28 U.S.C. § 1367

(a).

       Shortly after removal, Defendants moved to dismiss the federal and state claims,

and Barnett moved to recuse the district-court judge. The district court denied Barnett’s

                                                 4
motion to recuse. And it dismissed with prejudice the federal claims under Federal Rule of

Civil Procedure 12(b)(6) for failure to state a claim because the allegations of the complaint

would not support a finding that Defendants acted under color of state law.

       In the same order dismissing the federal claims, the district court applied the OCPA and

dismissed with prejudice the state claims. Before summarizing the district court’s ruling, a

further description of the state statute will be useful. The OCPA is one of a number of state

laws throughout the country that are designed to reduce the frequency of what are called

SLAPP lawsuits. (SLAPP is an acronym for “Strategic Lawsuit Against Public Participation.”

Krimbill v. Talarico, 

417 P.3d 1240, 1245

 (Okla. Civ. App. 2017)). As the Oklahoma Court of

Civil Appeals has explained: “Anti-SLAPP legislation appears to be the result of an

increasing tendency by parties with substantial resources to file meritless lawsuits against

legitimate critics, with the intent to silence those critics by burdening them with the time,

stress, and cost of a legal action. To [curb such lawsuits], anti-SLAPP acts typically

provide an accelerated dismissal procedure, available immediately after a suit is filed in

order to weed out meritless suits early in the litigation process.” Id.; see Anagnost v.

Tomecek, 

390 P.3d 707

, 709–10 (Okla. 2017) (“The stated purpose of the [OCPA] is to

encourage and safeguard the constitutional rights of persons to ‘petition, speak freely,

associate freely and otherwise participate in government to the maximum extent

permitted by law and, at the same time, protect the rights of a person to file a meritorious

lawsuit for a demonstrable injury.’ It accomplishes this goal by allowing parties to file

motions to dismiss legal actions if the legal action relates or is in response to free

speech.” (quoting 

Okla. Stat. tit. 12, § 1430

 )).

                                                    5
       Under the OCPA, if a defendant shows that the plaintiff’s claim relates to the

defendant’s exercise of the right to free speech1, to petition, or of association, then the plaintiff

must show by “clear and specific evidence a prima facie case for each essential element” of the

claim. Krimbill, 

417 P.3d at 1245

 (quoting § 1434(C)). If the plaintiff carries that burden, the

defendant can avoid further proceedings by proving “by a preponderance of the evidence” a

valid defense to the plaintiff’s claims. Id. (quoting § 1434(D)). If the plaintiff fails to show by

clear and specific evidence a prima facie case, or if the defendant proves a valid defense by a

preponderance of the evidence, the court must dismiss the plaintiff’s case. See 

Okla. Stat. tit. 12, § 1434

. The OCPA also provides for attorney fees to prevailing defendants. See 

id.

§ 1438.

       Although this court had held in Los Lobos Renewable Power, LLC v. Americulture,

Inc., 

885 F.3d 659, 673

 (10th Cir.), cert. denied, 

139 S. Ct. 591

 (2018), that the New Mexico

anti-SLAPP statute does not apply in federal court, the district court did not believe itself

bound by our precedent. The district court correctly understood Los Lobos to be “carefully

limited to the New Mexico law it addresses.” Barnett v. Hall, Estill, Hardwick, Gable, Golden

& Nelson, P.C., No. 18-cv-00064-TCK-FHM, 

2018 WL 4038117

, at *3 (N.D. Okla. Aug. 23,

2018). Our prior opinion was “based on the text of the New Mexico law and distinguish[ed] it

from other anti-SLAPP statutes . . . ‘that shift substantive burdens of proof, or alter substantive

standards.’” 

Id.

 (quoting Los Lobos, 

885 F.3d at 670

). The district court reasoned that the




1
 The statute defines Exercise of the right of free speech to mean “a communication
made in connection with a matter of public concern.” 

Okla. Stat. tit. 12, § 1431

(3).
                                                    6
OCPA was different, because it “employs . . . substantive standards.” 

Id.

 And unlike the New

Mexico courts, which have held that the New Mexico law was procedural, an Oklahoma court

has “indicate[d] that the OCPA is not merely procedural.” 

Id.

 (citing Steidley v. Cmty.

Newspaper Holdings, Inc., 

383 P.3d 780

, 785–87 (Okla. Civ. App. 2016)).

       Applying the OCPA framework, the district court first ruled that Barnett’s state-law

claims related to Defendants’ exercise of the right to free speech because the communication

alleged by the complaint (a report of a threat) was related to health or safety, which is a matter

of public concern under the OCPA. The burden then fell on Barnett to establish by clear and

specific evidence his prima facie case. The district court walked through the elements of each

of his state-law claims and concluded that he had presented no evidence to support them. It

dismissed with prejudice the state-law claims.

       I.      DISCUSSION

         We address in turn (1) the dismissal of the federal § 1983 claims, (2) the dismissal

of the state claims under the OCPA, and (3) the denial of the motion to recuse.

               A.      Section 1983 Claims

       We review de novo the dismissal of a complaint under Rule 12(b)(6). See Christy

Sports, LLC v. Deer Valley Resort Co., 

555 F.3d 1188, 1191

 (10th Cir. 2009). “In doing so,

we ask whether there is plausibility in the complaint. The complaint does not need detailed

factual allegations, but the factual allegations must be enough to raise a right to relief above the

speculative level.” 

Id.

 (citations, brackets, and internal quotation marks omitted). “Threadbare

recitals of the elements of a cause of action, supported by mere conclusory statements, do not

suffice.” Ashcroft v. Iqbal, 

556 U.S. 662, 678

 (2009).

                                                   7
       “To state a claim for relief in an action brought under § 1983, [a plaintiff] must

establish that [he was] deprived of a right secured by the Constitution or laws of the United

States, and that the alleged deprivation was committed under color of state law.” Am. Mfrs.

Mut. Ins. Co. v. Sullivan, 

526 U.S. 40

, 49–50 (1999). Missing from Barnett’s complaint

were allegations sufficient to establish action by Defendants under color of state law.

“[T]he under-color-of-state-law element of § 1983 excludes from its reach merely private

conduct, no matter how discriminatory or wrongful.” Id. at 50 (internal quotation marks

omitted). “Where a litigant seeks to hold a private actor accountable as a state actor for

constitutional deprivations, we have applied various analyses and referred to them as the

‘nexus test,’ the ‘public function test,’ the ‘joint action test,’ and the ‘symbiotic relationship

test.’” Wittner v. Banner Health, 

720 F.3d 770, 775

 (10th Cir. 2013). At the heart of each

test is “whether the conduct allegedly causing the deprivation of a federal right [is] fairly

attributable to the State.” Wasatch Equal. v. Alta Ski Lifts Co., 

820 F.3d 381, 387

 (10th

Cir. 2016) (internal quotation marks omitted).

       Defendants are all private actors. Barnett argues that they nevertheless acted under

color of state law because their actions satisfied the tests for joint action and symbiotic

relationship with the AG’s office. We disagree.

       The first test is whether the private party was “a willful participant in joint action

with the State or its agents.” Gallagher, 49 F.3d at 1453. Barnett contends that his

complaint establishes joint action because it alleges that Defendants conspired with state

officers. See Sigmon v. CommunityCare HMO, Inc., 

234 F.3d 1121, 1126

 (10th Cir.

2000) (“[O]ne way to prove willful joint action is to demonstrate that the public and

                                                  8
private actors engaged in a conspiracy.”). “When a plaintiff seeks to prove state action

based on a conspiracy theory, a requirement of the joint action charge is that both public

and private actors share a common, unconstitutional goal.” 

Id.

 (ellipsis and internal

quotation marks omitted).

       But Barnett failed to adequately allege a conspiracy. His claim of a conspiracy is

conclusory, devoid of allegations of specific facts in support. See Beedle v. Wilson, 

422 F.3d 1059, 1073

 (10th Cir. 2005) (“[T]he plaintiff must specifically plead facts tending to

show agreement and concerted action.” (internal quotation marks omitted)). All he alleges

is that Defendants filed a false threat report with officials in the AG’s office. But

“furnishing information to law enforcement officers, without more, does not constitute

joint action under color of state law.” Schaffer v. Salt Lake City Corp., 

814 F.3d 1151, 1157

 (10th Cir. 2016). “Rather, joint action arises only when an officer’s decision to

initiate an arrest or a prosecution was not independent of a private party’s influence.” 

Id.

The complaint contains no factual allegations to support a plausible inference that the AG’s

office acted under Defendants’ influence beyond their being a source of information.

       Barnett’s argument under the symbiotic-relationship test fares no better. That test

has rarely been applied, see Gallagher, 49 F.3d at 1451, and has been limited to

substantial, intimate long-term relationships, see Burton v. Wilmington Parking

Authority, 

365 U.S. 715, 725

 (1961) (public parking authority that leased space to

segregated restaurant on which the parking facility was financially dependent had “so far

insinuated itself into a position of interdependence with [the restaurant] that it must be

recognized as a joint participant in the [racially discriminatory] activity”); Milo v.

                                                 9
Cushing Municipal Hosp., 

861 F.2d 1194, 1196

 (10th Cir. 1988) (city “substantially

participated in the funding, creation, and financial structure of the Hospital,” and three

city commissioners were trustees that oversaw the hospital).

       The situation alleged here is a far cry from what has been held to be a symbiotic

relationship. There are no factual allegations regarding any past relationship whatsoever

between Defendants and the AG’s office; and there is certainly no indication that either is

dependent on the other. We reject Barnett’s suggestion that the isolated incident

described in the complaint could establish a symbiotic relationship.

       To salvage his claims, Barnett makes two cursory procedural arguments. He says

that the district court erred by (1) denying him leave to amend his complaint, and (2) sua

sponte dismissing the federal claims as to Hall Estill. Both can be disposed of quickly.

The first fails on two grounds. One is that he did not request leave to amend at any time in

district court. See Burnett v. Mortg. Elec. Registration Sys., Inc., 

706 F.3d 1231

, 1238 n.4

(10th Cir. 2013) (“Where a plaintiff does not move for permission to amend the complaint,

the district court commits no error by not granting such leave.”). He excuses his failure on

the ground that he was awaiting a ruling on the motion to dismiss before trying to cure any

possible defects. But cases are not to be litigated piecemeal. The court should not have to

address repeated “improvements” to the complaint. When a party faces a motion to

dismiss and it believes that it can overcome objections with an amendment to the pleading,

it should seek leave to amend at that time. Efficient adjudication of disputes requires that

the party present its best effort to state a claim before the court addresses the motion to

dismiss. See Goldstein v. MCI WorldCom, 

340 F.3d 238, 254

 (5th Cir. 2003) (“[W]e have

                                                 10
upheld the denial of leave to amend when the moving party . . . attempted to present

theories of recovery seriatim to the district court.”); Pallottino v. City of Rio Rancho, 

31 F.3d 1023, 1027

 (10th Cir. 1994) (similar). The second ground for rejecting his argument

that he should be allowed to amend his complaint is that Barnett has not explained on

appeal “how a proposed amendment would cure the deficiencies identified by the district

court.” Hall v. Witteman, 

584 F.3d 859, 868

 (10th Cir. 2009).

       Similarly, we will not reverse the sua sponte dismissal of the federal claims against

Hall Estill because Barnett offers no argument to support the claims, or possible amended

claims, against the firm. See McKinney v. Okla. Dep’t of Human Servs., 

925 F.2d 363

, 365

(10th Cir. 1991) (sua sponte dismissal “is not reversible error when it is patently obvious

that the plaintiff could not prevail on the facts alleged . . . and allowing him an opportunity

to amend his complaint would be futile” (internal quotation marks omitted)).

       We therefore affirm the district court’s dismissal of Barnett’s § 1983 claims against

all Defendants.

              B. State Claims

       The district court dismissed Barnett’s state claims under the dismissal procedure of

the OCPA. It applied that state statute because it saw no conflict between it and the

Federal Rules of Civil Procedure for two reasons: First, it said that “the OCPA is a

statement of the substantive policy of the State of Oklahoma,” whereas the Federal Rules

“do not provide any policy goals, nor any burden shifting and changes to substantive

standards to enact these goals.” Barnett, 

2018 WL 4038117

 at *4. Second, it said that a



                                                11
motion to dismiss under the OCPA “will not interfere with the operation” of Federal Rules

of Civil Procedure 8, 12, or 56. 

Id.

       We are not so sure. We recognize that our decision in Los Lobos, 

885 F.3d at 673

n.8, which rejected the application in federal court of the New Mexico anti-SLAPP

statute, is not controlling. In that case we simply held that the Federal Rules governed

because the state law was strictly procedural (merely requiring that standard procedures

be expedited). See 

id.

 at 668–72. Nevertheless, we did say that it was “very much

debatable” whether the New Mexico statute could operate alongside the Federal Rules

without conflict. 

Id.

 at 673 n.8. 2

       More closely in point is the persuasive reasoning of the D.C. Circuit in an opinion

by then-Judge Kavanaugh holding that the District of Columbia anti-SLAPP statute could

not apply in federal district court. See Abbas v. Foreign Policy Grp., LLC, 

783 F.3d 1328

(D.C. Cir. 2015). Relying on Shady Grove Orthopedic Associates, P.A. v. Allstate

Insurance Co., 

559 U.S. 393

 (2010), the circuit court adopted a more expansive view than

that expressed by the district court here of when state law conflicts with a Federal Rule. It




2
  Although the federal court’s jurisdiction over the state-law claims in Los Lobos was
based on diversity of citizenship whereas here the federal court had supplemental
jurisdiction over the state-law claims, we apply in both circumstances the same analysis
in determining whether federal or state law governs. See BancOklahoma Mortg. Corp. v.
Capital Title Co., 

194 F.3d 1089, 1103

 (10th Cir. 1999) (The rule in diversity cases that
the federal court “applies the substantive law . . . of the forum state . . . also applies when
a federal court exercises supplemental jurisdiction over state law claims in a federal
question lawsuit.”) (internal quotation marks and citation omitted)).


                                                12
read the portion of the lead opinion in Shady Grove that was joined by a majority of the

Supreme Court 3 as stating that a valid Federal Rule prevails over state law if the Federal

Rule “answers the same question as the state law or rule.” Abbas, 

783 F.3d at 1333

(brackets and internal quotation marks omitted). The opinion goes on to say that “Federal

Rules of Civil Procedure 12 and 56 ‘answer the same question’ [as the D.C. statute] about

the circumstances under which a court must dismiss a case before trial. . . . Under the

Federal Rules, a plaintiff is generally entitled to trial if he or she meets the Rules 12 and 56

standards to overcome a motion to dismiss or for summary judgment. But the D.C. Anti–

SLAPP Act nullifies that entitlement in certain cases.” 

Id.

 at 1333–34.

       Perhaps there is in practice no real difference between the OCPA and the Federal

Rules because the same analysis is to be used for both application of the OCPA dismissal

standards and the Federal Rules for dismissal for failure to state a claim and summary

judgment. 4 If there is a difference, however, an analysis under the Federal Rules Enabling

Act, 

28 U.S.C. § 2072

, would be necessary to determine whether the Federal Rules could




3
  Justice Stevens wrote a concurring opinion. His opinion joined the plurality opinion of
four other Justices on this issue but differed on how to analyze whether the Rules
Enabling Act limits the application of a Federal Rule.
4
  In Krimbill, 

417 P.3d at 1247

, the Oklahoma Court of Civil Appeals said that “disputed
questions of material fact cannot be resolved in an OCPA dismissal proceeding,” suggesting
that there may be no significant difference between the Oklahoma statutory procedures and the
Federal Rules. See Abbas, 

783 F.3d at 1335

 n.3 (noting the possibility that “if a State anti-
SLAPP act did in fact exactly mirror Federal Rules 12 and 56,” it may not be preempted by the
Federal Rules.)

                                                 13
be applied. 5 See Abbas, 

783 F.3d at 1333

 (“A federal court exercising diversity jurisdiction

should not apply a state law or rule if (1) a Federal Rules of Civil Procedure answers the

same question as the state law or rule and (2) the Federal Rule does not violate the Rules

Enabling Act.” (brackets and internal quotation marks omitted)). But we need not decide

that issue because Barnett’s state-law claims come before us only under the district court’s

supplemental jurisdiction.

       The Supreme Court has encouraged the practice of dismissing state claims or

remanding them to state court when the federal claims to which they are supplemental have

dropped out before trial. See, e.g., Carnegie–Mellon Univ. v. Cohill, 

484 U.S. 343

, 350 n.7

(1988) (“[I]n the usual case in which all federal-law claims are eliminated before trial, the

balance of factors to be considered under the pendent jurisdiction doctrine—judicial economy,

convenience, fairness, and comity—will point toward declining to exercise jurisdiction over




5
  Abbas concluded that “Rules 12 and 56 are valid under the Rules Enabling Act”
because “pleading standards and rules governing motions for summary judgment are
procedural.” 

783 F.3d at 1337

. We should note, however, that it decided that the
concurring opinion of Justice Stevens in Shady Grove did not create Supreme Court law
on how to determine whether a federal rule is authorized by the Rules Enabling Act and
instead followed Sibbach v. Wilson & Co., 

312 U.S. 1

 (1941). But we are bound by
Tenth Circuit precedent which has said that Justice Stevens’s opinion controls the
analysis on that question. See Los Lobos, 

885 F.3d at 668

 n.3; Garman v. Campbell Cty.
Sch. Dist. No. 1, 

630 F.3d 977

, 983 n.6 (10th Cir. 2010).
        Of course, state anti-SLAPP statutes differ in their particulars, and the analyses of
different circuits do not always agree. Several circuits have applied such state statutes in
federal court. See Godin v. Schencks, 

629 F.3d 79, 89

 (1st Cir. 2010) (Maine’s anti-SLAPP
statute); United States ex rel. Newsham v. Lockheed Missiles & Space Co., 

190 F.3d 963, 973

(9th Cir. 1999) (California’s statute); cf. Henry v. Lake Charles Am. Press, L.L.C., 

566 F.3d 164, 168

 (5th Cir. 2009) (applying Louisiana’s statute without analysis).
                                                  14
the remaining state-law claims”); 

id.

 (stating that in a case removed from state court the district

court had discretion to either dismiss the state-law claims or remand them to state court);

United Mine Workers of Am. v. Gibbs, 

383 U.S. 715, 726

 (1966) (“Certainly, if the federal

claims are dismissed before trial, even though not insubstantial in a jurisdictional sense, the

state claims should be dismissed as well.”). This court has followed suit when the claims over

which the district court had original jurisdiction are dismissed on pretrial motion, especially

when the state claims raise novel issues of state law. See VR Acquisitions, LLC v. Wasatch

Cty., 

853 F.3d 1142, 1149

 (10th Cir. 2017) (reversing district court’s dismissal with

prejudice of state-law claims and remanding for dismissal without prejudice); Merrifield

v. Bd. of Cty. Comm’rs for Cty. of Santa Fe, 

654 F.3d 1073, 1086

 (10th Cir. 2011) (reversing

district court’s grant of summary judgment on state claim because “the interest in comity—

leaving to the states to decide novel questions of state-law—clearly predominates here.”);

Brooks v. Gaenzle, 

614 F.3d 1213, 1230

 (10th Cir. 2010) (reversing district court’s grant of

summary judgment on state claim because disputed issue of Colorado law was “best left for a

state court’s determination.”); Ball v. Renner, 

54 F.3d 664, 669

 (10th Cir. 1995) (reversing

summary judgment “[w]here a state law cause of action is thus in a process of current

evolution, it is particularly appropriate for the federal courts to leave the continuing

development and application of that cause of action to the state courts.”).

       In this case the district court resolved the federal-law claims at a very early stage

of the litigation. This in itself counsels dismissal without prejudice of the state-law

claims at the same time. In addition, there is a novel issue of state law regarding

attorney-fee awards under the OCPA. The OCPA provides that “the court shall award to

                                                  15
the moving party . . . [c]ourt costs, reasonable attorney fees and other expenses incurred in

defending against the legal action as justice and equity may require.” 

Okla. Stat. tit. 12, § 1438

(A) (emphasis added). Oklahoma courts have not addressed whether the justice-and-

equity clause modifies all items in the series (court costs, reasonable attorney fees, and

other expenses) or only the last item (other expenses); 6 and neither did the district court

when it declined to award attorney fees on the ground of justice and equity. Thus, comity

considerations add an additional thumb on the scale in favor of dismissal. Further, we

note that the federal-law issue of whether to apply anti-SLAPP statutes like the OCPA in

federal court is a challenging one and has divided the circuits. It makes sense to wait to

decide the issue until we must do so, perhaps after helpful development of both federal

law and Oklahoma case law interpreting the statute.

       We therefore reverse the dismissal of the state claims under the OCPA and remand

with instructions to dismiss them without prejudice or remand them to state court.

Because of our disposition of this issue, we need not address Barnett’s remaining

challenges to the OCPA.

               C.     Recusal

       Barnett challenges the district judge’s denial of his motion to recuse. A district

judge “shall disqualify himself in any proceeding in which his impartiality might



6
  The OCPA was modeled on the Texas anti-SLAPP statute. See Krimbill, 

417 P.3d at 1244

 (“Oklahoma’s Act, which became effective in 2014, mirrors that of the Texas
Citizens’ Participation Act.”). And the Texas Supreme Court has interpreted the Texas fee
provision, which contains the same text as the OCPA, to “limit the justice-and-equity
modifier to other expenses.” Sullivan v. Abraham, 

488 S.W.3d 294, 299

 (Tex. 2016).
                                                  16
reasonably be questioned.” 

28 U.S.C. § 455

(a). “[W]hat matters is not the reality of bias

or prejudice but its appearance.” Liteky v. United States, 

510 U.S. 540, 548

 (1994).

Thus, “[t]he test under § 455(a) is not whether the judge believes he or she is capable of

impartiality.” Burke v. Regalado, 

935 F.3d 960, 1054

 (10th Cir. 2019). Rather, we ask

whether a reasonable person, fully informed of the relevant facts, would “question the

judge’s impartiality.” 

Id.

 We review for abuse of discretion the denial of a motion to

disqualify. See 

id. at 1052

.

       On appeal Barnett argues two grounds for recusal. First, he contends that financial

contributions made by the district judge’s spouse to TU, her alma mater, created an

appearance of impropriety. The district judge had been previously asked to recuse

himself in an unrelated case, see Ross v. Univ. of Tulsa, No. 14-CV-00484-TCK-PJC,

2017 U.S. Dist. LEXIS 38650

, at *1 (N.D. Okla. Mar. 17, 2017), where he explained the

nature of the contributions. In 2011 his spouse purchased a life-insurance policy for

$16,000. Under the policy TU would receive $100,000 upon her death to endow a

scholarship for student athletes in her name. In its announcement of the scholarship, TU

attributed the contributions to both the district judge and his spouse. The policy,

however, was purchased with her separate funds, and the donation does not appear on the

district judge’s tax returns.

       In our view, the district judge properly rejected the motion for recusal. Three

features of the contribution combine to compel this conclusion. To begin with, this is not

a case where the judge (or a family member) has a financial interest in the litigation. The

contribution was not an investment. The contribution could imply bias only by showing

                                               17
enthusiasm for TU. Courts have repeatedly rejected claims that similar actions reflecting

enthusiastic loyalty to a university in themselves require recusal in litigation involving

the institution. See, e.g., Roe v. St. Louis Univ., 

746 F.3d 874, 886

 (8th Cir. 2014)

(“Alumni connections are not a reasonable basis for questioning a judge’s impartiality,

even if alumni contribute financially or participate in educational activities.”); Chalenor v.

Univ. of N.D., 

291 F.3d 1042

, 1049–50 (8th Cir. 2002) (stating, in Title IX suit by student

wrestlers against university, that “[b]oth the fact that Judge Webb is an alumnus of the

University and the fact that he has contributed financially to the University are immaterial,

unless the facts indicated a specific and particular interest in the wrestling program or some

other particularly relevant problem.”); U.S. ex rel. Hochman v. Nackman, 

145 F.3d 1069

,

1076 (9th Cir. 1998) (“The judge’s main contact with USC was a small yearly contribution

to the law school’s alumni association. This insignificant contact would not reasonably

lead one to question his impartiality, in particular because USC is not a named party and

because the relationship between this case and USC’s law school is virtually

nonexistent.”); Lunde v. Helms, 

29 F.3d 367, 371

 (8th Cir. 1994) (in sex-discrimination suit

in connection with dismissal from medical school, court said that it did “not think that

making alumni contributions or participating in university educational programs, without

more, is a reasonable basis for questioning the judge’s impartiality.”); Wu v. Thomas, 

996 F.2d 271, 272, 275

 (11th Cir. 1993) (in retaliation claim brought by professors against

university, the judge’s “status as an [unsalaried] adjunct professor and his past

contributions to the University” did not warrant recusal). Indeed, a decision by this circuit

held that a judge is not disqualified from hearing a case involving a university despite the

                                                18
judge’s being a director of that university’s alumni association. See Willner v. Univ. of

Kan., 

848 F.2d 1023, 1026, 1028

 (10th Cir. 1988). It is not at all clear to us that a $16,000

contribution to TU shows a stronger attachment than being a director of the alumni

association.

       The second important feature of the contribution is that it is not specifically tied to

activity related to the litigation. The contribution was for scholarships for student athletes.

The litigation has no apparent connection to TU’s athletic program. See Chalenor, 291

F.3d at 1049–50 (in Title IX suit by student wrestlers against university, judge’s ties to

university were “immaterial, unless the facts indicated a specific and particular interest in

the wrestling program or some other particularly relevant problem”); Nackman, 145 F.3d at

1070, 1076 (in suit against physicians at university’s school of medicine, judge’s

connections were with law school, whose relationship to the case was “virtually

nonexistent”). This is quite different from contributing to an advocacy organization that

presents that advocacy in litigation before the judge. Cf. Guide to Judiciary Policy, Vol.

2B, Ch. 2, Published Advisory Opinion No. 40 (2009) (acknowledging the propriety of

judicial participation in nonprofit civic, charitable, education, religious and social

organizations, but advising that a judge should not participate in organizations “if it is

likely that the organization will either be engaged in proceedings that would ordinarily

come before the judge or will be regularly engaged in adversary proceedings in any

court.”).

       Finally, and perhaps most importantly, the contribution was distinctly made by the

judge’s spouse, not the judge himself. The enthusiasm was that of the spouse and should

                                                 19
not be attributed to the judge. In present-day society we do not treat a married couple as

single-minded on public issues. See Perry v. Schwarzenegger, 

630 F.3d 909, 912

 (9th Cir.

2011) (concluding that motion for recusal based on wife’s opinions as executive director of

the American Civil Liberties Union of Southern California is “based upon an outmoded

conception of the relationship between spouses. . . . Because my wife is an independent

woman, I cannot accept Proponents’ position that my impartiality might reasonably be

questioned under § 455(a) because of her opinions or the views of the organization she

heads.”). We cannot conclude that the spouse’s support for TU, without more, would

allow a reasonable person to question the judge’s impartiality. Cf. ClearOne Commc’ns,

Inc. v. Bowers, 

643 F.3d 735

, 776–77 (10th Cir. 2011) (no recusal required where judge’s

spouse was of counsel to law firm that had done unrelated work for litigant and spouse had

not worked for litigant); Nachshin v. AOL, LLC, 

663 F.3d 1034, 1041

 (9th Cir. 2011)

(judge’s husband’s serving as unpaid board member of nonprofit that stood to gain $25,000

as proposed cy pres beneficiary from litigation did not require recusal; the spouse would

not benefit “in any way other than to enable [the nonprofit] to continue providing legal

services to the indigent”).

       Barnett’s second argument for recusal is that the district judge was biased against

him after being apprised by Defendants of his abusive and threatening behavior toward the

court. Ordinarily, however, recusal for bias cannot be based on a judge’s reaction, or

potential reaction, to events during the course of the litigation. Thus, it is well settled that

“a motion to recuse cannot be based solely on adverse rulings.” Willner, 

848 F.2d at 1028

.

But the principle extends beyond that specific application. For example, a judge is not

                                                 20
required to recuse himself because of “baseless personal attacks on or suits against the

judge by a party,” or “threats or other attempts to intimidate the judge.” United States v.

Cooley, 

1 F.3d 985, 994

 (10th Cir. 1993). As the Supreme Court has recognized: “The

judge who presides at a trial may, upon completion of the evidence, be exceedingly ill

disposed towards the defendant, who has been shown to be a thoroughly reprehensible

person.” Liteky, 510 U.S. at 550–51. Nevertheless, “opinions formed by the judge on the

basis of facts introduced or events occurring in the course of the current proceedings, or

of prior proceedings, do not constitute a basis for a bias or partiality motion unless they

display a deep-seated favoritism or antagonism that would make fair judgment

impossible.” Id. at 555.

       In this case, there is no reason to think that the district judge was influenced in his

merits rulings by the reports of Barnett’s misconduct communicated to the court by

Defendants’ counsel or by a totally inappropriate vulgar message left by Barnett on the

court’s voicemail. The district judge’s order rejecting the recusal motion was a model of

propriety. The district judge did not abuse his discretion in declining to recuse himself.

       II.    CONCLUSION

       We REVERSE the dismissal of the state-law claims and REMAND them with

instructions either to remand them to state court or to dismiss them without prejudice.

We DENY as moot Defendants’ cross-appeal request for attorney fees, costs, expenses,

and sanctions under the OCPA.




                                                 21