Roe v. St. John's University

21-1125
Roe v. St. John’s University

UNITED STATES COURT OF APPEALS
FOR THE SECOND CIRCUIT
August Term, 2021
(Argued: April 12, 2022 Decided: January 31, 2024)
Docket No. 21-1125

RICHARD ROE,
Plaintiff-Counter-Defendant-Appellant,

v.

ST. JOHN’S UNIVERSITY,
Defendant-Appellee,

JANE DOE,
Defendant-Counter-Claimant-Appellee.

Before: SACK, PARKER, AND MENASHI, Circuit Judges. *

St. John’s University disciplined male plaintiff “Richard Roe” for allegedly
sexually assaulting two women—“Jane Doe” and “Mary Smith”—on separate
occasions. Roe brought suit against St. John’s University in the United States
District Court for the Eastern District of New York, alleging that it violated his
rights under Title IX of the Education Amendments of 1972 and under state
contract law. Roe also sued Doe under state law for allegedly defaming him in
an anonymous tweet that accused Roe of sexual assault. The district court
(Pamela K. Chen, J.) granted St. John’s University’s motion to dismiss Roe’s suit
under Federal Rule of Civil Procedure 12(b)(6), concluding that Roe had failed to
state a Title IX claim against St. John’s University. The district court also

* Judge Amalya L. Kearse, originally a member of the panel, recused herself from this case after
it had been argued. Judge Barrington D. Parker, chosen at random, was subsequently added to
the panel in her stead.
21-1125
Roe v. St. John’s University

declined to exercise supplemental jurisdiction over Roe’s state-law breach of
contract and defamation claims. Roe timely appealed, arguing that he had
adequately pleaded facts raising a minimal plausible inference of sex
discrimination in St. John’s University’s disciplinary procedures. For the reasons
set forth below, we disagree. We therefore
AFFIRM the judgment of the district court.

Judge Parker concurs in a separate opinion.

Judge Menashi dissents in a separate opinion.

PETER G. EIKENBERRY (Michael Valentine, on
the brief, The Law Office of Michael
Valentine, Brooklyn, NY), Law Office of
Peter G. Eikenberry, New York, NY, for
Plaintiff-Counter-Defendant-Appellant;

LYLE S. ZUCKERMAN (Michael J. Goettig, on
the brief), Davis Wright Tremaine LLP, New
York, NY, for Defendant-Appellee;

CHARDAIE C. CHARLEMAGNE (Victoria L.
Stork, on the brief, Baker & Hostetler LLP,
New York, NY), Baker & Hostetler LLP,
San Francisco, CA, for Defendant-Counter-
Claimant-Appellee.

SACK, Circuit Judge:

INTRODUCTION

St. John’s University (“SJU”), headquartered in the New York City

borough of Queens, disciplined male plaintiff Richard Roe for allegedly sexually

2
21-1125
Roe v. St. John’s University

assaulting two women, Jane Doe and Mary Smith, 1 in different countries, several

months apart. Roe claims that anti-male bias influenced SJU’s adjudication of the

accusations against him, and based thereon, brought suit against SJU in the

United States District Court for the Eastern District of New York for violating his

rights under Title IX of the Education Amendments of 1972 (“Title IX”), 20 U.S.C.

§ 1681 et seq., and under state contract law. Roe also sued Doe for defamation

based on an alleged anonymous tweet that falsely accused Roe of sexual assault.

Roe further contends that SJU’s failure to adequately investigate his claims

regarding the tweet subjected SJU to liability under Title IX and for breach of

contract.

SJU moved to dismiss Roe’s complaint pursuant to Federal Rule of Civil

Procedure 12(b)(6). The district court (Pamela K. Chen, J.) granted SJU’s motion

and dismissed Roe’s suit because, the court concluded, Roe had failed to allege

sufficient facts to support a minimal plausible inference of sex-based 2

discrimination by SJU. The court also declined to exercise supplemental

jurisdiction over Roe’s state-law breach of contract and defamation claims. Roe

1“Richard Roe,” “Jane Doe,” and “Mary Smith” are pseudonyms adopted for the purposes of
this litigation.
2 This opinion uses the terms “sex” and “gender” interchangeably.

3
21-1125
Roe v. St. John’s University

timely appealed, arguing principally that the district court’s dismissal of his Title

IX claims was erroneous. For the reasons set forth below, we disagree and

therefore affirm the judgment of the district court.

We emphasize at the outset that our focus is on the decisions made and

related actions taken by university officials patrolling the behavior of the

university’s students. Despite their operation as a panel in a manner much like a

federal or state court adjudicating a case or controversy, the officials did not

constitute a court of law. Moreover, while this case’s facts involve charges of

highly offensive sexual predation by a male student against two female fellow

students, we express no opinion as to whether the university selected appropriate

punishments for the male student’s alleged transgressions. Instead, our review

is principally concerned with whether the district court correctly concluded that

the male student failed to plausibly allege that anti-male bias influenced the

university’s disciplinary process in violation of his rights under Title IX.

4
21-1125
Roe v. St. John’s University

BACKGROUND

I. Factual Background

A. The Doe Incident

According to the allegations in Richard Roe’s complaint in the district

court: On April 12, 2018, Roe and Jane Doe were both SJU students studying in

Paris, France, lodging in an SJU dormitory there. On their first night in the city,

Roe and Doe visited a local club with several other SJU students. Doe asked Roe

to dance with her, and he agreed. “[W]hen Roe was back in the dorm, he went to

Doe’s room and found her awake and on her phone.” JA 12. “Doe invited Roe

into her room and, thereafter, Doe took Roe’s right hand and placed it upon her

fully clothed breast.” Id. Roe “immediately said, ‘I am not interested in sex,’”

and Doe responded, “[t]hen, get the hell out of here.” Id. Roe then left. 3

B. The Doe Incident’s Aftermath

Doe submitted a complaint to SJU accusing Roe of sexual misconduct. SJU

notified Roe of Doe’s complaint on September 4, 2018, some five months after the

3Inasmuch as the issue before us is whether Roe’s complaint filed in the district court
“contain[s] sufficient factual matter, accepted as true,” to state a plausible claim for relief,
Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009), Doe’s allegations about Roe’s misconduct have, at
most, limited relevance to the questions at hand.

5
21-1125
Roe v. St. John’s University

events allegedly occurred. The following month, on October 3, 2018, SJU’s

“Conduct Board” held a “Conduct Hearing” with respect to Doe’s allegations.

Roe attended.

Although Roe asserts that SJU was required to follow its own internal

policies and procedures when determining the validity of Doe’s complaint

against him, Roe has not alleged that this Conduct Hearing was required to

provide him with all the procedural protections that he would receive if his case

were tried in a federal court. Indeed, SJU’s Student Code of Conduct and

Conduct Process (“Student Code of Conduct”) states that the applicable

evidentiary standard in a sexual misconduct case is a preponderance of the

evidence, a much lower burden of proof than the beyond a reasonable doubt

standard that would apply in a federal criminal case. 4 The Student Code of

Conduct also states that “[s]trict conformity to the legal rules of evidence shall

not be required at hearings.” JA 227.

4Roe did not include SJU’s Student Code of Conduct as an exhibit to his complaint, but SJU did
include the document as an exhibit to its motion to dismiss. The district court nonetheless
concluded that it could consider the Student Code of Conduct without converting SJU’s motion
to dismiss into one for summary judgment because Roe heavily relied upon the document in his
complaint and because the Student Code of Conduct is publicly available on SJU’s website. Roe
v. St. John’s Univ., No. 19-CV-4694 (PKC) (RER), 2021 WL 1224895, at *14 (E.D.N.Y. Mar. 31,
2021). Roe has given us no reason to question that conclusion, and Roe himself quotes from the
Student Code of Conduct in his appellate briefs.

6
21-1125
Roe v. St. John’s University

At the October 2018 Conduct Hearing, Roe stood accused of multiple

charges in connection with the events of April 12. But the Conduct Board

concluded that he had violated only one applicable Student Code of Conduct

provision: that prohibiting non-consensual sexual contact. The Conduct Board

sent Roe a letter informing him of this finding and notifying him that the Student

Conduct Administrator had determined that Roe would receive, in addition to

other sanctions, a one-semester suspension for his misconduct.

The Conduct Board’s stated rationale for its misconduct finding was that:

[Roe] admitted in [sic] engaging in physical contact of a sexual nature
with [Doe], and the evidence demonstrated a lack of affirmative
consent to engage in such contact. Such evidence included [Doe’s]
intoxication, as described by multiple witnesses, and [Roe’s]
assertion, which was not disputed, that he was not impaired by
alcohol.

JA 80–81. 5 Roe appealed his suspension to the SJU administration’s “Conduct

Appeals Board” on November 7, 2018. The Appeals Board affirmed the decision

and suspension on January 8, 2019, reasoning in part that “[Roe] had admitted to

5Exhibit F in support of SJU’s motion to dismiss is an incident adjudication report that contains
handwritten notes from the October 2018 Conduct Hearing that provide additional
explanations for the Conduct Board’s decision. We conclude that consideration of Exhibit F was
not necessary to resolve this case and therefore do not reach SJU’s argument that the district
court erred by excluding it.

7
21-1125
Roe v. St. John’s University

the physical touching at issue here and, thus, should have expected a sanction of

some kind.” JA 82.

C. The Smith Incident

According to Roe’s complaint, another woman, Mary Smith, accused Roe

of sexually assaulting her in December 2018, while Roe was serving his

suspension based on the Doe allegations. Roe alleges in his complaint that on the

night of December 15, 2018, Roe and Smith were both at a local bar in New York

City where Smith became intoxicated. After the bar closed, a friend of Roe’s

agreed to drive Smith back to Roe’s house. A short time after arriving there,

Smith went outside to sleep on a concrete patio adjacent to the house. Roe

invited Smith to sleep on a couch in the living room instead because it was

raining and near freezing outside. Smith accepted. Because Roe was concerned

that Smith’s intoxication might cause her to vomit, “he propped her up on her

side with her back against the sofa to ensure that she could not choke.” JA 28.

Roe then fell asleep at the foot of the couch, and Smith left Roe’s home the next

morning without incident. 6

6As with the discussion of the Doe incident above, inasmuch as the issue presented in
connection with the motion to dismiss Roe’s complaint is whether Roe’s complaint “contain[s]
sufficient factual matter, accepted as true,” to state a plausible claim for relief, Iqbal, 556 U.S. at

8
21-1125
Roe v. St. John’s University

D. The Tweets

Roe alleges in his complaint that, on January 4, 2019, shortly after Smith

spent the night at Roe’s home during the previous month, the hashtag

“#SurvivingSJU” was created, by whom we do not know. Roe further alleges

that more than 2,000 tweets were posted under the hashtag on January 4 and

that, as a result, it was the “number one hashtag” trending on Twitter in the

United States from January 4 to January 6, 2019. JA 15, 17. SJU’s student

newspaper reported that in response to the hashtag and the public criticism that

the tweets conveyed, SJU would investigate all claims posted under

#SurvivingSJU.

Roe contends that he was the subject of one among the thousands of tweets

published under the hashtag #SurvivingSJU. The tweet, which was posted by an

anonymous user on January 4, the same day that the hashtag was created,

included a picture of Roe and contained the message “[Roe] was allowed to stay

abroad after raping me with no travel restrictions. Only got half a semester

suspension.” JA 72.

678, the complete details of Smith’s allegations against Roe in response are not relevant to this
discussion.

9
21-1125
Roe v. St. John’s University

Roe asserts that “Doe almost instantly ‘liked’ the anonymously posted

tweet, and less than 30 minutes later, posted in another tweet from her personal

account that she had written a ‘final statement’ for the University Conduct

Hearing . . . instead of ‘writing essays’ for her classes.” JA 15–16. Roe claims that

“only Doe, Roe and [SJU] officials were informed of Roe’s suspension and the

reason for it” and that “the only person who could have authored the [allegedly

defamatory] tweet was Doe.” JA 16. Roe asserts that shortly after the tweet was

published, a male SJU student threatened him via phone calls and text messages

and a female SJU student punched him in the face at a bar.

On January 7, 2019, Roe’s attorney brought the tweet to the attention of

SJU officials. On January 16, SJU’s Director of Title IX Compliance informed Roe

that SJU could not reprimand the individual who posted the tweet because it

could not determine his or her identity. On July 18, 2019, Roe’s attorney again

contacted an SJU official in an attempt to determine how SJU would respond to

the allegedly defamatory and harassing tweet. SJU’s attorney denied that SJU

had any obligation to investigate Roe’s claim that Doe was the tweet’s author

because SJU did not have any evidence implicating Doe.

10
21-1125
Roe v. St. John’s University

E. The Smith Incident’s Aftermath

On January 5, 2019—the day after the creation of #SurvivingSJU and the

dissemination of the tweet accusing Roe of sexual assault—Smith filed a

complaint with SJU against Roe accusing him of sexually assaulting her during

the preceding month. On January 7, SJU ordered Roe and Smith to refrain from

contacting one another. 7 The next day, SJU suspended Roe pending further

investigation. SJU officially notified Roe of Smith’s complaint on June 6, 2019,

and scheduled a Conduct Hearing for September 17, 2019, to review Smith’s

accusations.

Thomas Foley, an SJU attorney, chaired the panel that conducted the

September 17 hearing. Foley had also chaired the panel that conducted the

hearing regarding Doe’s allegations against Roe in October of the previous year.

Following the September 17 hearing, the panel concluded that Roe had violated

SJU’s policy governing non-consensual sexual contact but not SJU’s policy

governing non-consensual sexual penetration. The panel noted that the evidence

supporting its determination included “key inconsistencies in [Roe’s] testimony”

and “accounts of [Roe’s] housemates.” JA 143. The panel also concluded that

7 SJU sent Roe another no contact order on January 8.

11
21-1125
Roe v. St. John’s University

“key aspects of [Roe’s] account of the events lacked credibility, including his

testimony that he turned [Smith’s] face toward the couch in the event she

vomited.” Id.

On October 2, 2019, Jack Flynn—SJU’s Director of Student Conduct—

informed Roe that he was expelled from SJU, effective immediately, for his

misconduct. Roe appealed his expulsion to SJU’s Conduct Appeals Board on

October 25, 2019. SJU’s Appeals Board affirmed the SJU Conduct Board’s

findings and sanctions in all respects and closed the matter on January 15, 2020.

See JA 152, 157 (reasoning that “[t]he evidence showed that [Roe] digitally

penetrated [Smith’s] vagina” and that “the sanction in this case—expulsion—was

the only one justified by the evidence and the record”).

II. Procedural History

On August 14, 2019, several months before his expulsion, Roe filed this

lawsuit against SJU and Doe in the United States District Court for the Eastern

District of New York. On February 18, 2020, Roe filed a second amended

complaint that alleged that SJU had engaged in sex-based discrimination in

violation of Title IX and had breached an implied contract with Roe. He further

claimed that Doe had defamed him by authoring and disseminating the

12
21-1125
Roe v. St. John’s University

anonymous tweet. On May 8, 2020, Doe asserted counterclaims alleging

intentional infliction of emotional distress, negligent infliction of emotional

distress, and prima facie tort against Roe.

On March 31, 2021, the district court granted SJU’s motion to dismiss Roe’s

claims in their entirety. See generally Roe v. St. John’s Univ., No. 19-CV-4694 (PKC)

(RER), 2021 WL 1224895 (E.D.N.Y. Mar. 31, 2021). The court held that Roe had

failed to state a claim under Title IX; it also declined to exercise supplemental

jurisdiction over Roe’s state-law claims and over Doe’s counterclaims, dismissing

them.

Roe appealed the district court’s judgment to this Court. Doe did not

appeal the dismissal of her counterclaims.

DISCUSSION

I. Standard of Review

We review de novo the district court’s grant of SJU’s motion to dismiss

under Federal Rule of Civil Procedure 12(b)(6). Cornelio v. Connecticut, 32 F.4th

160, 168 (2d Cir. 2022). A complaint must plead “enough facts to state a claim to

relief that is plausible on its face.” Bell Atl. Corp. v. Twombly, 550 U.S. 544, 570

(2007). “A claim is plausible ‘when the plaintiff pleads factual content that

13
21-1125
Roe v. St. John’s University

allows the court to draw the reasonable inference that the defendant is liable for

the misconduct alleged.’” Matson v. Bd. of Educ., 631 F.3d 57, 63 (2d Cir. 2011)

(quoting Ashcroft v. Iqbal, 556 U.S. 662, 678 (2009)). Although all factual

allegations contained in the complaint are assumed to be true, this rule does not

extend “to legal conclusions. Threadbare recitals of the elements of a cause of

action, supported by mere conclusory statements, do not suffice.” Iqbal, 556 U.S.

at 678 (citing Twombly, 550 U.S. at 555).

II. Roe’s Argument Under Rule 12(b)(6)

Roe first argues that the district court violated Federal Rule of Civil

Procedure 12(b)(6) by improperly considering Doe’s allegations regarding Roe’s

behavior when granting the motion to dismiss Roe’s complaint. The district

court did recount Doe’s competing allegations against Roe in its memorandum

and order. Roe, 2021 WL 1224895, at *4. But, as discussed below, Roe has argued

that SJU discriminated against him on the basis of his sex by investigating Doe’s

and Smith’s accusations of sexual assault more seriously than his allegation that

Doe harassed and defamed him in an anonymous tweet. This theory required

the district court to give at least some consideration to Doe’s and Smith’s

allegations against Roe to compare them with Roe’s accusations.

14
21-1125
Roe v. St. John’s University

Moreover, the district court’s memorandum and order decided not only

whether Roe had stated a plausible claim against SJU, but also whether to

exercise supplemental jurisdiction over Doe’s counterclaims against Roe. Roe,

2021 WL 1224895, at *11, *23. Deciding whether to exercise jurisdiction over

Doe’s counterclaims is an issue that also justifies some consideration of Doe’s

counterclaims’ alleged factual bases.

Finally, the district court recognized its obligation to accept the allegations

in Roe’s complaint as true when deciding SJU’s motion to dismiss. See id., at *12.

And “[i]n the absence of contrary indications, courts are generally presumed to

know the laws that govern their decisions and to have followed them.” United

States v. Banks, 464 F.3d 184, 190 (2d Cir. 2006). Other than noting that the district

court “referenc[ed] the allegations of Doe’s counterclaims” when issuing its

decision, Plaintiff-Counter-Defendant-Appellant’s Br. 22, Roe has given us no

reason to conclude that the district court violated its obligations under Rule

12(b)(6). We therefore reject Roe’s argument that we should vacate the district

court’s decision because the district court summarized Doe’s recollection of

events when describing this case’s factual background.

15
21-1125
Roe v. St. John’s University

III. Roe’s Title IX Claims

Roe alleges that SJU violated Title IX by discriminating against him

because of his sex in, broadly speaking, two different ways. Roe first claims that

anti-male bias influenced SJU’s disciplinary proceedings. He also contends that

SJU’s failure to investigate his allegations regarding the anonymous tweet that

accused him of sexual assault demonstrated deliberate indifference to a hostile

educational environment.

A. Roe’s Complaints Regarding SJU’s Disciplinary Proceedings

Title IX provides that “[n]o person in the United States shall, on the basis

of sex, be excluded from participation in, be denied the benefits of, or be

subjected to discrimination under any education program or activity receiving

Federal financial assistance.” 20 U.S.C. § 1681(a). 8 Hence, “Title IX bars the

imposition of university discipline where gender is a motivating factor in the

decision to discipline.” Yusuf v. Vassar Coll., 35 F.3d 709, 715 (2d Cir. 1994). A

Title IX complaint alleging that a university discriminated against a plaintiff on

the basis of the plaintiff’s sex when disciplining the plaintiff sufficiently alleges

8Although SJU is a private university, it receives federal funds, and Roe has alleged that SJU is
therefore subject to Title IX. SJU does not contend otherwise.

16
21-1125
Roe v. St. John’s University

that the university acted with the discriminatory intent required for a successful

Title IX claim when, “like a complaint under Title VII, . . . it pleads specific facts

that support a minimal plausible inference of such discrimination.” Doe v.

Columbia Univ., 831 F.3d 46, 56 (2d Cir. 2016). This minimal-plausible-inference-

of-discrimination standard is “low,” id. at 48, but failure to meet it is fatal, see,

e.g., Doe v. N.Y. Univ., 438 F. Supp. 3d 172, 181, 183–86 (S.D.N.Y. 2020) (noting

that “Title IX claims require evidence of intentional discrimination” and granting

motion to dismiss the plaintiff’s Title IX claim in part because the plaintiff’s

allegations failed to establish a plausible inference of intentional discrimination).

“[G]enerally,” a plaintiff “attacking a university disciplinary proceeding

on grounds of gender bias” asserts that the university is liable under one or both

of two separate theories: an erroneous outcome theory or a selective

enforcement theory. Yusuf, 35 F.3d at 715; see also Doe v. Colgate Univ., 760 F.

App’x 22, 30 (2d Cir. 2019) (summary order). The essence of an erroneous

outcome claim is that due at least in part to the plaintiff’s sex, the university

wrongly concluded that the plaintiff committed misconduct. Yusuf, 35 F.3d at

715. To successfully plead such a claim, the plaintiff must allege (1) “facts

sufficient to cast some articulable doubt on the accuracy of the outcome of the

17
21-1125
Roe v. St. John’s University

disciplinary proceeding” and (2) “circumstances suggesting that gender bias was

a motivating factor behind the erroneous finding.” Id.; see also Colgate Univ., 760

F. App’x at 30. The essence of a selective enforcement claim is that “regardless of

the [plaintiff’s] guilt or innocence,” the university initiated disciplinary

proceedings against the plaintiff or disciplined the plaintiff more severely at least

in part because of the plaintiff’s sex. Yusuf, 35 F.3d at 715.

Roe argues on appeal that he has successfully pleaded in his complaint

that SJU is liable under both an erroneous outcome theory and a selective

enforcement theory. 9 Roe contends that both theories are viable in this case

9 We do not dispute the Dissent’s assertion that the erroneous outcome and selective
enforcement theories described in Yusuf are not necessarily the only ways in which a plaintiff
may show that a university’s disciplinary proceedings exhibited sex-based bias. Dissent 4–5; see
also Yusuf, 35 F.3d at 715 (noting that plaintiffs’ claims will “generally” fall within these two
categories). In this case, however, Roe’s challenges to SJU’s disciplinary proceedings do fit
within Yusuf’s erroneous outcome and selective enforcement categories, and Roe himself
framed his claims around these two theories of sex-based discrimination. See Plaintiff-Counter-
Defendant-Appellant’s Br. 24, 30–31. Determining whether Roe has sufficiently alleged that he
was the victim of sex discrimination by analyzing the sufficiency of Roe’s erroneous outcome
and selective enforcement theories is also consistent with our precedents, even those that
postdate Columbia University. See, e.g., Radwan v. Manuel, 55 F.4th 101, 130–32 (2d Cir. 2022)
(utilizing Yusuf’s framework when resolving a Title IX claim); Colgate Univ., 760 F. App’x at 30,
33 (same).
We do, however, disagree with the Dissent’s suggestion that reliance on Yusuf is incompatible
with the applicable standards under Title IX and Rule 12(b)(6). Dissent 5–6. Several of our
recent Title IX cases have cited Yusuf without any suggestion that it placed too high a pleading
burden on plaintiffs. See, e.g., Vengalattore v. Cornell Univ., 36 F.4th 87, 103 (2d Cir. 2022);
Columbia Univ., 831 F.3d at 55–56; Colgate Univ., 760 F. App’x at 30.

18
21-1125
Roe v. St. John’s University

because his complaint contains sufficient factual allegations for a court crediting

them to plausibly infer that his sex partially motivated SJU’s decisions to

discipline Roe for sexual assault and to not investigate Roe’s allegations about

the anonymous tweet. We analyze Roe’s erroneous outcome and selective

enforcement claims in turn.

1. Roe’s Erroneous Outcome Claims

Roe asserts in his complaint that his sex was a motivating factor in SJU’s

decisions to discipline him for allegedly sexually assaulting Doe and Smith.

Roe’s argument fails because, even assuming that his complaint alleges “facts

sufficient to cast some articulable doubt on the accuracy of the outcome of the

disciplinary proceeding[s],” Roe’s complaint does not sufficiently allege

“circumstances suggesting that gender bias was a motivating factor behind the

erroneous finding[s].” Yusuf, 35 F.3d at 715. 10

Beginning with SJU’s handling of Doe’s complaint, Roe’s allegations

indicating that his sex influenced SJU’s actions can be grouped into two general

categories. First, Roe contends that his factual allegations, if proven, indicate that

10It may bear re-emphasis here that the issue before us is not whether SJU’s determinations
respecting Roe’s behavior were incorrect, but if so, whether they were incorrect as a result of
gender bias on the part of SJU.

19
21-1125
Roe v. St. John’s University

SJU provided a baseless explanation for its erroneous determination that Roe had

committed sexual misconduct when—according to Roe—Doe grabbed his hand

and put it on her breast. Roe asserts that SJU’s justification for its decision to

discipline him is misguided inasmuch as SJU based its decision on the facts that

Roe “admitted in [sic] engaging in physical contact of a sexual nature with” Doe

when she could not affirmatively consent to that contact due to her intoxication.

JA 80–81; see also Plaintiff-Counter-Defendant-Appellant’s Br. 33 (“Obviously, the

sexual contact to which Roe admitted was a violation of the Student Code by his

accuser, not conduct constituting a violation of the Student Code by him.”).

Accepting Roe’s version of events as true, as we must at this stage of the

litigation, SJU erroneously concluded that Roe violated SJU’s Student Code of

Conduct by engaging in non-consensual sexual contact with Doe. But it does not

follow that SJU reached this allegedly erroneous outcome due to gender bias. As

our cases have long held, such an allegation of an erroneous outcome, absent any

additional allegations of fact indicating bias on account of sex, does not state a

claim under Title IX. See, e.g., Columbia Univ., 831 F.3d at 57; Yusuf, 35 F.3d at 715

(explaining that a plaintiff proceeding on an erroneous outcome theory must

allege facts that cast doubt on the disciplinary proceeding’s outcome and

20
21-1125
Roe v. St. John’s University

“allegation[s] relating to a causal connection between the flawed outcome and

gender bias”).

Roe’s assertion that SJU provided a faulty explanation for its erroneous

conclusion also does not convince us that the district court erred by dismissing

Roe’s complaint. SJU’s stated reasoning for its conclusion was that Roe admitted

to engaging in physical contact of a sexual nature with Doe while she was

intoxicated. Roe therefore asserts that SJU’s decision supports an inference of

bias because it “[found] him in violation of the St. John’s rules based upon the

conduct of Doe taking his hand and placing it upon her breast.” JA 19. Even

reading this, as the Dissent does, as an allegation that “SJU accepted his version

of events,” Dissent 9, we cannot turn a blind eye to “an obvious alternative

explanation” for alleged facts that undermine a plaintiff’s theory of liability,

Twombly, 550 U.S. at 567. Here, the obvious alternative explanation, based on the

facts alleged by Roe, is that SJU accepted Roe’s concession that he engaged in

sexual contact with Doe but did not credit his assertion that Doe initiated the

contact. 11

11We note that the SJU panel that decided Smith’s complaint against Roe concluded that “key
aspects of [Roe’s] account . . . lacked credibility.” JA 143. While SJU did not explicitly note this
finding in its explanation for its decision to discipline Roe for the Doe incident, factfinders are

21
21-1125
Roe v. St. John’s University

Furthermore, the allegation that a university conducted its disciplinary

proceedings in a less-than-flawless manner does not automatically permit a

factfinder to reasonably infer that a university has committed sex discrimination.

For instance, allegations of a deficient Title IX investigation may not be enough

to enable a plaintiff to survive a motion to dismiss. See Doe v. Samford Univ., 29

F.4th 675, 689 (11th Cir. 2022). A university’s failure to comply with its internal

Title IX policies may similarly be insufficient to support a minimal plausible

inference of sex discrimination. See id. at 688 (“A deviation from a Title IX policy

is not, in and of itself, a violation of Title IX.”). 12 Finally, even allegations of

“potentially serious flaws” in a Title IX plaintiff’s disciplinary proceedings may

fail to allege “sufficient facts to support a plausible inference that the

irregularities are attributable to sex bias.” Doe v. Stonehill Coll., Inc., 55 F.4th 302,

permitted to make decisions on the basis of unstated credibility assessments in settings such as
criminal proceedings that are more formal than school disciplinary hearings. See Doe v. Samford
Univ., 29 F.4th 675, 691 (11th Cir. 2022). As noted above, Roe’s October 2018 Conduct Hearing
was not required to provide many of the procedural protections afforded in a federal criminal
trial such as utilization of the beyond a reasonable doubt standard and adherence to the Federal
Rules of Evidence.
12The Dissent discusses Samford at length based on its suggestion that bias against respondents
in disciplinary proceedings does not necessarily reflect bias on the basis of sex. See Dissent 17–
21. Because the allegations in Roe’s complaint do not require this court to reach this question,
we refrain from responding.
22
21-1125
Roe v. St. John’s University

334 (1st Cir. 2022). In short, alleged “procedural errors are not inevitably a sign

of sex bias.” Id.

Roe’s complaint lacks the factual allegations of significant investigatory

and procedural irregularity that have raised a plausible inference of

discrimination in similar cases. For instance, the male plaintiff in Columbia

University alleged that his university conducted a biased investigation of the

sexual assault allegations against him. See 831 F.3d at 49–50. According to the

male plaintiff’s complaint, the Title IX investigator’s questioning of him “was

akin to cross-examination calculated to elicit a confession,” id. at 49, while the

questioning of the plaintiff’s accuser lacked leading questions and was more

neutral, id. at 50. The plaintiff in Columbia University also alleged that he

reviewed the Title IX investigator’s notes from their previous meeting and that

the notes “inaccurately and inadequately paraphrased” his recollection of events.

Id. at 50. Finally, the plaintiff alleged that the Title IX investigator failed to

interview specific witnesses who could have supported his recollection of events.

See id. at 52 (“[The investigator’s report] did not include reference to witnesses

who could have supported Plaintiff’s defense, allegedly because [the

investigator] had declined to follow the leads Plaintiff had given her.”).

23
21-1125
Roe v. St. John’s University

Similarly, in Menaker v. Hofstra University, 935 F.3d 20 (2d Cir. 2019), 13 the

male plaintiff accused of committing sexual harassment alleged that he provided

the university that investigated those allegations with documents that proved

that the accusations against him were false, id. at 29. Similarly, the plaintiff

alleged that an official involved in the disciplinary process knew that at least one

of the accusations was false and believed the complaint to be a “ploy.” Id. at 28–

29. The male plaintiff also identified particular students who could provide

useful information, but whom the university did not interview. Id. at 29.

Roe’s complaint does not contain similar allegations pointing to specific

ways in which SJU’s investigation of the accusations against him was deficient

and biased. He merely complains that SJU justified its decision against him with

an unsatisfying explanation. Cf. Doe v. Univ. of So. Ind., 43 F.4th 784, 797 (7th Cir.

2022) (“What we have left are procedural choices that could arguably be

considered mistakes. They are not enough to show a likely bias against men.”).

We do not think that this allegation rises to the level of the allegations of

13Menaker was brought by a male university employee under Title VII of the Civil Rights Act of
1964 after he was terminated following a student’s complaint of sexual harassment. 935 F.3d at
26–28. But as we explained in Menaker, “[w]e apply similar principles in both Title VII and Title
IX when seeking to identify discriminatory intent.” Id. at 32 (discussing Columbia University’s
application to Title VII cases).

24
21-1125
Roe v. St. John’s University

objectively deficient investigations in Columbia University and Menaker. See

Samford Univ., 29 F.4th at 688 (“[A]llegations that are ‘merely consistent with’

liability ‘stop[] short of the line between possibility and plausibility.’” (second

alteration in original) (quoting Twombly, 550 U.S. at 557)).

Despite these deficiencies, the Dissent argues that we should conclude that

SJU’s justification for its allegedly erroneous decision to discipline Roe is itself

enough to support an inference of sex bias. Dissent 13–14. But such a conclusion

would conflict with this Circuit’s precedents, see, e.g., Yusuf, 35 F.3d at 715

(“[A]llegations of a . . . flawed proceeding that has led to an adverse and

erroneous outcome combined with a conclusory allegation of gender

discrimination is not sufficient to survive a motion to dismiss.”), and Title IX’s

text, which prohibits discrimination only “on the basis of sex,” 20 U.S.C.

§ 1681(a). And while it may be true that in some cases an erroneous decision is

sufficiently “inexplicable” to warrant inferring that the university reached its

decision due to sex-based bias, see Doe v. Oberlin Coll., 963 F.3d 580, 588 (6th Cir.

2020), this is not such a case. Roe’s allegations regarding SJU’s treatment of

Doe’s accusations against him do not give us “grave” doubts as to “the merits of

the decision itself,” id., and they are not serious enough to give rise to a minimal

25
21-1125
Roe v. St. John’s University

plausible inference of sex discrimination, see id. (noting that in that case the

alleged facts favored the male plaintiff to such a degree that “one could regard

this as nearly a test case regarding the College’s willingness ever to acquit a

respondent”).

Roe’s second category of allegations supporting his argument that his sex

influenced SJU’s review of Doe’s complaint against him consists of allegations

that SJU treated Doe differently than Roe. For instance, Roe alleges that SJU

suspended Roe and not Doe for conduct that, according to Roe’s complaint, Doe

initiated and to which Roe did not consent. Roe also points to SJU’s decision not

to investigate his allegation that Doe published a tweet accusing Roe of sexually

assaulting her. We do not think that these allegations create a minimal plausible

inference of sex discrimination.

As we will discuss further when analyzing Roe’s selective enforcement

theory of discrimination, Roe’s allegations regarding SJU’s different treatment of

Roe and Doe are not persuasive because Roe and Doe were not similarly

situated. Although Roe’s complaint alleges that it was Doe who initiated sexual

contact with Roe, Roe does not allege that he reported Doe to SJU for sexual

misconduct. Cf. Doe v. Rollins Coll., 77 F.4th 1340, 1354 (11th Cir. 2023) (affirming

26
21-1125
Roe v. St. John’s University

rejection of male plaintiff’s selective enforcement theory of discrimination in part

because male plaintiff did not file a sexual misconduct complaint against alleged

female victim). We also do not view Roe’s allegation that Doe harassed and

defamed him in an anonymous tweet as comparable to Doe’s allegation that Roe

sexually assaulted her in light of the differences between the nature and, indeed,

seriousness of the two offenses. 14 Roe cites Prasad v. Cornell University, Civ. A.

No. 5:15-CV-322, 2016 WL 3212079 (N.D.N.Y. Feb. 24, 2016), for the proposition

that “differential treatment between the complainant and respondent in the

adjudicatory process” can create a plausible inference of sex discrimination,

Plaintiff-Counter-Defendant-Appellant’s Br. 31. But Roe has not explained how

SJU treated Roe and Doe differently when reviewing Doe’s complaint other than

by not suspending Doe for her alleged behavior. As noted, there were no formal

sexual assault charges against Doe.

In short, Roe argues that—accepting his allegations as true, as we must on

a motion to dismiss—the SJU panel that reviewed Doe’s complaint got it wrong.

Roe supports his argument by offering a description of events in which his

14For these reasons, especially inasmuch as the tweet was not comparable to Roe’s alleged
behavior, even if Roe had alleged that Doe published the tweet non-anonymously, which he did
not, we do not think that his allegations would give rise to a minimal plausible inference of sex
discrimination.
27
21-1125
Roe v. St. John’s University

behavior was completely innocent and by noting the weaknesses in SJU’s

explanation for its decision to discipline him. But Roe’s second amended

complaint contains no plausible allegations linking SJU’s alleged error to gender

bias. And in light of the absence of “circumstances suggesting that gender bias

was a motivating factor behind [SJU’s] erroneous finding,” Yusuf, 35 F.3d at 715,

the district court did not err in concluding that Roe’s allegations regarding Doe’s

complaint failed to articulate a plausible Title IX claim.

Roe also alleges that his sex influenced SJU’s review of Smith’s complaint

against him.

[W]here a university (1) takes an adverse action against a student or
employee, (2) in response to allegations of sexual misconduct, (3)
following a clearly irregular investigative or adjudicative process, (4)
amid criticism for reacting inadequately to allegations of sexual
misconduct by members of one sex, these circumstances provide the
requisite support for a prima facie case of sex discrimination.

Menaker, 935 F.3d at 33. Roe argues that his complaint’s allegations meet this

standard in light of the #SurvivingSJU “tweet storm” and the procedural

irregularities that he argues impacted SJU’s review of Smith’s complaint. We

disagree.

Beginning with Roe’s allegations about the “tweet storm,” he claims that

#SurvivingSJU was the number one hashtag trending on Twitter in the United
28
21-1125
Roe v. St. John’s University

States between January 4 and January 6, 2019, and that thousands of tweets

under the hashtag criticized SJU’s treatment of women’s sexual assault claims.

These allegations do weigh in favor of Roe’s theory of sex discrimination. See

Menaker, 935 F.3d at 33. But we have also made clear that public pressure,

standing alone, is not sufficient to permit a plaintiff to survive a motion to

dismiss. Instead, the pressure must be “combined with clear procedural

irregularities in a university’s response to allegations of sexual misconduct.” Id.

(emphasis added); see also id. at 33 n.48 (suggesting that evidence of clear

procedural irregularity is more important than evidence of outside pressure).

Roe has failed to identify any clear procedural irregularities in either SJU’s

immediate or ultimate response to Smith’s complaint.

Regarding SJU’s immediate response to Smith’s complaint, around the

same time that #SurvivingSJU was trending, SJU ordered Roe and Smith to

refrain from contacting each other and notified Roe that he would be

immediately suspended due to, as SJU later officially informed Roe, Smith’s

29
21-1125
Roe v. St. John’s University

sexual assault accusation. 15 These actions were entirely consistent with SJU’s

internal procedures.

SJU’s Student Code of Conduct describes interim actions that the

university may institute after receipt of a complaint. The Student Code of

Conduct states that interim actions, including but not limited to “an interim

suspension; a ‘no contact’ order with the student who raised the concern or with

other students involved or with knowledge of the matter; . . . [and] limitation of

privileges to engage in specified University activities,” may be issued “for any

student accused of a violation.” JA 223. The Student Code of Conduct also

states:

Interim actions may be issued in the following circumstances: (1) to
ensure the physical or emotional safety and well-being of members of
the University or its property; (2) to ensure the student’s own physical
or emotional safety and well-being; or (3) if the student poses an
ongoing threat or disruption of the normal operations of the
University.

15Around this time, SJU’s Appeals Board affirmed the Conduct Board’s decision to suspend Roe
for sexually assaulting Doe. But since the Appeals Board merely affirmed the Conduct Board’s
October decision for the reasons the Conduct Board gave for its actions, see JA 83 (“[Roe,] while
sober, engaged in physical conduct of a sexual nature with [Doe] . . . [while] lack[ing]
affirmative consent to do so.”), and Roe has not alleged any procedural irregularities in SJU’s
review of Doe’s complaint other than the Conduct Board’s incomplete justification for its earlier
decision, the January “tweet storm” does not change our conclusion that Roe failed to plausibly
allege that his sex influenced SJU’s review of Doe’s complaint.

30
21-1125
Roe v. St. John’s University

....

Once imposed, an interim suspension takes effect immediately. A
suspension pending a hearing is not a University sanction, and no
notation of it will be made in the student’s transcript or file.

JA 223.

SJU issued a no contact order between Roe and Smith in January 2019.

SJU’s Student Code of Conduct explicitly allows for SJU to issue such an order,

and it does not appear to us to be irregular for SJU to have issued this order after

Smith accused Roe of sexually assaulting her. While Roe complains that he was

suspended on a temporary basis “without any explanation or an opportunity to

present his case,” Plaintiff-Counter-Defendant-Appellant’s Br. 41, this

suspension was consistent with SJU’s internal procedures, which allow for the

issuance of an interim suspension “[a]t any time during the Student Conduct

Process, and at the discretion of the Vice President of Student Affairs, the Dean of

Students or designee.” JA 223. The suspension also appears reasonable since

Roe was accused of sexually assaulting a student while serving the suspension

he received for mistreating another student. We do not see any evidence of

“clear procedural irregularities” impacting these interim decisions. Menaker, 935

F.3d at 33.

31
21-1125
Roe v. St. John’s University

We also are not convinced that Roe has identified any particularly

significant irregularities in SJU’s ultimate decision regarding Smith’s complaint.

Roe focuses his attention on two principal aspects of the disciplinary process: (1)

the Conduct Hearing panel’s composition and deliberations and (2) the Appeals

Board’s conclusions.

First, Roe asserts that “as a result of the bias of [SJU] officials from the

tweet storm, it’s [sic] attorney[, Thomas Foley,] improperly served as chair of the

Smith hearing panel.” Plaintiff-Counter-Defendant-Appellant’s Br. 35. 16 But Roe

16While Roe did not mention his claim about Foley’s appointment as chair in Roe’s counseled
brief in opposition to SJU’s motion to dismiss, Roe now raises this claim on appeal. Because it
was not raised previously, we need not address it. See, e.g. In re Nortel Networks Corp. Sec. Litig.,
539 F.3d 129, 132 (2d Cir. 2008) (per curiam). Because Roe devotes so much of his attention on
appeal to this claimed irregularity, we nonetheless exercise our discretion to evaluate this claim.
We conclude that it is meritless.
We decline to exercise our discretion to consider the other alleged procedural irregularities that
Roe did not raise on appeal—but on which the Dissent now focuses—such as the claim that Roe
was denied the opportunity to present an impact statement regarding the Smith incident and
the Dissent’s assertion that SJU’s delay in deciding Smith’s complaint violated a separate SJU
policy that the district court deemed inapplicable. As this Court has explained, “[w]e think it
reasonable to hold appellate counsel to a standard that obliges a lawyer to include his most
cogent arguments in his opening brief, upon pain of otherwise finding them waived.”
McCarthy v. SEC, 406 F.3d 179, 186 (2d Cir. 2005). This practice “promotes the orderly briefing
and consideration of appeals.” Id.; see also Fed. R. App. P. 28(a)(8) (stating that appellant’s brief
“must contain . . . appellant’s contentions and the reasons for them, with citations to the
authorities and parts of the record on which the appellant relies”); JP Morgan Chase Bank v. Altos
Hornos de Mex., S.A. de C.V., 412 F.3d 418, 428–29 (2d Cir. 2005) (affirming district court’s order
granting defendant’s motion to dismiss and holding that “arguments not made in an appellant’s
opening brief are waived even if the appellant pursued those arguments in the district court or
raised them in a reply brief”); LoSacco v. City of Middletown, 71 F.3d 88, 92 (2d Cir. 1995)

32
21-1125
Roe v. St. John’s University

has not demonstrated that Foley’s appointment as chair was irregular; 17 SJU

policy states that SJU “reserves the right to have an attorney as a part of the

Conduct Board hearing process.” JA 227. Moreover, Roe alleges that Foley

chaired the Doe panel in October 2018, which further weakens Roe’s assertion

that the January 2019 “tweet storm” caused SJU to appoint Foley to the Smith

panel.

Roe also complains that Foley and the other members of the Smith panel

deliberated privately with two other SJU officials—Jack Flynn, SJU’s Director of

Student Conduct, and John Breheny, an impartial investigator—before ruling on

the charges against Roe. Roe appears to assert that the Smith panel’s

deliberations with these officials violated SJU’s Student Code of Conduct’s

statement that after the conclusion of all testimony, “the Conduct Board shall

meet in private to deliberate the matter.” JA 228. But in part because SJU’s

Student Code of Conduct explicitly provides that “[h]earing panel members may

separately question [a] Student Conduct Administrator or Title IX Investigator as

(“Although [the appellant] stridently opposed the motion, he did not raise this issue in his
appellate brief. Consequently, he has abandoned it.”).
17Roe does claim that “St. John’s Code mandated that the three-member body be chaired by one
of its own.” Plaintiff-Counter-Defendant-Appellant’s Br. 35. Roe has not explained, though,
why Foley would not qualify as one of SJU’s “own.”

33
21-1125
Roe v. St. John’s University

appropriate,” id., it is not clear to us that the Smith panel’s expanded

deliberations violated SJU’s policies. In any event, any such deviation would

appear to us to be minor, insufficient to persuade us that SJU violated Title IX.

See Harris v. Niagara Mohawk Power Corp., 252 F.3d 592, 599 (2d Cir. 2001) (“[T]he

mere fact that an employer failed to follow its own internal procedures does not

necessarily suggest that the employer was motivated by illegal discriminatory

intent.” (citation omitted)); Samford Univ., 29 F.4th at 688 (“A deviation from a

Title IX policy is not, in and of itself, a violation of Title IX.”). 18 Roe also does not

identify any SJU policy that prevented the Smith panel from discussing Roe’s

case “before and after the hearing” or otherwise explain how these additional

deliberations demonstrate sex-based bias. Plaintiff-Counter-Defendant-

Appellant’s Br. 38.

18 Our conclusion that this deviation would be minor is based in part on our agreement with the
district court that, due to Flynn’s position, it “would seem to be the expected norm” for him to
be involved in the disciplinary process. Roe, 2021 WL 1224895, at *22. We also note that while
the Dissent claims that Roe has plausibly alleged that Flynn tainted the Smith panel with anti-
male bias because Flynn was the subject of criticism directed at SJU during the tweet storm for
its handling of Title IX complaints, Dissent 38–40, Roe’s complaint does not allege that the tweet
storm subjected Flynn to a harsher level of criticism than the many other SJU officials involved
in Title IX proceedings. In addition, Roe alleges that Flynn was involved in the October 2018
panel that considered Doe’s allegations and that predated the January 2019 tweet storm, further
undermining the suggestion that Flynn’s involvement in SJU’s disciplinary proceedings after
the tweet storm was irregular.

34
21-1125
Roe v. St. John’s University

Second, Roe complains that when the Appeals Board affirmed the Smith

panel’s conclusion that Roe violated SJU’s prohibition against non-consensual

sexual contact, the Appeals Board stated that “[t]he evidence showed that [Roe]

digitally penetrated [Smith’s] vagina while she was asleep.” JA 152. But the

Smith panel determined that Roe was guilty of violating only SJU’s policy

against non-consensual sexual contact—and not non-consensual sexual

penetration—despite Smith’s description of Roe’s alleged assault. JA 142. 19

However, since the Appeals Board recognized that the Smith panel determined

that Roe violated only the policy against non-consensual sexual contact, see JA

152 (listing the two charges against Roe and then stating the one charge the

Conduct Board found Roe violated), we do not consider the Appeals Board’s

statement regarding the evidence against Roe to be a major procedural

irregularity. 20

19 The Dissent describes the Smith panel’s conclusion as “perplexing.” Dissent 33. But as the
Dissent recognizes, the record does not contain all the evidence the Smith panel considered
when it reached its conclusion. See id. at 34 n.16 (explaining that the evidence against Roe
included accounts of Roe’s housemates that are not in the record). We therefore do not see any
firm basis upon which we can conclude that the Smith panel’s decision was perplexing or, more
importantly, whether the decision was perplexing in a way that lends support to Roe’s theory of
sex-based discrimination.
20For similar reasons, we are not persuaded by Roe’s claim that the Appeals Board acted highly
irregularly by allegedly misstating the evidence regarding Doe’s complaint against Roe. See
Plaintiff-Counter-Defendant-Appellant’s Br. 14–15 (noting that the Appeals Board stated that

35
21-1125
Roe v. St. John’s University

In sum, the alleged procedural irregularities are not serious enough to

support a claim of sex-based discrimination. See Menaker, 935 F.3d at 34 n.50

(“[W]e emphasize that our standard requires clear irregularities to raise an

inference of bias . . . . [M]inimal irregularities (absent other indicia of bias) do

not suffice to suggest discrimination.”). And in the absence of clear procedural

irregularities, Roe’s allegations regarding the “tweet storm” are insufficient to

support denial of the motion to dismiss. See id. at 33 & n.48. We conclude,

therefore, that even after considering Roe’s allegations concerning the Doe

incident and the Smith incident together, Roe has failed to articulate a viable

erroneous outcome claim because he has failed to plausibly allege

“circumstances suggesting that gender bias was a motivating factor behind

[SJU’s allegedly] erroneous finding[s]” that Roe sexually assaulted Doe and

Smith. Yusuf, 35 F.3d at 715.

2. Roe’s Selective Enforcement Claim

Roe also argues that SJU engaged in selective enforcement by deciding

Doe’s and Smith’s sexual assault complaints against him while failing to

Doe was unconscious, but the Doe panel found that she was intoxicated); see also Samford Univ.,
29 F.4th at 688–89 (explaining that the availability of obvious alternative “lawful explanations”
for procedural irregularities, including “ineptitude [and] inexperience,” undermine the viability
of a plaintiff’s Title IX claim).

36
21-1125
Roe v. St. John’s University

investigate his theory that Doe published the anonymous tweet accusing him of

sexual assault. The district court correctly ruled that Roe’s claim of selective

enforcement “fail[s] in the absence of allegations that SJU treated a similarly

situated student [of the opposite sex] differently.” Roe, 2021 WL 1224895, at *17.

As another district court in our Circuit recently explained, “following Columbia,

courts in this circuit have consistently dismissed selective enforcement claims

absent allegations that a school treated similarly situated members of the

opposite sex—that is, members of the opposite sex facing comparable

disciplinary charges—differently.” N.Y. Univ., 438 F. Supp. 3d at 182 (internal

quotation marks and citation omitted) (collecting cases).

To repeat: Roe and both Doe and Smith were not, in this context and for

these purposes, similarly situated. Roe accused Doe of anonymously publishing

an allegedly harassing, defamatory tweet. Doe and Smith accused Roe of sexual

assault. The allegation that SJU investigated complaints of sexual assault more

thoroughly than an allegation of circulation of a harassing and defamatory tweet

does not demonstrate selective enforcement. The conduct alleged was not

similar. Moreover, SJU observed that it could not “sanction the individual who

posted the tweet because [it could] not confirm their identity.” JA 95. Doe’s and

37
21-1125
Roe v. St. John’s University

Smith’s complaints, by contrast, were about alleged sexual attacks by a specific

individual whose identity was not in doubt. We therefore reject Roe’s selective

enforcement theory of sex discrimination.

3. Conclusion

For the foregoing reasons, we affirm the district court’s rejection of Roe’s

erroneous outcome and selective enforcement theories of discrimination. While

Roe has alleged a few significant facts that do indeed weigh in favor of his

claims, his allegations as a whole are not sufficient to plausibly support a

minimal inference of sex-based discrimination.

We note that, accepting Roe’s claims as true, SJU’s treatment of him and

the complaints against him may well not have been up to the standards that

would apply if the issues that the university officials decided were adjudicated in

a federal court proceeding. But Doe’s and Smith’s complaints against Roe were

not decided by Article III judges in a federal court—they were decided by school

officials in a school setting. Hence, even if we accept Roe’s claimed innocence

and agree that SJU’s conduct of Roe’s disciplinary proceedings was not flawless,

we conclude that, on the record before us, Roe has not alleged facts that could

38
21-1125
Roe v. St. John’s University

reasonably support a minimal plausible inference of sex discrimination. And that

is the issue before us.

B. Roe’s Hostile Educational Environment Claim

Roe brings a separate Title IX claim asserting that SJU’s failure to

investigate his allegations regarding the anonymous tweet subjected him to a

hostile educational environment in violation of Title IX. This claim fails because

the abuse that Roe claims to have suffered as a result of the tweet was not

sufficiently severe or pervasive to support a hostile environment claim.

Title IX permits a plaintiff to recover damages when he or she is subjected

to a hostile environment that constitutes discrimination on the basis of sex and

deprives the plaintiff of the ability to enjoy the benefits of an educational

program receiving federal funds. See Davis v. Monroe Cty. Bd. of Educ., 526 U.S.

629, 633 (1999); Papelino v. Albany Coll. of Pharmacy of Union Univ., 633 F.3d 81, 89

(2d Cir. 2011). The “plaintiff must show that he subjectively perceived the

environment to be hostile or abusive and that the environment objectively was

hostile or abusive, that is, that it was permeated with discriminatory

intimidation, ridicule, and insult sufficiently severe or pervasive to alter the

conditions of his educational environment.” Papelino, 633 F.3d at 89. A school

39
21-1125
Roe v. St. John’s University

may be liable for its deliberate indifference to acts of harassment committed by

students against other students. Davis, 526 U.S. at 633.

The district court properly dismissed Roe’s hostile environment claim

because the alleged harassment of Roe through an allegedly defamatory tweet in

this case was not “so severe, pervasive, and objectively offensive that it can be

said to [have] deprive[d him] of access to the educational opportunities or

benefits provided by the school.” Davis, 526 U.S. at 650. Roe’s claims stem from

a single anonymous tweet accusing him of sexual assault. 21 To be sure, the single

tweet was broadly disseminated, but “[g]enerally, incidents [of harassment] must

be more than episodic [to justify a hostile environment claim]; they must be

sufficiently continuous and concerted in order to be deemed pervasive.” Demoret

v. Zegarelli, 451 F.3d 140, 149 (2d Cir. 2006) (citation omitted). But cf. Ferris v.

Delta Air Lines, Inc., 277 F.3d 128, 136 (2d Cir. 2001) (explaining that in extreme

cases, such as rape, a single incident of abuse can give rise to a hostile

21Roe’s complaint does not indicate that he asked SJU to initiate disciplinary proceedings
against the female student who allegedly struck him in the face or the male student who
allegedly sent him threatening messages after the tweet was published. Instead, Roe focuses his
attention on SJU’s failure to investigate the individual who published the tweet. Regardless,
even if Roe’s additional allegations were considered part of his hostile environment claim, they
would not change our decision in light of the events’ sporadic and disconnected nature. See
Demoret v. Zegarelli, 451 F.3d 140, 149 (2d Cir. 2006).

40
21-1125
Roe v. St. John’s University

environment claim). We conclude that the single anonymous tweet brought to

the attention of SJU here was not, standing alone, sufficiently severe for a court

to conclude that Roe was denied access to an educational benefit even assuming

that the tweet was, as Roe asserts, false and offensive. Roe’s hostile environment

claim is therefore fatally deficient.

CONCLUSION

We have considered Roe’s remaining arguments on appeal and conclude

that they are without merit. For the foregoing reasons, we AFFIRM the

judgment of the district court.

41
 21-1125
Roe v. St. John's University

1
2 PARKER, Circuit Judge, concurring:

3 I concur in the Court’s opinion. In doing so, I am of course aware that

4 university disciplinary processes can present sensitive and difficult adjudicative

5 concerns. See Vengalattore v. Cornell University, 36 F.4th 87, 114-15 (2d Cir. 2022)

6 (Cabranes, J., concurring); Doe v. Miami Univ., 882 F.3d 579, 599-603 (6th Cir.

7 2018); Doe v. Baum, 903 F.3d 575, 581-85 (6th Cir. 2018); Plummer v. University of

8 Houston, 860 F.3d 767, 778-84 (5th Cir. 2017) (Jones, J., dissenting); Doe v. Univ. of

9 Cincinnati, 872 F.3d 393, 403-07 (6th Cir. 2017); Messeri v. DiStefano, 480 F. Supp.

10 3d 1157, 1164-66 (D. Colo. 2020); Doe v. Univ. of Mississippi, 361 F. Supp. 3d 597,

11 608-14 (S.D. Miss. 2019); Doe v. N. Michigan Univ., 393 F. Supp. 3d 683, 693-95

12 (W.D. Mich. 2019); Walter v. Queens Coll., 390 F. Supp. 3d 382, 402-03 (E.D.N.Y.

13 2019). While these issues will undoubtedly come before us in the future, I

14 believe the district court decided this case correctly.

15 I write separately to draw attention to the complicated circumstances

16 under which Saint John’s University (“SJU”) adjudicated Doe’s and Smith’s

17 sexual assault allegations against Roe, circumstances that may require university

18 adjudicators to make difficult decisions in the face of competing considerations.

1
 21-1125
Roe v. St. John's University

1 While Roe’s complaint, whose well-pleaded allegations the Court

2 rightfully takes as true at the motion-to-dismiss stage, alleged that it was Doe

3 who initiated sexual contact with Roe before subsequently submitting a sexual

4 assault complaint against him to SJU, Doe’s account of her encounter with Roe

5 differed markedly. Doe and Roe were both SJU students studying abroad in

6 Paris. According to Doe, on the night of the alleged assault, Roe pulled her away

7 from her friends at the club “to dance with him and then he kissed her.” JA 267.

8 Doe further claimed that Roe made her feel uncomfortable “immediately” at the

9 club. She “signaled her discomfort to her friend who came over and pulled Doe

10 away from [Roe] and they rushed to the bathroom together to get away from”

11 Roe. Id. Doe then left the club to return to her residence. After Doe returned to

12 her room and fell asleep, “[Roe] entered Doe’s dorm room while Doe was still

13 sleeping.” Doe recalled that she “awoke to [Roe] on top of her in bed.” Id. Doe’s

14 “pants had been pulled down and [Roe] was digitally penetrating her vagina,”

15 causing Doe to be “frozen in shock” and unable to speak or scream. Doe

16 eventually managed to turn herself over and asked Roe to stop, but instead “felt

17 [Roe] begin to masturbate against her back.” Id. Doe asked Roe to stop once

18 more, and he finally left. Doe’s account is, of course, not relevant to the opinion’s

2
 21-1125
Roe v. St. John's University

1 Rule 12(b)(6) analysis, see Majority Op. at 5 n.3, but it demonstrates that during

2 the hearing, SJU was faced with competing accounts of what had transpired

3 between the two.

4 As with Doe’s allegations, the majority opinion correctly refrains from

5 considering Smith’s underlying accusations in its analysis. See id. at 9 n.6.

6 However, both Smith’s allegations and the context in which they were made

7 further underscore that extensively second-guessing SJU’s handling of these

8 accusations is a road we should proceed down with caution.

9 Contrary to Roe’s recollection of their interaction, Smith reported “falling

10 asleep on the living room couch and awakening to discover [Roe] . . . with his

11 hand inside her pants” and “penetrating her vagina with his fingers” without her

12 consent. JA 28. Subsequently, she alleged violations of SJU’s policies against

13 non-consensual sexual contact and non-consensual sexual penetration. The

14 issues surrounding SJU’s consideration of the Smith incident were arguably

15 more complex because they followed the Doe hearing and SJU’s imposition of a

16 suspension on Roe for assaulting Doe. Given this sequence of events, the

17 opinion concludes, SJU reasonably suspended Roe pending further investigation

18 into Smith’s accusations. See Majority Op. at 32.

3
 21-1125
Roe v. St. John's University

1 Roe argues on appeal that his due process rights were violated by SJU’s

2 handling of both the Doe and Smith adjudicative processes. But the Court’s

3 opinion correctly notes that any potential procedural errors in SJU’s adjudication

4 do not ipso facto demonstrate sex bias against Roe, as he claims. See id. at 21-23.

5 As the opinion recognizes, school officials are responsible for the

6 education and wellbeing of their students, and they are obligated to balance

7 myriad interests, particularly those of the accused, when dealing with allegations

8 of sexual assault. These officials are not federal judges presiding at criminal

9 trials. See id. at 39. Thus, we should be cautious about reflexively imposing the

10 requirements and expectations of an Article III tribunal on them. Accordingly, I

11 concur.

4
21-1125
Roe v. St. John’s University

MENASHI, Circuit Judge, dissenting:

When St. John’s University (“SJU”) notified him that a fellow
student had accused him of sexual misconduct, Richard Roe (“Roe”)
sought to defend himself. He offered an account of the events
according to which his accuser, a fellow classmate in the Paris study-
abroad program named Jane Doe (“Doe”), grabbed his hand and
placed it on her body without Roe’s consent. SJU said it accepted
Roe’s version of events—according to which he was not a perpetrator
but a victim—but the university suspended him anyway.

While serving that suspension, Roe was accused of sexual
misconduct by another SJU student, Mary Smith (“Smith”). The
accusation came a single day after SJU became subject to widespread
public criticism that it did not protect female students from sexual
assault. Among those criticisms was an anonymous tweet accusing
Roe of assaulting Jane Doe in Paris. SJU swiftly issued an order
prohibiting Roe from contacting Smith. But within hours of that
order—and following a newspaper article reporting on the criticism
of SJU—the university altered course, suspending Roe pending an
investigation into the Smith complaint. After several months and
numerous procedural irregularities, SJU expelled Roe. Roe then filed
this Title IX action, alleging that SJU discriminated against him on the
basis of sex.

The court admits that, accepting the allegations of the
complaint, SJU “erroneously concluded that Roe violated SJU’s
Student Code of Conduct,” relied on “a faulty explanation for its
erroneous conclusion,” and “conducted its disciplinary proceedings
in a less-than-flawless manner.” Ante at 20-22. The court purports to
credit Roe’s allegation that “SJU provided a baseless explanation” and
a “misguided” justification for its decision to discipline Roe. Id. at 20.
Yet the court imagines that there may be an “alternative explanation”
for SJU’s flawed treatment of Roe—one which SJU never articulated—
that would not involve sex-based bias. Id. at 21. In other words, the
court accepts that Roe alleged irregular treatment but concludes that
he still cannot state a claim because the court itself can supply a reason
for that treatment that does not reflect sex-based bias.

In reaching this conclusion, the court fails to credit the
allegations of Roe’s complaint and makes factual inferences in favor
of SJU—in violation of well-settled principles for reviewing the grant
of a motion to dismiss. And the court misconstrues our precedents
such that, after today’s decision, most Title IX plaintiffs alleging
improper discipline will fail to state a claim.

The court seems to be uncomfortable with the requirement of
Title IX that schools walk a “razor’s edge” when evaluating
allegations of sexual misconduct. Doe v. Univ. of Illinois, 138 F.3d 653,
679 (7th Cir. 1998) (Posner, C.J., dissenting from the denial of
rehearing en banc). “One student’s demand for a quick response to
her harassment complaint will conflict with the alleged harasser’s
demand for due process.” Davis v. Monroe Cnty. Bd. of Educ., 526 U.S.
629, 682 (1999) (Kennedy, J., dissenting). Title IX requires schools to
balance both interests; when it fails to do so, it is subject to liability for
discrimination. “A recipient’s treatment of a complainant or a
respondent in response to a formal complaint of sexual harassment
may constitute discrimination on the basis of sex under title IX.”
34 C.F.R. § 106.45(a) (emphasis added); see also id. § 106.45(b)(1)
(providing that a grievance process must “[t]reat complainants and
respondents equitably”) (emphasis added). The court discounts those
authorities providing that the denial of adequate process to a
respondent—that is, “basic principles of fairness and due process”—
amounts to discrimination in violation of Title IX. Vengalattore v.

2
Cornell Univ., 36 F.4th 87, 115 (2d Cir. 2022) (Cabranes, J., concurring). 1
And it splits from other circuits that have held that an “inexplicable”
disciplinary decision raises an inference of sex-based bias. Doe v.
Oberlin Coll., 963 F.3d 580, 588 (6th Cir. 2020). Instead, the court holds
that plausible allegations of an erroneous conclusion, flawed
procedures, and baseless reasoning do not even state a claim under
the civil rights laws. I dissent.

I

I begin with the pleading standard that Roe’s complaint must
meet to survive a motion to dismiss. In Doe v. Columbia University, we
held that a complaint alleging a Title IX violation must meet the “low
standard described in Littlejohn v. City of New York, 795 F.3d 297 (2d
Cir. 2015).” 831 F.3d 46, 48 (2d Cir. 2016). Under that standard:

[A] complaint under Title IX, alleging that the plaintiff
was subjected to discrimination on account of sex in the
imposition of university discipline, is sufficient with
respect to the element of discriminatory intent, like a

1 The court emphasizes that a university disciplinary process does not
necessarily include the same procedural protections as a court. See ante at 4,
6, 22 n.11, 38. It therefore does not matter, says the court, that “SJU’s
treatment of [Roe] and the complaints against him may well not have been
up to the standards that would apply if the issues that the university
officials decided were adjudicated in a federal court proceeding.” Id. at 38.
But the question here is whether SJU’s disciplinary actions violated Title IX,
not federal court procedural standards. And it is well-established that
“procedural irregularities in the investigation and adjudication of the
accusations” by a university disciplinary process may indicate
discriminatory treatment that violates Title IX. Menaker v. Hofstra Univ., 935
F.3d 20, 31 (2d Cir. 2019). So we must consider the procedural regularity of
SJU’s disciplinary process and cannot excuse irregularities on the ground
that a university is not a court.

3
 complaint under Title VII, if it pleads specific facts that
support a minimal plausible inference of such
discrimination.

Id. at 56; see also Menaker, 935 F.3d at 30 (“A plaintiff need only allege
facts that give plausible support to a minimal inference of
discriminatory motivation.”) (internal quotation marks omitted).

Before our decision in Columbia University, the leading case
related to student discipline was Yusuf v. Vassar College, 35 F.3d 709
(2d Cir. 1994). In Yusuf, we identified the “erroneous outcome” and
“selective enforcement” tests. Id. at 715. When evaluating a Title IX
claim under the “erroneous outcome” test, we ask whether a plaintiff
alleged (1) “facts sufficient to cast some articulable doubt on the
accuracy of the outcome of the disciplinary proceeding,” and
(2) “particular circumstances suggesting that gender bias was a
motivating factor behind the erroneous finding.” Id. 2 When
evaluating a Title IX claim under the “selective enforcement” test, we
ask whether a plaintiff alleged that “the severity of the penalty and/or
the decision to initiate the proceeding was affected by the student’s
gender.” Yusuf, 35 F.3d at 715.

It is important to note, however, that “[t]he tests in Yusuf … do
not capture the full range of conduct that could lead to liability under
Title IX.” Doe v. Samford Univ., 29 F.4th 675, 687 (11th Cir. 2022).

2 Title IX prohibits educational programs receiving federal assistance from
discriminating “on the basis of sex.” 20 U.S.C. § 1681(a) (emphasis added).
Yusuf used the term “gender” to mean “sex”; it did not distinguish between
these terms. Though such a distinction might be an issue in other cases, see,
e.g., Adams ex rel. Kasper v. Sch. Bd. of St. Johns Cnty., 57 F.4th 791, 814 (11th
Cir. 2022) (concluding that “reading ‘sex’ to include ‘gender identity’”
would not “comport with the plain meaning of ‘sex’ at the time of Title IX’s
enactment”), the issue is not presented in this case.

4
Rather, those tests “simply describe ways in which a plaintiff might
show that sex was a motivating factor in a university’s decision,” Doe
v. Purdue Univ., 928 F.3d 652, 667 (7th Cir. 2019) (Barrett, J.). Our
opinion in Yusuf does not hold that the “erroneous outcome” and
“selective enforcement” tests are the exclusive ways for a Title IX
plaintiff to state a claim. The opinion says only that plaintiffs
challenging university disciplinary proceedings “can be expected to
fall generally within [the] two categories” it describes. 35 F.3d at 715
(emphasis added).

We know that the Yusuf tests are not exclusive because our
subsequent decisions in Columbia University and Menaker did not even
attempt to fit the plaintiffs’ allegations into the Yusuf categories.
Instead, we explained that the key inquiry was whether a plaintiff’s
allegations support a “minimal plausible inference” that he was
“subjected to discrimination on account of sex in the imposition of
university discipline.” Columbia Univ., 831 F.3d at 56. While the Yusuf
tests may be helpful for particular types of cases, a plaintiff’s failure
to fit into the Yusuf formula is not a basis for granting a motion to
dismiss.

A recent decision in this area confirms that we apply the
“minimal plausible inference” standard rather than the Yusuf
framework. In Vengalattore v. Cornell University, the plaintiff claimed
that sex-based bias motivated Cornell in its investigation of the
plaintiff’s alleged sexual misconduct and its decision to discipline
him. We held that the complaint’s allegations supported an inference
of sex-based bias. See 36 F.4th at 109. Nowhere in the opinion did we
describe the plaintiff’s claim as an “erroneous outcome” claim per
Yusuf. Instead, we asked whether the allegations “support a minimal
plausible inference” of sex-based discrimination. Id. at 106.

5
 One important reason to apply the Columbia University
standard rather than the Yusuf framework is that Yusuf suggested that
a plaintiff must show a “particularized … causal connection between
the flawed outcome and gender bias.” 35 F.3d at 715. Such a
requirement would impose a heightened pleading burden on Title IX
plaintiffs, which conflicts with the “low standard described in
Littlejohn” that a plaintiff need only provide allegations to support a
“minimal inference” of discriminatory motivation. Columbia Univ.,
831 F.3d at 48. Yusuf’s confusion on this point may be understandable
because Yusuf was decided not only before Littlejohn but before the
Supreme Court decisions on which Littlejohn relied. 3 But we have
repeatedly admonished district courts to apply the Littlejohn standard
and not to impose a heightened “particularized” pleading burden on
Title IX plaintiffs. See Columbia Univ., 831 F.3d at 55 n.8 (“We have …
cautioned district courts against imposing too high a burden on
plaintiffs alleging discrimination at the 12(b)(6) stage.”); Dawson v.
N.Y.C. Transit Auth., 624 F. App’x 763, 770 (2d Cir. 2015) (“At the
pleading stage, district courts would do well to remember this
exceedingly low burden that discrimination plaintiffs face even after
they have survived a motion to dismiss.”); see also Schwake v. Ariz. Bd.
of Regents, 967 F.3d 940, 949 (9th Cir. 2020) (“There is no heightened
pleading standard for Title IX claims.”).

Today, the court abandons the low pleading standard we
described in Columbia University. Although it cites our more recent
cases, see ante at 17, the court declines to apply those standards
because “Roe’s challenges to SJU’s disciplinary proceedings … fit

3 See Littlejohn, 795 F.3d at 310 (concluding that “Iqbal’s [pleading]
requirement applies to Title VII complaints of employment discrimination”
in light of Ashcroft v. Iqbal, 556 U.S. 662 (2009), Bell Atl. Corp. v. Twombly, 550
U.S. 544 (2007), and Swierkiewicz v. Sorema N.A., 534 U.S. 506 (2002)).

6
within Yusuf’s erroneous outcome and selective enforcement
categories” and because Roe’s brief “frame[s] his claims around these
two theories of sex-based discrimination.” Id. at 18 n.9. That
explanation fails for three reasons. First, it is an overstatement. Roe’s
brief mentions the erroneous outcome and selective enforcement
tests, see, e.g., Appellant’s Br. 21, 23, but it also references Columbia
University’s “minimal plausible inference of … discrimination”
framework throughout and without reference to the Yusuf tests, see,
e.g., id. at 31 (arguing that “[t]he plaintiff need only allege facts that
support a minimal plausible inference of sexual discrimination on the
part of the university” such as “differential treatment between the
complainant and respondent in the adjudicatory process”). Roe does
not “frame[]” his appeal around the Yusuf tests.

Second, even if he had, that would not be a reason to confine
our analysis to those tests. In Columbia University, the Title IX plaintiff
largely did frame his brief around the Yusuf tests, citing the case over
a hundred times. See generally Brief for Plaintiff-Appellant-Cross-
Appellee, Doe v. Columbia University, 831 F.3d 46 (2d Cir. 2016)
(No. 15-1536), ECF No. 55. Rather than confine our analysis to Yusuf,
we announced the proper pleading standard without reference to
those tests. 4

4 The court asserts that confining our analysis to the erroneous outcome
and selective enforcement tests is “consistent with our precedents, even
those that postdate Columbia University,” citing one opinion and one
summary order. Ante at 18 n.9 (citing Radwan v. Manuel, 55 F.4th 101, 130-
32 (2d Cir. 2022); Doe v. Colgate Univ., 760 F. App’x 22, 30, 33 (2d Cir. 2019)).
But in both Radwan and Colgate University, we emphasized that the plaintiffs
expressly pursued one particular theory. See Radwan, 55 F.4th at 130 (“In
this case, Radwan pursues only a theory of selective enforcement.”); Colgate

7
 Third, the relevant inquiry focuses on the complaint. Roe’s
complaint does not even allude to the Yusuf tests. 5 Yet it
unmistakably pleads sex discrimination. See, e.g., App’x 19
(“St. John’s discriminatory bias against Roe as a male student is
evident.”); id. at 24 (alleging that SJU “should not have suspended
Roe without the due process required by its own policies and
procedures”); id. at 34 (“The Conduct Appeals Board[’s] bias against
Roe as a male student is reflected in their making factual findings
beyond their purview and jurisdiction.”). Roe has not conceded the
proposition that Yusuf—and Yusuf alone—controls this case.

Even if the court’s analysis under the two Yusuf tests were
correct—and I think it is not—the court ignores our precedents by
failing to evaluate, outside those tests, whether Roe’s complaint
pleaded facts that “support a minimal plausible inference of …
discrimination.” Columbia Univ., 831 F.3d at 56.

II

The complaint did plead such facts—even more clearly than a
typical Title IX complaint regarding student discipline. When a
university adjudicates a claim of sexual misconduct, it often must
evaluate two conflicting stories. If the respondent challenges the
university’s decision under Title IX, he typically alleges that the

Univ., 760 F. App’x at 30 (“John Doe’s claim proceeds under an ‘erroneous
outcome’ theory of gender bias.”).
5 That is unsurprising because a complaint must plead facts, not legal
theories. See Northrop v. Hoffman of Simsbury, Inc., 134 F.3d 41, 46 (2d Cir.
1997) (“[T]he failure in a complaint to cite a statute, or to cite the correct
one, in no way affects the merits of a claim. Factual allegations alone are
what matters.”) (quoting Albert v. Carovano, 851 F.2d 561, 571 n.3 (2d Cir.
1988)); see also Newman v. Silver, 713 F.2d 14, 15 n.1 (2d Cir. 1983) (“[F]ederal
pleading … is by statement of claim, not by legal theories.”).

8
university accepted the complainant’s version of events even though
the evidence did not support it. This is not such a case. Instead, Roe
alleged that SJU accepted his version of events. That allegation is
plausible because the university, in announcing its disciplinary
decision, explained that it relied on Roe’s having “admitted” to
“engaging in physical contact of a sexual nature with the
complainant” and the evidence indicating “a lack of affirmative
consent to engage in such contact.” App’x 14. But Roe alleged that the
only contact to which he admitted involved Doe taking his hand
without his consent. This version of events indicates that Roe engaged
in no misconduct, and yet the complaint alleged that SJU nevertheless
found him responsible for sexual misconduct and suspended him.
These allegations support “a minimal plausible inference” that Roe
was “subjected to discrimination on account of sex in the imposition
of university discipline.” Columbia Univ., 831 F.3d at 56.

A

When reviewing a motion under Rule 12(b)(6), “the only facts
to be considered are those alleged in the complaint, and the court
must accept them, drawing all reasonable inferences in the plaintiff’s
favor, in deciding whether the complaint alleges sufficient facts to
survive.” Id. at 48. 6 In today’s opinion, however, the court affirms the

6 Although the concurring opinion acknowledges that we must assume the
truth of Roe’s allegations in the complaint, it nevertheless recounts factual
allegations from Doe and Smith that do not appear in the complaint. Ante
at 2-3 (Parker, J., concurring). The concurrence insists that it is necessary to
consider these alternative factual accounts to give “context” to the
“complicated circumstances under which [SJU] adjudicated … [the]
allegations against Roe.” Id. at 3, 1. But the only “context” relevant to
deciding a motion to dismiss is what is contained in the complaint and “any
written instrument attached to it as an exhibit or any statements or

9
district court’s Rule 12(b)(6) dismissal on a basis other than the
allegations of the complaint. Despite purporting to credit Roe’s
allegations that SJU’s justification of its decision was “faulty” and
“baseless,” ante at 20-21, the court supplies an “alternative
explanation” for that decision and decides that, given that alternative
explanation, it is impossible to infer that SJU discriminated against
Roe on the basis of sex, id. at 21. This is a departure from the standard
principles for reviewing a dismissal for failure to state a claim.

Roe’s second amended complaint makes the following
allegations. Roe and Doe were students studying abroad in France.
On April 12, 2018, Roe was in Paris at a local club to celebrate a
birthday with fellow SJU students studying abroad. Doe asked Roe to
dance, and they did. Sometime later, Doe told Roe that she was
returning to the SJU dormitory where both were staying. Doe gave
Roe her room number and asked him to “check on her” when he
arrived at the dorm. App’x 12. Hours later, Roe went to Doe’s room
and found her awake and on her phone. “Doe invited Roe into her
room and, thereafter, Doe took Roe’s right hand and placed it upon
her fully clothed breast.” Id. Roe then told Doe, “I am not interested
in sex.” Id. She replied, “Then, get the hell out of here,” and Roe left.
Id.

On September 4, 2018, SJU notified Roe that Doe had filed an
incident report with four complaints against him: (1) non-consensual

documents incorporated in [the complaint] by reference.” Chambers v. Time
Warner, Inc., 282 F.3d 147, 152 (2d Cir. 2002) (quoting Int’l Audiotext Network,
Inc. v. AT&T Co., 62 F.3d 69, 72 (2d Cir. 1995)). Doe’s and Smith’s alternative
factual allegations do not qualify. The concurrence’s recitation of factual
allegations beyond those that appear in the complaint has no purpose but
to introduce disputed facts into our review of a motion to dismiss. By
contrast, I rely on the factual allegations of the operative complaint.

10
sexual contact on April 12, 2018, (2) sexually inappropriate conduct
on April 13, 2018, (3) non-consensual sexual contact on April 13, 2018,
and (4) non-consensual sexual penetration on April 13, 2018. See id. at
303-04. About a month later, Roe attended an SJU Conduct Board
hearing to evaluate Doe’s complaint. At the hearing, Roe said that Doe
“took [Roe’s] right hand and placed it upon her fully clothed breasts.”
Id. at 14.

In a letter dated October 15, 2018, the Conduct Board notified
Roe that he had been found in violation of SJU rules for non-
consensual sexual contact on April 13, 2018, but he was not found to
have committed the other three alleged violations. The sole basis for
the Conduct Board’s conclusion was announced in its letter:

The respondent [Roe] admitted in engaging in physical
contact of a sexual nature with the complainant [Doe],
and the evidence demonstrated a lack of affirmative
consent to engage in such contact. Such evidence
included the complainant’s intoxication, as described by
multiple witnesses, and the respondent’s assertion,
which was not disputed, that he was not impaired by
alcohol.

Id.

The Conduct Board did not specify the “physical contact of a
sexual nature” to which Roe “admitted.” Roe alleged, however, that
“the only ‘physical contact of a sexual nature’ [to which he] ‘admitted’
was that ‘Doe took his right hand and placed it upon her fully clothed
breasts.’” Id. Roe concludes, therefore, that the Conduct Board’s
finding must be based on “Doe taking his hand and placing it upon
her breast.” Id. at 19. This factual allegation is neither conclusory nor
speculative. Accordingly, “[u]nder settled principles of adjudication,

11
we must … accept the well-pleaded facts in the complaint as true.”
Washington v. Barr, 925 F.3d 109, 120 (2d Cir. 2019).

Indeed, this is exactly how the district court evaluated this case.
The district court accepted the truth of Roe’s allegations that the
Conduct Board relied on his version of events, and it dismissed his
complaint on that basis:

[T]he Conduct Board found that Plaintiff “admitted” that
he and Doe had engaged in sexual contact, and even
though Plaintiff told SJU’s investigator that Doe had
initiated that contact, the Conduct Board concluded that
Doe was unable to affirmatively consent to the contact
due to her intoxication. In other words, even assuming
that the Conduct Board accepted Plaintiff’s account of
the events, i.e., that Doe had placed Plaintiff’s hand on
Doe’s breast, the Board found that Doe did not, and
could not, have done so knowingly and voluntarily due
to intoxication, and that Plaintiff, who was “not impaired
by alcohol,” would have known that.

Roe v. St. John’s Univ., No. 19-CV-4694, 2021 WL 1224895, at *18
(E.D.N.Y. Mar. 31, 2021). The district court thus concluded that even
though the well-pleaded allegations of the complaint establish that
Doe rather than Roe initiated the contact, SJU appropriately found
Roe rather than Doe responsible for sexual misconduct because Doe
had not consented to her own conduct.

B

The district court’s reasoning is indefensible, as SJU admits on
appeal. SJU acknowledges that if the contact occurred as Roe
admitted, then Doe rather than Roe likely violated SJU’s sexual
misconduct policy. See Oral Argument Audio Recording at 22:19 (“If
12
that was the version of events, Roe’s conduct would not violate Policy
703 of the university. … In fact, perhaps [Doe] would have been in
violation of Policy 703.”).

Our inquiry into the rationale behind the Conduct Board’s
finding would normally end with the well-pleaded allegations of the
complaint. See Doe v. Amherst Coll., 238 F. Supp. 3d 195, 216 (D. Mass.
2017) (“At this stage of the litigation the court is required to credit
Doe’s reading of the Hearing Board’s decision on this issue, provided
… the reading is plausible.”). In this case, the complaint plausibly
alleged that the Conduct Board accepted Roe’s admitted version of
events, that those events involved no wrongdoing, and that the
Conduct Board disciplined him anyway. Perhaps the Conduct
Board’s decision could be explained by further evidence, but that
must await later stages of the litigation. See Sabir v. Williams, 52 F.4th
51, 65 (2d Cir. 2022) (Sack, J.) (“[A]lthough the facts at trial or
summary judgment might show otherwise, we cannot manufacture
such facts out of thin air.”) (internal quotation marks and alteration
omitted); see also Cronin v. Aetna Life Ins. Co., 46 F.3d 196, 206 (2d Cir.
1995) (“[The defendant’s] characterization [of a letter] might, if
proffered through sworn testimony, raise questions of fact to be
resolved at trial. It does not, however, provide a basis for entry of
judgment in [the defendant’s] favor as a matter of law.”).

Even if there were a possible explanation for the Conduct
Board’s decision aside from sex-based bias, moreover, that
explanation would again need to await future stages of the litigation.
See Sassaman v. Gamache, 566 F.3d 307, 313 (2d Cir. 2009) (“The choice
between plausible interpretations … is a question of fact to be
resolved by a jury.”). For two reasons, Roe’s allegations about the Doe
incident support “a minimal plausible inference” that Roe was

13
“subjected to discrimination on account of sex in the imposition of
university discipline.” Columbia Univ., 831 F.3d at 56.

First, the Conduct Board’s reasoning indicates that it treated
Doe more favorably than Roe. See Littlejohn, 795 F.3d at 312 (“An
inference of discrimination can arise from circumstances including …
the more favorable treatment of employees not in the protected
group.”). The Conduct Board decided that Doe could not consent
because she was intoxicated, and it held Roe responsible on that basis.
But the Conduct Board disregarded the fact that Roe also did not
consent to the contact. In doing so, the Conduct Board treated Doe’s
lack of consent more favorably than Roe’s. Such differential treatment
supports an inference of sex-based bias. See Doe v. Rollins Coll., 352
F. Supp. 3d 1205, 1211 (M.D. Fla. 2019) (holding that allegations
plausibly state a claim for sex discrimination when the information
“collected during the investigation could have equally supported
disciplinary proceedings against Jane Roe for also violating the Sexual
Misconduct Policy”); Doe v. Univ. of Scis., 961 F.3d 203, 210 (3d Cir.
2020) (holding that allegations plausibly state a claim of sex
discrimination when the university identified the male “as the
initiator of sexual activity, notwithstanding the comparable
intoxication of both participants”). 7

7 The court sidesteps this argument, suggesting that the disparate
treatment at issue in this case is that “SJU suspended Roe and not Doe.”
Ante at 26. But SJU treated Roe and Doe differently when it decided that
Doe did not consent because she was intoxicated, disregarded Roe’s own
lack of consent, and held Roe responsible for the contact. Both Roe and Doe
were similarly situated in that they both did not consent to the contact. The
court also asserts that this dissent relies only on SJU’s justification as
creating an inference of sex-based bias, see id. at 25, but I have just explained
that the complaint also alleged differential treatment.

14
 Second, the Conduct Board’s otherwise “inexplicable” decision
supports an inference of sex-based bias. Oberlin Coll., 963 F.3d at 588.
Such an inference is reasonable because “the more outrageous the
decision is, the less likely it is that any errors were made in good
faith,” and “when the erroneous decision ceases to be consistent with
good-faith mistake, the explanation of improper bias becomes
sufficiently likely to cross ‘the line between possibility and
plausibility.’” Samford Univ., 29 F.4th at 690 (quoting Twombly, 550
U.S. at 557).

C

The court avoids the conclusion that Roe has stated a claim by
making factual inferences in favor of SJU. The Conduct Board’s
justification for its decision may have been “faulty” and “baseless,”
the court says, but the faulty decision-making might have been the
result of something other than sex-based discrimination. Ante at 20-
21. Because Roe does not have specific evidence “suggesting that
gender bias was a motivating factor behind the erroneous findings,”
the court decides that its own alternative explanation must prevail
and the complaint must be dismissed. Id. at 19 (alteration omitted)
(quoting Yusuf, 35 F.3d at 715).

The court hypothesizes that SJU’s faulty decision may have an
explanation that would not violate Title IX. Specifically, the court
speculates that SJU may have “accepted Roe’s concession that he
engaged in sexual contact with Doe but did not credit his assertion
that Doe initiated the contact.” Id. at 21. Nothing in the complaint or
in SJU’s letter explaining its decision supports this theory of the
university’s decision, so the court posits that SJU must have relied on
a justification for its decision that it never articulated, such as
“unstated credibility assessments.” Id. at 22 n.11.

15
 The court identifies no allegations in the complaint that support
this or any other alternative explanation. Drawing such inferences
violates the requirement that, in reviewing the grant of a motion to
dismiss, we “draw all reasonable inferences in the plaintiffs’ favor.” In
re Platinum and Palladium Antitrust Lit., 61 F.4th 242, 258 (2d Cir. 2023)
(emphasis added) (quoting In re Aluminum Warehousing Antitrust
Litig., 833 F.3d 151, 157 (2d Cir. 2016)). Instead of making inferences
in Roe’s favor, the court “manufacture[s] … facts out of thin air” to
dismiss Roe’s complaint. Sabir, 52 F.4th at 65 (quoting Salahuddin v.
Goord, 467 F.3d 263, 275 (2d Cir. 2006)).

In doing so, the court supposes that SJU’s otherwise
inexplicable decision might have been motivated by something other
than a “bias against men.” Ante at 24 (quoting Doe v. Univ. of S. Ind.,
43 F.4th 784, 797 (7th Cir. 2022)). This speculation conflicts with the
allegations in the complaint, which suggest that SJU was biased
against men. For example, the complaint alleged that SJU was
criticized by “over 2,000 women” for “allowing accused men to remain
on campus.” App’x 24 (emphasis added). By raising the specific issue
of sexual misconduct by men, Roe pleads “minimal evidence of
pressure on the university to act based on invidious stereotypes,”
which we have said “will permit a plausible inference of sex
discrimination.” Menaker, 935 F.3d at 33.

D

The court relies on cases from other circuits that suggest that a
university may, consistent with Title IX, impose student discipline
based on discriminatory procedures that reflect “a pro-complainant,
anti-respondent bias.” Samford Univ., 29 F.4th at 690. Under this view,
“discrimination against respondents is not discrimination on the basis
of sex and does not permit a reasonable inference of an anti-male bias

16
because both men and women can be respondents.” Id. (internal
quotation marks and citations omitted). This view is unpersuasive
because an anti-respondent bias is a sex-based bias. See Oberlin Coll.,
963 F.3d at 587 (“[T]he 100 percent responsibility rate—in cases where
most if not all the respondents were male—supports an inference
regarding bias in the hearings themselves.”); Doe v. Miami Univ., 882
F.3d 579, 593 (6th Cir. 2018) (deciding that “the statistical evidence …
shows a pattern of gender-based decision-making” in part because
“nearly ninety percent of students found responsible for sexual
misconduct … have male first names”).

It is true that “both men and women can be respondents.” But
both men and women can be complainants too—and yet we would
never say that a university may escape liability for deliberate
indifference to sexual assault because it may have been motivated by
an anti-complainant bias instead of an anti-female bias. To the
contrary, we say that deliberate indifference to cases of sexual assault
by the university ratifies the perpetrator’s targeting of the
complainant on the basis of sex, and that satisfies the requirement of
demonstrating a sex-based bias. See Davis, 526 U.S. at 644-45 (“If a
funding recipient does not engage in harassment directly, it may not
be liable for damages unless its deliberate indifference … cause[s]
students to undergo harassment or make them liable or vulnerable to
it.”) (internal quotation marks and alteration omitted); Folkerson v.
Circus Circus Enters., Inc., 107 F.3d 754, 756 (9th Cir. 1997) (explaining
that an employer will be liable under Title VII “for sexual harassment
on the part of a private individual” when “the employer either ratifies
or acquiesces in the harassment by not taking immediate and/or

17
corrective actions when it knew or should have known of the
conduct”). 8

We have held that accusations of sexual misconduct work the
same way. In Menaker, we explained that a complainant’s decision to
accuse a respondent of “sexual misconduct” is “significant, and it
suggests that [the respondent’s] sex played a part in [the
complainant’s] allegations,” such that a “rational finder of fact” could
“infer that such an accusation was based, at least in part, on … sex.”
935 F.3d at 38. In turn, the complainant’s sex-based “intent may be
imputed to [the university]” when the university “controlled … the
very complaint process by which [the complainant] sought to
effectuate her allegedly discriminatory intent.” Id. at 38-39. “[I]nsofar
as [the university] negligently or recklessly implemented [the
complainant’s] discriminatory design”—as evidenced by “procedural
irregularities” in the complaint process—the university has
discriminated on the basis of sex. Id. at 39.

Under the alternative approach of some courts, however, a
university’s treatment of a complainant or respondent will not
implicate Title IX if it could have been the product of “anti-
complainant” or “anti-respondent” bias. If a university may be
excused from biased treatment of a respondent because “both men
and women can be” respondents, Samford Univ., 29 F.4th at 690
(quoting Doe v. Univ. of Denver, 952 F.3d 1182, 1197 (10th Cir. 2020)),
then a university may also be excused from acting with deliberate

8 See also Catherine A. MacKinnon, In Their Hands: Restoring Institutional
Liability for Sexual Harassment in Education, 125 Yale L.J. 2038, 2042-43 (2016)
(“As the accounts and data below demonstrate, sexual harassment is
gender-based because it is directed against a woman because she is a
woman or affects women disproportionately.”) (internal quotation marks
and alteration omitted).

18
indifference to sexual assault because “both men and women can be”
victims of sexual assault as well, id. This approach departs from the
established case law in this context. See, e.g., Davis, 526 U.S. at 643
(holding that “deliberate indifference to known acts of harassment …
amounts to an intentional violation of Title IX”).

The precedents of our circuit do not allow a complaint alleging
improper discipline to be dismissed on the ground that the university
may have implemented the complainant’s discriminatory design for
its own non-sex-based reasons. As in Menaker, in this case the
complaint plausibly alleged that the complainant made a sex-based
accusation against the respondent, and a factfinder “could plausibly
conclude that [the university] was negligent or reckless in acting on
[the complainant’s] allegations” in violation of Title IX. 935 F.3d at 39.

If a university could escape liability under Title IX because it
was biased only against “respondents” rather than men, our case law
would look very different. In Columbia University, we concluded that
the plaintiff had pleaded actions “motivated … by pro-female, anti-
male bias” because the administrators “declined to seek out potential
witnesses [whom the] Plaintiff had identified as sources of
information favorable to him” and violated “procedures designed to
protect accused students.” 831 F.3d at 56-57. Such conduct by the
administrators was also consistent with an “anti-respondent” bias.
But we did not accept such an explanation, observing instead that “[i]t
is not the court’s function in ruling on a motion to dismiss for
insufficiency of the complaint to decide which was the defendant’s
true motivation.” Id. at 56 n.10. Columbia argued that the allegations
did “not support an inference of intentional sex discrimination”
because “any motivation on the part of the [disciplinary] panel to
demonstrate that it takes [sexual assault] complaints seriously is not
the same thing as a motivation to discriminate against an accused

19
male student.” Id. at 57. In other words, Columbia argued that the
university was biased against the respondent for a reason other than
sex. We rejected that argument because it “fails to recognize the
court’s obligation to draw reasonable inferences in favor of the
sufficiency of the complaint. Iqbal does not require that the inference
of discriminatory intent supported by the pleaded facts be the most
plausible explanation of the defendant’s conduct. It is sufficient if the
inference of discriminatory intent is plausible.” Id.

We applied identical principles in Menaker and Vengalattore. In
neither case did we ask whether the university’s behavior could be
explained by an alternative non-sex-based bias such as “anti-
respondent bias.” In every case alleging that a respondent was treated
unfairly because of sex, the university’s conduct could also be
explained by “anti-respondent bias.” If that were a reason for
dismissing the complaint, then Columbia University, Menaker, and
Vengalattore were all wrongly decided.

E

Today’s decision heightens the Title IX pleading standard. In
Columbia University we said that the alleged procedural irregularities
did “not necessarily relate to bias on account of sex” and that other
factual allegations gave “ample plausible support to a bias with
respect to sex.” Id. (emphasis added). But we did not establish a rigid
pleading requirement according to which plaintiffs must rebut
alternative possible explanations or establish more than “a minimal
plausible inference” that the plaintiff was “subjected to
discrimination on account of sex in the imposition of university
discipline.” Id. at 56. We did not hold that every plausible claim must
be alleged the same way. Nor did we foreclose the possibility that we
would join those other circuits that have recognized that the merits of

20
a university’s decision may support an inference of sex-based bias.
See, e.g., Purdue Univ., 928 F.3d at 669; Oberlin Coll., 963 F.3d at 588;
Samford Univ., 29 F.4th at 690. According to those courts, an erroneous
decision that is “inexplicable,” Oberlin Coll., 963 F.3d at 588, or
“perplexing,” Purdue Univ., 928 F.3d at 669, can raise an inference of
sex-based bias. Unlike the plaintiff in Columbia University, Roe has
alleged that the Conduct Board’s reasoning, on its face, is inexplicable
except by reason of bias. The court’s decision today creates a split with
those circuits that have held that such a decision raises an inference
of sex-based bias.

Moreover, in Menaker we expressly declined to “define
precisely what sort of irregularities meet the standard of ‘clearly
irregular investigative or adjudicative process’” and thereby made
clear that we have not limited plaintiffs to alleging a particular factual
scenario. 935 F.3d at 34. 9 We emphasized that when a school
“accept[s] an unsupported accusatory version” of events “over that of
the accused,” it supports an inference of bias. 935 F.3d at 34 (alteration
omitted) (quoting Columbia Univ., 831 F.3d at 57). Roe did not allege
that SJU accepted Doe’s unsupported version of events over his. He
alleged something even more inexplicable absent bias: that SJU
accepted Roe’s version of events—according to which he was not a

9
For this reason, our role in Title IX student discipline cases is not limited
to determining whether the evidence substantially favored Roe but the
university inexplicably favored Doe. See Columbia Univ., 831 F.3d at 57.
Students may be the victims of discrimination even when the evidentiary
record is thin. We must determine whether the allegations support a
minimal inference of sex-based discrimination by the defendant. Roe’s
allegations focus on the Conduct Board’s reasoning, not the weight of the
evidence, and the reasoning supports an inference of bias on the part of SJU.

21
perpetrator but a victim—and nevertheless found that he had
committed a violation.

In the Title VII context, we have never restricted plaintiffs to
alleging a specific factual scenario. Rather, “we routinely look to a
wide variety of (often subtle) indications that a consideration
prohibited by Title VII played a role in the employer’s conduct.”
Gregory v. Daly, 243 F.3d 687, 695 (2d Cir. 2001); cf. Furnco Constr. Corp.
v. Waters, 438 U.S. 567, 577 (1978) (“The method suggested in
McDonnell Douglas for pursuing this inquiry … was never intended to
be rigid, mechanized, or ritualistic.”). For this reason, a plaintiff may
support an inference of discrimination in many ways, including but
not limited to “the more favorable treatment of employees not in the
protected group[] or the sequence of events leading to the plaintiff’s
discharge.” Littlejohn, 795 F.3d at 312 (internal quotation marks
omitted). A plaintiff may also allege a “‘mosaic’ of intentional
discrimination by identifying ‘bits and pieces of evidence’ that
together give rise to an inference of discrimination.” Vega v. Hempstead
Union Free Sch. Dist., 801 F.3d 72, 87 (2d Cir. 2015) (quoting Gallagher
v. Delaney, 139 F.3d 338, 342 (2d Cir. 1998)). We should not treat Title
IX claims differently. “[W]e have long interpreted Title IX by looking
to the caselaw interpreting Title VII.” Vengalattore, 36 F.4th at 103
(internal quotation marks and alterations omitted); see also Columbia
Univ., 831 F.3d at 55-56 (“Title VII cases provide the proper
framework for analyzing Title IX discrimination claims.”). 10

10 I note that some circumstances might implicate relevant distinctions
between Title VII and Title IX. See, e.g., Jackson v. Birmingham Bd. of Educ.,
544 U.S. 167, 175 (2005) (describing differences in statutory language); Soule
v. Conn. Ass'n of Schs., Inc., 90 F.4th 34, 63 (2d Cir. 2023) (Menashi, J.,
concurring) (noting “important differences between” Title VII and Title IX);

22
 Today’s decision introduces two errors into our Title IX case
law. First, it departs from the straightforward pleading standard of
Columbia University that a Title IX plaintiff alleging improper
discipline must plead “specific facts that support a minimal plausible
inference” of discrimination. 831 F.3d at 56. Instead, the court reverts
to Yusuf in requiring such a plaintiff to plead facts that track the
erroneous outcome and selective enforcement tests.

Second, the court’s decision imposes a new heightened
pleading requirement for Title IX plaintiffs. The plaintiff must
preemptively refute alternative explanations of biased treatment by
pleading “additional allegations of fact indicating bias on account of
sex.” Ante at 20. In the court’s view, alleged treatment that involves
an erroneous conclusion, flawed procedures, and a baseless
rationale—even in the context of a sexual misconduct adjudication—
does not “support a minimal plausible inference of sex
discrimination.” Id. at 39. In other words, a Title IX plaintiff must
plead “direct, smoking gun, evidence of discrimination.” Vega, 801
F.3d at 86 (quoting Richards v. N.Y.C. Bd. of Educ., 668 F. Supp. 259, 265
(S.D.N.Y. 1987)). We long ago rejected such a pleading requirement
because discrimination is “elusive,” Tex. Dep’t of Cmty. Affs. v. Burdine,
450 U.S. 248, 255 n.8 (1981), and “clever men may easily conceal their
motivations,” Robinson v. 12 Lofts Realty, Inc., 610 F.2d 1032, 1043 (2d
Cir. 1979) (quoting United States v. City of Black Jack, 508 F.2d 1179,
1185 (8th Cir. 1974)).

Today’s decision provides cover for clever university
administrators. Going forward, a Title IX defendant may cite today’s

Adams, 57 F.4th at 811 (“Title IX, unlike Title VII, includes express statutory
and regulatory carve-outs for … separate living and bathroom facilities.”).
This case does not involve those circumstances.

23
decision to argue that as long as its irregular treatment of the plaintiff
might be explained on some alternative ground—even one
unsupported by allegations in the complaint—the complaint must be
dismissed unless the plaintiff has direct evidence that the irregular
treatment was based specifically on a protected characteristic. That
will prevent plaintiffs from stating otherwise meritorious claims.

III

The circumstances surrounding Roe’s expulsion also allow for
a minimal plausible inference of discrimination. This conclusion
follows from “the general principle that where a university (1) takes
an adverse action against a student or employee, (2) in response to
allegations of sexual misconduct, (3) following a clearly irregular
investigative or adjudicative process, (4) amid criticism for reacting
inadequately to allegations of sexual misconduct by members of one
sex, these circumstances provide the requisite support for a prima facie
case of sex discrimination.” Menaker, 935 F.3d at 33. Roe plausibly
alleged that SJU experienced “at least some pressure … to react more
forcefully to allegations of male sexual misconduct” and that his
expulsion followed a “sufficiently irregular process” that “raise[s] an
inference of bias.” Id. at 34.

A

Roe alleged that his expulsion followed a “tweet storm” that
criticized SJU for its handling of sexual assault complaints. On
January 4, 2019, the hashtag “#SurvivingSJU” appeared on Twitter
and became the top trending hashtag for the next two days. App’x 15.
Over 2,000 tweets “detail[ed] alleged experiences of sexual
misconduct that had taken place at St. John’s” and “criticiz[ed] St.
John’s treatment of female complainants of sexual assault.” Id. at 17.
One of those tweets criticized SJU specifically for mishandling the

24
Doe complaint. That tweet read: “[Roe] was allowed to stay abroad
after raping me with no travel restrictions. Only got half a semester
suspension.” Id. at 72. The tweet faulted SJU for its failure to impose
harsher interim relief and a harsher ultimate sanction on Roe. On
January 8, SJU’s student newspaper The Torch reported on the tweet
storm and SJU’s public statement in response that it would
“investigate all claims.” Id. at 74.

The tweet storm “provides a backdrop that, when combined
with other circumstantial evidence of bias in [a] specific proceeding,
gives rise to a plausible claim.” Purdue Univ., 928 F.3d at 669. We have
said that, when combined with procedural irregularity, “even
minimal evidence of sex-based pressure on the university is sufficient
to establish bias on account of sex.” Menaker, 935 F.3d at 33 n.48.

As in Menaker, the university in this case faced sustained
“internal criticism for its assertedly inadequate response to male
sexual misconduct on campus.” Id. at 27. The pressure in this case is
even more acute than in Menaker because Roe was specifically
identified as part of the campaign of criticism. The anonymous tweet
included Roe’s name and picture, accused him of rape, and
condemned SJU for failing to punish him adequately. Roe alleged that
one day after the tweet, a male student at SJU called Roe and sent text
messages threatening to “fuck [Roe] up.” App’x 16. He alleged that a
female student “punched [him] in his face at a bar.” Id. at 17. Treating
these allegations as true and making all reasonable inferences in Roe’s
favor—as, again, we must on a motion to dismiss—it is plausible that
the tweet accusing Roe of rape was viewed by SJU students who were
upset by Roe’s alleged behavior and SJU’s response. It is difficult to
imagine that SJU did not face even “minimal” pressure as a result.
Menaker, 935 F.3d at 33.

25
 The district court discounted the allegations of pressure
because Roe did not allege that SJU “continued to experience any such
scrutiny when it conducted the Smith hearing over nine months” after
the tweet storm. 2021 WL 1224895, at *22. In Menaker, however,
fourteen months passed between public criticism of the university, see
935 F.3d at 27 (“By May 2015, the national press had identified Hofstra
as one of several universities under investigation … for possible
mishandling of sexual misconduct claims.”), and the university’s
consideration of the relevant Title IX complaint, see id. at 28 (noting
that the university first “summoned Menaker to a meeting” shortly
“after receiving the July 2016 Kaplan Letter”). We did not hold that
the public pressure needed to remain constant throughout the
university’s adjudication of the claim. See id. at 33 (“[W]e reject the
District Court’s attempt to limit Doe v. Columbia to cases where the
public pressure on a university is particularly acute.”).

But even if we did require acute public pressure, Roe’s
allegations support a plausible inference that “such pressure …
affect[ed] how the University treated [him] in sexual misconduct
disciplinary proceedings on the basis of sex.” Schwake, 967 F.3d at 948.
SJU issued Roe an interim suspension on January 8, 2019—just four
days after the hashtag began trending and the anonymous tweet was
publicized. The suspension not only marked the start of the
disciplinary process that led to Roe’s expulsion but was itself a denial
of “the benefits of … any education program or activity” under
Title IX. 20 U.S.C. § 1681(a). The temporal proximity between public
pressure and the imposition of discipline is much closer in this case
than in Menaker. Moreover, unlike the criticism in Columbia University
and Menaker, the criticism here addressed SJU’s treatment of Roe
specifically. That means there is a plausible basis for inferring that the
pressure put on SJU in January 2019 affected its decisions to suspend

26
Roe four days later and ultimately to expel him nine months later. The
court does not disagree. See ante at 28-29 (“Roe’s allegations about the
‘tweet storm’ … do weigh in favor of Roe’s theory of sex
discrimination.”).

B

“The only remaining question … is whether [Roe’s expulsion]
followed a sufficiently irregular process to raise an inference of bias.”
Menaker, 935 F.3d at 34. Here, Roe has pleaded facts that, when taken
as true, reflect an irregular investigative and adjudicative process.
These irregularities support an inference of bias at every stage of the
Smith proceedings. “[A]t some point an accumulation of procedural
irregularities all disfavoring a male respondent begins to look like a
biased proceeding.” Doe v. Regents of Univ. of Cal., 23 F.4th 930, 941
(9th Cir. 2022). So it is here. 11

First, less than twenty-four hours after issuing the no-contact
order—which would have allowed Roe to return to campus but not
to contact Smith—SJU altered course and suspended Roe without a
hearing, an investigation, or even a description of the rules he had

11 The court refuses “to consider” some “procedural irregularities” that it
believes were not specifically mentioned in the appellate briefing. Ante at
32 n.16. But those procedural irregularities are facts alleged in Roe’s
complaint. See App’x 15 (impact statement); id. at 28-30 (timeline of Smith
complaint until expulsion). The court’s insistence that Roe has waived facts
alleged in his complaint violates the obligation on a motion to dismiss to
take the complaint’s “factual allegations to be true” and to determine
whether those facts “state[] a plausible claim for relief.” Harris v. Mills, 572
F.3d 66, 71-72 (2d Cir. 2009).

27
allegedly violated. See App’x 17-18. 12 Against the backdrop of the
pressure campaign, the decision to suspend Roe before investigating
Smith’s allegations supports an inference that the university
presumed that Roe was responsible. Title IX, like Title VII, requires
that “in the course of investigating [sexual misconduct] claims,”
universities “do not presume male [students] to be ‘guilty until
proven innocent’ based on invidious sex stereotypes.” Sassaman, 566
F.3d at 314. “[J]ust as the lack of investigation of a reported claim of
harassment may factor into the determination of an employer’s
liability for discrimination against the complainant, so too may it
indicate discrimination by an employer whose adverse determination
against the putative harasser otherwise bears indicia of prohibited
discrimination.” Id. at 315.

The suspension, moreover, represents a departure from SJU’s
Students’ Bill of Rights, which affords students the right to
“[p]articipate in a process that is fair, impartial, and provides
adequate notice and a meaningful opportunity to be heard.” App’x
117; see also Doe v. Baum, 903 F.3d 575, 581 (6th Cir. 2018) (holding that
an opportunity to be heard requires, among other things, that “if a
student is accused of misconduct, the university must hold some sort
of hearing before imposing a sanction as serious as expulsion or
suspension”). The failure to provide Roe adequate notice or an
opportunity to be heard before issuing the suspension supports an

12 SJU argues that the exhibits attached to Roe’s complaint demonstrate that
“[i]t is patently false for Roe to claim that he was not provided with notice
of the grounds for his suspension in January 2019.” Appellee’s Br. 32. But
Roe was informed only that he was suspended “pending the results of a
non-academic disciplinary investigation.” App’x 87. SJU did not inform
Roe of the allegations against him, the nature of the allegations, or why the
suspension was necessary.

28
inference of bias. See Doe v. William Marsh Rice Univ., 67 F.4th 702, 708
(5th Cir. 2023) (explaining that the university’s treatment of the
respondent was “deficient” in several ways, including that the
respondent was “prohibited from entering campus with only 24-
hours’ notice” and prior to having a “reasonable opportunity to
present his side of the story with the advice of counsel”).

The court responds that the interim suspension “was consistent
with SJU’s internal procedures” because the Student Code of Conduct
says that “[a]t any time” and “at the discretion” of university officials,
a student may be issued an “interim action[]” such as “an interim
suspension.” Ante at 30-31 (quoting App’x 223). Yet a university
cannot insulate itself from allegations that it behaved in a
discriminatory fashion simply by writing policies that give it
unfettered discretion to impose any sanction at any time. We would
never give an employer accused of discrimination such deference in
the imposition of discipline. Cf. Lauture v. IBM Corp., 216 F.3d 258, 262
(2d Cir. 2000) (“In the context of a racial discrimination claim brought
under federal law, the fact that employment was at will is simply not
dispositive.”).

Moreover, the court does not acknowledge that SJU also
imposes a condition precedent on interim suspensions. SJU’s “You
Are Not Alone” pamphlet explains that SJU will impose an interim
suspension only if the accused student “presents a continuing threat
to the health and safety of the community or the complainant.” App’x
120. 13 It is irregular, therefore, for SJU to impose an interim

13 The “You Are Not Alone” pamphlet provides students with “important
information about prohibited conduct, available resources on and off
campus, and ways to file a complaint” related to sexual misconduct. App’x
99. The pamphlet purports to be a student’s “first step” in understanding

29
suspension without making a finding of dangerousness. There is no
indication in the complaint or in the record that SJU determined that
Roe posed a “continuing threat” to the community prior to imposing
an interim suspension.

Second, SJU did not provide Roe with information about
Smith’s complaint according to the timeframe its policies require.
Policy 703—the applicable SJU policy for “any allegation of sexual
misconduct,” App’x 45—provides that “in most cases complaints will
be resolved within 60 days,” id. at 63. 14 And the Student Conduct

how SJU will resolve allegations of sexual misconduct. Id. at 100. SJU’s
Policy 703 states that students “should refer” to the “You Are Not Alone”
pamphlet to understand the resources available in that process. Id. at 52.
14 The district court determined that Roe’s “reliance on Policy 703 is
misplaced because the Policy makes clear that complaints against students
are investigated and adjudicated pursuant to procedures outlined in the
Student Code.” 2021 WL 1224895, at *22. But Roe alleged that in
“investigating” Doe’s and Smith’s complaints, SJU was “required to
follow” Policy 703. App’x 13. The court fails to credit that allegation on a
motion to dismiss, even though whether and to what extent Policy 703
applied is a disputed factual issue. The district court concluded that Policy
703 did not apply because the “[a]djudication” subsection of Policy 703
states that “[t]he applicable procedure for remedying a complaint depends
on whether the accused is a student, member of the faculty, or staff or
administrator.” Id. at 65. But Policy 703 explains that the only procedures
applicable specifically to faculty and staff are found in Section VIII, titled
“Adjudication of Complaints Against Faculty, Administrators and Staff.”
Id. at 66. The other sections of Policy 703 apply to students as well as faculty.
Those sections include the process by which a Title IX coordinator
investigates a complaint of sexual misconduct, including the
implementation of interim remedies. Id. at 61-64. For that reason, when SJU
notified Roe of his interim suspension, it “advise[d] [him] to review” the
Student Bill of Rights and Policy 703 “for all the information [he] may need.”
Id. at 89 (emphasis added). SJU did not reference the Student Code of

30
Process, which forms a part of SJU’s Student Code of Conduct,
provides that “hearings may be conducted within ninety (90) days of
when the incident was documented.” App’x 227. It took SJU six
months to inform Roe of the allegations and another three months to
hold a hearing. See Oberlin Coll., 963 F.3d at 586-87 (concluding that a
similar delay amounted to a clear procedural irregularity). SJU’s
failure to follow its own established timetable was especially serious
because Roe was suspended from school pending that proceeding.

Third, Roe was prevented from submitting an impact statement
before the sanction was imposed, despite Smith being allowed to do
so. The district court dismissed this allegation as insignificant because
“the Student Code does not provide for the respondent to submit an
impact statement while the appropriate sanction is being decided on.”
2021 WL 1224895, at *17. The fact that the university maintained a
biased policy before applying it to Roe does not mean that the policy
does not plausibly indicate bias. The inquiry in this case is whether
the alleged facts lead to a plausible inference of bias. A university’s
departure from its own policy may indicate such bias, but the policy
itself may demonstrate bias as well. Universities cannot avoid liability
under Title IX by instituting biased policies. See Regents of Univ. of Cal.,
23 F.4th at 940 n.14 (“[E]ven if the University’s policies in place at the
time condoned these procedures, Doe is entitled to allege that such
policies were inherently problematic, as he has done here.”); see also
Schwake, 967 F.3d at 950 (“Contrary to the University’s suggestion that
there can be no showing of gender bias because University policy
foreclosed an appeal, gender bias is a plausible explanation in light of
the background indicia of sex discrimination.”). If Roe were a faculty

Conduct at all. See Menaker, 935 F.3d at 34 n.50 (explaining that whether a
process is “clearly irregular” depends on, among other things, the
“expectations of the parties”).

31
member accused of the same conduct, he would have been able to
submit an impact statement, see App’x 64, which suggests that the
university has specifically designed the student conduct procedures
in a discriminatory way to make it easier to penalize students accused
of misconduct. 15

Fourth, the Conduct Board’s decision is, on its face, perplexing.
See Oberlin Coll., 963 F.3d at 588 (“[A] ‘perplexing’ basis of decision
can support an inference of sex bias.”) (quoting Purdue Univ., 928 F.3d
at 669). The Student Conduct Process states that the “purpose” of the
hearing before the Conduct Board is “to make findings of fact with
respect to the matter before the panel.” App’x 226 (emphasis added).
The Conduct Board, however, made no findings of fact apart from its
bottom-line conclusions that Roe violated the policy prohibiting non-
consensual sexual contact but did not violate the policy prohibiting
non-consensual sexual penetration. See id. at 142-43. It is irregular for
a disciplinary board tasked with making factual findings not to make
factual findings about the incident being investigated. Without such
findings, the Conduct Board’s conclusions are especially perplexing.
Smith’s version of events was that Roe engaged in non-consensual
penetration. See id. at 28. Roe denied that he engaged in any sort of

15 See Doe v. Brandeis Univ., 177 F. Supp. 3d 561, 607 (D. Mass. 2016) (“[T]he
lowering of the standard appears to have been a deliberate choice by the
university to make cases of sexual misconduct easier to prove—and thus
more difficult to defend, both for guilty and innocent students alike. It
retained the higher standard for virtually all other forms of student
misconduct. The lower standard may thus be seen, in context, as part of an
effort to tilt the playing field against accused students, which is particularly
troublesome in light of the elimination of other basic rights of the
accused.”); cf. 34 C.F.R. § 106.45(b)(1)(vii) (noting that it violates Title IX not
to “apply the same standard of evidence for formal complaints against
students as for formal complaints against employees, including faculty”).

32
sexual contact at all. The Conduct Board did not agree with either
version of events because it concluded that Roe had violated the
policy prohibiting non-consensual sexual contact but had not violated
the policy prohibiting non-consensual sexual penetration. On what
basis did the Conduct Board conclude that neither party’s account of
the encounter was correct? The evidence cited in the Conduct Board’s
decision does not provide an explanation. 16

The Conduct Board’s perplexing conclusion supports the
plausible inference that it decided to find Roe responsible for at least
one violation “in order to avoid further negative media attention and
to portray a stricter approach to sexual assault cases.” Doe v. Univ. of
Arkansas-Fayetteville, 974 F.3d 858, 865 (8th Cir. 2020); see also Regents
of Univ. of Cal., 23 F.4th at 939 n.12 (“[One] explanation might be that,
when confronted by a claim that lacked merit, the University rushed

16
That evidence consisted of “(1) key inconsistencies in [Roe]’s testimony;
(2) the accounts of [Roe]’s housemates; and (3) [that] key aspects of [Roe]’s
account of the events lacked credibility, including his testimony that he
turned [Smith]’s face toward the couch in the event she vomited.” App’x
143. We do not know what the housemates alleged because the Conduct
Board did not say. The court defends the Conduct Board on the ground that,
because “the record does not contain all the evidence the Smith panel
considered when it reached its conclusion,” there is no “firm basis” to
conclude that the Conduct Board’s decision was perplexing. Ante at 35 n.19.
But the lack of a complete record is not a proper ground for granting a
motion to dismiss. Instead, we must credit the allegations of the complaint.
In his complaint, Roe has plausibly alleged that the Conduct Board acted
irregularly because, despite being charged with making factual findings, it
issued only legal conclusions. Perhaps a developed record would
eventually reveal that the Conduct Board relied on sufficient evidence. But
we will never see a developed record because the court has erroneously
decided, at the motion-to-dismiss stage, to affirm the dismissal of the
complaint because the record has not been developed. This is yet another
departure from the standard principles for evaluating a motion to dismiss.

33
to judgment in issuing the two-year interim suspension and then
sought out a way to find the accused responsible for something in
order to justify its earlier actions.”); Miami Univ., 882 F.3d at 593
(concluding that an “unexplained discrepancy” supported an
inference of sex-based bias).

Fifth, Roe alleged that the Appeals Board engaged in erroneous
appellate factfinding that contradicted the conclusions of the Conduct
Board. The Appeals Board stated that Roe was found to have
“digitally penetrated [Smith’s] vagina while she was asleep at an
off-campus private house.” App’x 152. But the Conduct Board never
made such a finding. In fact, the Conduct Board found that Roe did
not engage in non-consensual sexual penetration. See id. at 30.
Referring to the Doe proceedings, the Appeals Board observed that
Roe “twice committed violations of a sexual nature” in which the
victim was “unconscious.” Id. at 157; see also id. at 153 (describing the
Doe case as involving “an unconscious victim”). But at no point did
the Conduct Board—in its evaluation of either the Doe complaint or
the Smith complaint—find that the complainant was “unconscious.”
The Appeals Board relied on these allegedly erroneous findings to
affirm Roe’s expulsion. The Appeals Board concluded that “the
sanction in this case—expulsion—was the only one justified” because
“Roe has now a demonstrable history of preying on women while
they were in a vulnerable state.” Id. at 157. 17

The court acknowledges that the Appeals Board’s factfinding is
a procedural irregularity. But the court announces that it does “not

17
Doe claimed that Roe engaged in digital penetration while she was
asleep, but the complaint plausibly alleged that the Conduct Board did not
credit those allegations. Perhaps “the facts at trial or summary judgment
might show otherwise,” but the plausible allegations at this stage support
an inference that the Appeals Board acted irregularly. Sabir, 52 F.4th at 65.

34
consider the Appeals Board’s statement regarding the evidence
against Roe to be a major procedural irregularity” and therefore the
irregularity is “not serious enough to support a claim of sex-based
discrimination.” Ante at 35-36 (emphases added). The court reasons
that the Appeals Board “recognized that the Smith panel determined
that Roe violated only the policy against non-consensual sexual
contact,” so its factual statement to the contrary was a minor
irregularity. Id. at 35. But that is backwards. The fact that the Appeals
Board recognized that the Conduct Board did not find that Roe
violated the policy against non-consensual sexual penetration makes
it more perplexing that the Appeals Board went on to rely on its own
improper finding that Roe did engage in such conduct.

In sum, the factfinding body (the Conduct Board) failed to find
facts but nevertheless concluded that Roe violated school policy and
expelled him. Then the appellate body (the Appeals Board)
improperly found its own new facts in order to ratify the expulsion.
The process reflects “the philosophy of Lewis Carroll’s Queen of
Hearts: ‘Sentence first—verdict afterwards.’ This is the opposite of
procedural regularity.” Menaker, 935 F.3d at 36 (footnote omitted)
(quoting Lewis Carroll, Alice’s Adventures in Wonderland (1865)).

Sixth, Roe alleged that SJU “failed to undertake an ‘impartial’
investigation” of the tweet that accused him of sexual assault after he
filed a complaint with the university. App’x 22. That failure, he
argues, “stands in stark contrast with St. John’s treatment of Smith’s
complaint.” Appellant’s Br. 23. The differential treatment of
complaints supports an inference of sex-based bias. See Doe v.
Princeton Univ., 30 F.4th 335, 344 (3d Cir. 2022) (“Doe has plausibly
alleged that he reported a violation that was not investigated by the
University. And that, in turn, plausibly supports the inference that sex
was a motivating factor in Princeton’s investigation.”); Amherst Coll.,

35
238 F. Supp. 3d at 218 (“[S]pecific factual allegations that the College
responded differently to similar reports when the genders of the
potential victims and aggressors were different … provide a
foundation from which a court can infer gender-based discrimination
may have played a role in the College’s responses.”). 18 Roe plausibly
alleged that “[a]bsent the sexual misconduct proceeding and the
alleged pressure that the University faced regarding its handling of
sexual misconduct complaints,” Roe’s complaint would have been
taken seriously. Schwake, 967 F.3d at 950-51. The failure to address
Roe’s complaint—coupled with the alleged pressure on SJU—
supports a plausible claim of sex discrimination. See Univ. of Scis., 961
F.3d at 211 (“[W]hen Doe’s allegations about selective investigation
and enforcement are combined with his allegations related to
pressure applied by the 2011 Dear Colleague Letter, we conclude that
he states a plausible claim of sex discrimination.”).

The court insists that “Roe and both Doe and Smith were not …
similarly situated” because “Roe accused Doe of anonymously
publishing an allegedly harassing, defamatory tweet” whereas “Doe
and Smith accused Roe of sexual assault.” Ante at 37. But Roe did not
allege that SJU investigated his complaint less “thoroughly.” Id. He
alleged that SJU “failed to undertake an ‘impartial’ investigation” at
all, even though such an investigation was required by SJU policy and

18
See also Doe v. Syracuse Univ., 341 F. Supp. 3d 125, 139 (N.D.N.Y. 2018)
(holding that allegations that the university sua sponte initiated sexual
assault proceedings against the male plaintiff, but not against a female
student, based on an incident in which both parties were intoxicated, state
a plausible selective enforcement claim).

36
Title IX. App’x 22. 19 SJU stated that “the University cannot sanction
the individual who posted the tweet because we cannot confirm their
identity.” App’x 22. But that is not a reason to forgo an investigation;
it is a reason to undertake one. Had SJU taken Roe’s complaint of
harassment seriously, it could have at least inquired of Doe whether
she sent the tweet, as Roe had alleged. Had its investigation yielded
no answers, perhaps then SJU could have concluded it was not
possible to confirm the sender’s identity. The court may be correct
that the seriousness of the complaints against Roe justified a greater
investigative effort than did Roe’s complaint against Doe. But that
does not justify SJU’s decision to undertake no investigation
whatsoever into Roe’s complaint. It is at least plausible that the refusal
to undertake an investigation indicates bias.

Seventh, Roe plausibly alleged that Director of Student
Conduct Jack Flynn tainted the Conduct Board’s decision. We have
held that an “impermissible bias of a single individual at any stage of
the … process may taint the ultimate [adverse] decision even absent
evidence of illegitimate bias on the part of the ultimate decision
maker, so long as the individual shown to have the impermissible bias
played a meaningful role in the process.” Back v. Hastings on Hudson
Union Free Sch. Dist., 365 F.3d 107, 125-26 (2d Cir. 2004) (alterations
omitted) (quoting Bickerstaff v. Vassar Coll., 196 F.3d 435, 450 (2d Cir.
1999)); see also Holcomb v. Iona Coll., 521 F.3d 130, 143 (2d Cir. 2008). If
a biased employee “influenc[ed] the adverse action by a non-biased

19 Roe notes that SJU’s director of Title IX compliance, Keaton Wong,
responded to his email, which might be understood as the first step in an
investigation. But Wong’s email reflected no interest in investigating. Wong
simply thanked Roe for his email, disclaimed any chance of identifying the
sender of the tweet, and forwarded Roe a copy of the “You Are Not Alone”
pamphlet. App’x 22.

37
decision-maker,” then the adverse decision is actionable. Columbia
Univ., 831 F.3d at 59. The criticisms that SJU faced during the tweet
storm implicated Flynn as the person who decided on interim actions
and sanctions. Under these circumstances, “[i]t is plausible that
[Flynn] was motivated to refute those criticisms by siding with the
accusing female and against the accused male.” Id. at 58. We have
previously held that a complaint states a plausible Title IX claim when
it alleges that an administrator who “had suffered personal criticism
in the student body for her role in prior cases in which the University
was seen as not taking seriously the complaints of female students”
had “significant influence, perhaps even determinative influence,
over the University’s decision.” Id. Roe alleged that here.

Roe alleged that Flynn “suspend[ed] Roe before
investigati[ng]” the Smith complaint “without due explanation.”
App’x 33. Roe further alleged that Flynn then manipulated the
conduct of the hearings to justify his suspension decision: Flynn
allegedly “[s]elect[ed] the members of both appeal and hearing
panels,” id.; “[c]oach[ed] the investigator to testify adversely [against]
Roe,” id. at 34; “conferred with the impartial investigator … and
members of the [Conduct Board] panel, both in advance of the
hearing and thereafter,” id. at 30; “[a]sk[ed] … to expedite the …
appeal decision due to pressure of over 2,000 tweets by female
students against St. John’s,” id. at 33; and “[c]onspir[ed]” with a
member of the Appeals Board to disallow Roe from submitting a
written impact statement, id. at 34. The fact that Flynn was
“responsible to ensure that all student conduct proceedings are
carried out in accordance with the Student Conduct Process,” id. at
155, is enough to support a reasonable inference that Flynn had
“supervisory authority [and] institutional influence in recommending
and thus influencing” the Conduct Board’s decision. Columbia Univ.,

38
831 F.3d at 59. Flynn was also responsible for deciding the
appropriate sanction. See App’x 230 (“The sanctions … are imposed
by the Office of Student Conduct.”). And it is plausible that the bias
that affected the interim suspension would also have motivated the
eventual expulsion. See Amherst Coll., 238 F. Supp. 3d at 223 (holding
that a claim of sex discrimination was plausible based on the
allegation that “the College intended [the plaintiff’s] punishment to
appease campus activists who sought the expulsion of a male
student”).

C

Like the irregularities alleged in Columbia University and
Menaker, SJU allegedly failed to follow its own procedures, failed to
investigate Roe’s complaint, and reached decisions that were
perplexing and based on incorrect or unproved versions of the facts.
Other circuits have held that similar allegations are enough to state a
claim under Title IX. See, e.g., Schwake, 967 F.3d at 951 (holding a claim
to be plausible based on allegations that the university failed to
consider the accused’s version of the alleged assault); Purdue Univ.,
928 F.3d at 669 (holding a claim to be plausible based on allegations
that the university’s Title IX investigator credited the story of the
female accuser over the male accused when the investigator had
never spoken with the accuser); Baum, 903 F.3d at 586 (holding a claim
to be plausible based on allegations that the appeals board exclusively
credited female testimony and rejected male testimony).

Like the allegations in those cases, Roe’s allegations provide
“circumstantial evidence of bias in [his] specific proceeding” that,
when combined with the allegations of sex-based pressure, “gives rise
to a plausible claim.” Baum, 903 F.3d at 586. “There is nothing
implausible or unreasonable about the Complaint’s suggested

39
inference that the panel adopted a biased stance in favor of the
accusing female and against the defending male … in order to avoid
further fanning the criticisms that [SJU] turned a blind eye to such
assaults.” Columbia Univ., 831 F.3d at 58.

IV

Independent of his claims for improper discipline, Roe alleges
that SJU was deliberately indifferent to a hostile educational
environment because it failed to respond to the anonymous tweet
accusing him of rape. I agree that the district court was correct to
dismiss this claim because the alleged harassment was not “so severe,
pervasive, and objectively offensive that it can be said to deprive
[him] of access to the educational opportunities or benefits provided
by the school.” Ante at 40 (quoting Davis, 526 U.S. at 650).

I note that the district court nevertheless erred when it
dismissed Roe’s claim on the ground that the tweet “was based on
[Roe’s] alleged past conduct, not his gender.” 2021 WL 1224895, at *20.
In Menaker, we held that a student’s false accusation of sexual
misconduct was plausibly “motivated, at least in part, by Menaker’s
sex.” 935 F.3d at 38. The student, we explained,

did not accuse Menaker of just any misconduct; she
accused him of sexual misconduct. That choice is
significant, and it suggests that Menaker’s sex played a
part in her allegations. A rational finder of fact could
therefore infer that such an accusation was based, at least
in part, on Menaker’s sex.

Id.; see also Cox v. Onondaga Cnty. Sheriff’s Dep’t, 760 F.3d 139, 149 (2d
Cir. 2014) (observing that “false statements … intended … to establish
a claim of racial harassment … could be viewed by a reasonable
observer as themselves racial harassment”). So too here. Assuming,
40
as we must on a motion to dismiss, that the tweet falsely accused Roe
of rape, then the tweet would represent sex-based harassment.

Still, for the other reason the court provides, I would affirm the
district court’s dismissal of Roe’s claim alleging deliberate
indifference.

* * *

Universities subject to Title IX must ensure that students are not
denied educational opportunities because of sex-based
discrimination. In the context of discipline arising from alleged sexual
misconduct, schools must investigate allegations of misconduct, lest
the victims be deprived of educational opportunities, and must afford
procedural protections to the accused, lest those students be similarly
deprived. See 34 C.F.R. § 106.45(a). Universities are not free to “favor
the accusing female over the accused male, so as to protect … the
University from accusations that [the University] had failed to protect
female students from sexual assault.” Columbia Univ., 831 F.3d at 57.
A university may not “expel first and ask questions later.” Foster v.
Bd. of Regents of Univ. of Mich., 982 F.3d 960, 969 (6th Cir. 2020).

Yet today the court tells SJU that it need not provide defensible
reasons for suspending men accused of sexual assault. It authorizes a
university to provide an accused student with flawed procedures and
then to rely on “a faulty explanation for its erroneous conclusion” that
he should be expelled. Ante at 21. And it tells SJU that a litany of
procedural irregularities will be excused absent specific evidence that
refutes whatever alternative explanation for the irregularities a court
might imagine. That approach is inconsistent with our precedents and
with Title IX.

None of these considerations necessarily mean that Roe should
ultimately prevail in this case. See Purdue Univ., 928 F.3d at 670 (“To
41
be sure, John may face problems of proof, and the factfinder might
not buy the inferences that he’s selling.”). Perhaps SJU followed a
more regular process and had a more reasoned justification for its
decisions than the complaint describes. But “[w]e are limited here to
the allegations in the complaint and the evidence in the attachments
to it. … [A]lthough the facts at trial or summary judgment might show
otherwise, we cannot manufacture such facts out of thin air” but
should require SJU to establish those facts. Sabir, 52 F.4th at 64-65
(internal quotation marks and alteration omitted). Because Roe’s
“allegations raise a plausible inference that he was denied an
educational benefit on the basis of his sex … his claim should have
made it past the pleading stage,” and we ought to reverse the district
court’s “premature dismissal of it.” Purdue Univ., 928 F.3d at 670.
Accordingly, I dissent.

42