Foreseeability, Special Relationships, and Campus Liability: Sixth Circuit Narrows Negligence Exposure and Sets High Bar for IIED and ELCRA Claims in Chen v. Hillsdale College

Introduction

In Grace Chen v. Hillsdale College, No. 24-1788 (6th Cir. Aug. 28, 2025), the U.S. Court of Appeals for the Sixth Circuit addressed whether a private college can be held liable under Michigan law for student-on-student sexual assaults and for its handling of those incidents. Two students alleged assaults by classmates, criticized Hillsdale’s investigatory response, and sued for negligence, intentional infliction of emotional distress (IIED), and sex discrimination under Michigan’s Elliott-Larsen Civil Rights Act (ELCRA). The district court dismissed the case, and a split Sixth Circuit panel affirmed. Chief Judge Sutton wrote for the court; Judge White concurred in part and dissented in part, concluding the IIED claim should proceed.

The decision clarifies several important points of Michigan law as applied in federal court: the absence of a generalized duty on colleges to prevent criminal acts by third parties, the stringent foreseeability requirements to trigger any duty under a “special relationship,” the demanding “extreme and outrageous” threshold for IIED in the campus investigation context, and the evidentiary showing required for disparate treatment and disparate impact claims under ELCRA. The panel also offers guidance on certification requests to state supreme courts after adverse federal rulings.

Summary of the Judgment

Negligence: Affirmed dismissal. Under Michigan law, colleges owe no general duty to protect students from criminal acts of other students. Any “special relationship” duty arises only in narrow, controlled contexts and extends solely to imminent, foreseeable harms to readily identifiable victims. Plaintiffs did not allege Hillsdale had specific notice of an imminent risk to them.
Intentional Infliction of Emotional Distress: Affirmed dismissal (majority). While plaintiffs alleged insensitive, even hurtful comments and a deficient process, the conduct did not meet Michigan’s “extreme and outrageous” standard when viewed in the aggregate, especially given the college’s contemporaneous efforts to investigate and impose some discipline. Judge White dissented on this claim, concluding the alleged comments, threats, and stonewalling by officials plausibly met the IIED threshold when combined with the power imbalance and victims’ vulnerability.
ELCRA Sex Discrimination: Affirmed dismissal. Plaintiffs failed to allege that similarly situated male and female complainants received differential treatment (disparate treatment). For disparate impact, generalized societal statistics about sexual assault are insufficient absent school-specific facts or data showing that Hillsdale’s policies caused a sex-based disparity.
Certification: The court declined to certify state-law questions post-judgment, noting plaintiffs chose a federal forum and did not seek certification below.

Analysis

Precedents Cited and Their Influence

1) Negligence duty and foreseeability

Krass v. Tri-County Security, Inc. (Mich. Ct. App. 1999): Restates the default Michigan rule—no duty to protect another from the criminal acts of third parties.
Prosser & Keeton on Torts § 33 (5th ed. 1984): Under ordinary circumstances, an actor may assume others will obey criminal law; the burden of continuous precautions typically outweighs the risk. The majority leans on this principle to resist converting colleges into insurers of student safety.
In re Certified Question from Fourteenth District Court of Appeals of Texas (Mich. 2007): Michigan’s duty analysis weighs relationship of the parties, foreseeability of harm, burden on the defendant, and risk nature. The court deploys this multi-factor approach in rejecting a duty here.
Williams v. Cunningham Drug Stores, Inc. (Mich. 1988): Recognizes exceptions for “special relationships” (e.g., common carriers, innkeepers), but only in controlled contexts. The Sixth Circuit applies this exception narrowly to higher education settings.
MacDonald v. PKT, Inc. (Mich. 2001) and Graves v. Warner Bros. (Mich. Ct. App. 2002): Foreseeability requires an “imminent” risk to a “readily identifiable” victim; general awareness of crime trends is not enough. This standard is dispositive: plaintiffs alleged no specific warning signs or histories tied to the assailants.
Tame v. A L Damman Co. (Mich. Ct. App. 1989): A proprietor need not “turn its establishment into a fortress.” The court paraphrases this to caution against limitless campus security obligations.
In re Doe (Mich. Ct. App. 2006): No school duty where the incident occurred outside school hours and activities; highlights “control” over the setting as a prerequisite to a special relationship. The court analogizes to college dorm and off-campus apartment settings.
Out-of-state comparators: Davidson v. UNC (N.C. Ct. App. 2001) (special relationship during cheerleading practice); Avila v. Citrus Community College (Cal. 2006) (special duties to athletes during sports); Regents of the Univ. of Cal. v. Superior Court (Cal. 2018) (duty where university knew of assailant’s violent history and the attack occurred in a classroom); Barlow v. State (Wash. 2024) (no duty for unforeseeable off-campus harm). These comparisons underscore that when universities are in a supervisory position during school activities and have specific risk knowledge, duties may arise—but not here.
O’Neal v. MCC Mecosta, LLC (Mich. Ct. App. 2023) and Dawe v. Bar-Levav & Associates, P.C. (Mich. Ct. App. 2010): Michigan courts demand notice of a “specific situation” of peril; prior “inappropriate” conduct is not equivalent to notice of imminent assault (O’Neal), whereas explicit homicidal fantasies paired with a gun at the office suffice (Dawe). These contrasts emphasize why plaintiffs’ generalized policy criticisms and anonymous posts did not allege foreseeability.
National higher-education cases: Freeman v. Busch (8th Cir. 2003), Guest v. Hansen (2d Cir. 2010), McCauley v. Univ. of the V.I. (3d Cir. 2010): Modern higher education is not in loco parentis; colleges do not bear parental-like safety duties over adult students. The Sixth Circuit leans into this consensus.
Title IX analogs rejected: Ware v. Univ. of Vermont (D. Vt. 2024), Karasek v. Regents of Univ. of Cal. (9th Cir. 2020), Simpson v. Univ. of Colo. Boulder (10th Cir. 2007). The court distinguishes Title IX’s statutory framework and funding-based obligations from Michigan negligence law; even on their own terms, those cases turned on either dismissals or unique foreseeability/supervision facts.

2) IIED threshold

Teadt v. Lutheran Church Missouri Synod (Mich. Ct. App. 1999); Bernhardt v. Ingham Regional Medical Center (Mich. Ct. App. 2002): Articulate Michigan’s well-known IIED elements and the “so outrageous in character” standard.
Meek v. Michigan Bell Telephone Co. (Mich. Ct. App. 1991); Duran v. Detroit News (Mich. Ct. App. 1993); Rosenberg v. Rosenberg Bros. Special Account (Mich. Ct. App. 1984); Ledsinger v. Burmeister (Mich. Ct. App. 1982): Illustrate the high bar for “extreme and outrageous” conduct and contexts that can elevate otherwise offensive conduct (e.g., abuse of power in a public accommodation).
Conway v. Detroit Public Schools Community District (Mich. Ct. App. 2023): Supports considering “aggregated circumstances,” which the majority does, concluding the whole does not cross the line.
Dissent’s authorities: Restatement (Second) of Torts § 46 cmts. e, f, i (abuse of authority and vulnerability can make conduct extreme; recklessness suffices), and out-of-jurisdiction cases (Brandon v. County of Richardson (Neb. 2001); Drejza v. Vaccaro (D.C. 1994); Snyder v. Smith (S.D. Ind. 2014)). The dissent analogizes campus officials to authority figures whose belittling, threatening conduct toward vulnerable victims can be “extreme and outrageous.”

3) ELCRA discrimination

Lytle v. Malady (Mich. 1998): ELCRA supports disparate treatment and disparate impact theories.
Betty v. Brooks & Perkins (Mich. 1994): The “essence” is whether similarly situated persons are treated differently because of sex.
Garvin v. Detroit Board of Education (Mich. Ct. App. 2013): Requires factual allegations indicating differential treatment would have occurred if the victim’s sex were different.
Alspaugh v. Commission on Law Enforcement Standards (Mich. 2001) and Texas Department of Housing & Community Affairs v. Inclusive Communities Project, Inc. (U.S. 2015): Disparate impact demands proof that a facially neutral policy causes a class-based disparity; generalized societal disparities “without more” are insufficient.
Burkhardt v. Flint Community Schools (Mich. Ct. App. 2020): Anecdotes alone do not “adequately explain how” a policy creates a disproportionate burden.

4) Certification after adverse federal judgment

City of Columbus v. Hotels.com, L.P. (6th Cir. 2012); State Auto Property & Casualty Insurance v. Hargis (6th Cir. 2015): Certification is disfavored when the plaintiff chose federal court and seeks certification only after an adverse judgment and did not ask below.

Legal Reasoning

Negligence: No duty and no foreseeability

The core of the negligence analysis is Michigan’s general rule: no duty to protect against third-party criminal acts. A narrow exception exists for “special relationships” where a defendant exercises control or holds a custodial/supervisory role (innkeepers, common carriers, certain school contexts), but even then the duty extends only to imminent, foreseeable risks to identifiable victims. The court stressed two gaps:

Relationship/control: The alleged assaults occurred in a dorm room and an off-campus apartment, not during a school-controlled activity. Absent a school-sponsored or supervised context (e.g., practice, class, travel under coach supervision), Hillsdale did not occupy a “unique position” of control to prevent the harm.
Foreseeability: Plaintiffs alleged no facts showing Hillsdale knew or should have known of specific, imminent threats by the assailants to these plaintiffs (e.g., prior complaints about these assailants, explicit threats, observable pre-assault conduct suggesting imminent harm). Anonymous postings or a student newspaper critique of procedures do not equate to knowledge of a specific, imminent risk.

The court emphasized the danger of converting foreseeability into omniscience and colleges into insurers. It also distinguished Title IX-based arguments because Hillsdale does not receive federal funds and because Title IX’s statutory obligations are not coextensive with Michigan negligence doctrine.

IIED: Aggregated circumstances do not meet the “extreme and outrageous” bar

Applying Michigan’s stringent standard, the majority accepted that plaintiffs alleged insensitive and even threatening remarks, delays, and limited enforcement, but it weighed these alongside:

Prompt connections to outside investigators;
Statements crediting belief in at least one complainant’s account;
Some disciplinary action (e.g., probation, community service, team suspension).

Considering the whole and the obligation to afford some fairness to accused students, the court concluded the alleged conduct, while objectionable, was not “atrocious and utterly intolerable in a civilized community.”

The dissent would allow the IIED claim to proceed, stressing:

The alleged abuse of authority (lawyers and administrators) over vulnerable victims;
Statements that downplayed or denied the assault, mischaracterized consent, and threatened “consequences” for continued complaints;
Restatement (Second) of Torts § 46 comments recognizing that authority and victim vulnerability can transform conduct into “extreme and outrageous;”
Persuasive analogies to cases condemning law-enforcement mistreatment of sexual-assault victims.

The split reveals a meaningful, fact-intensive line: when do insensitive or discouraging institutional responses become “extreme” enough for IIED under Michigan law?

ELCRA: No plausible disparate treatment or disparate impact

For disparate treatment, plaintiffs needed to allege that similarly situated male and female complainants were treated differently. The complaint lacked any comparator allegations (e.g., how Hillsdale treated a male victim reporting assault), focusing instead on male perpetrators. The court found a “dearth” of facts suggesting that outcomes would have differed if a complainant were male.

For disparate impact, plaintiffs relied on general nationwide statistics about sexual assault and argued that Hillsdale’s policy chilled reporting and favored men. The court held that ELCRA requires school-specific facts or data tying a facially neutral policy to a disproportionate burden on women at that school. The complaint did not identify metrics, internal statistics, or policy mechanisms that caused a disparate impact at Hillsdale. The court also rejected attempts to infer discriminatory impact from Hillsdale’s religious mission statements, cautioning against conclusory allegations and noting record evidence (a newspaper article) describing Hillsdale as “atypically safe” with commitments to respond “swiftly” and “with compassion and respect.”

Certification

After losing in federal court, plaintiffs asked the Sixth Circuit to certify state-law questions to the Michigan Supreme Court. The panel declined, highlighting three points: plaintiffs chose a federal forum, did not request certification below, and offered no good explanation for seeking certification only after an adverse judgment. This reinforces the Sixth Circuit’s general reluctance to permit post-judgment certification as a second bite at the apple.

Impact and Implications

Negligence exposure for Michigan colleges: The decision significantly narrows negligence exposure for campuses regarding peer-on-peer crimes absent clear, specific notice of an imminent risk to an identifiable student and a setting where the school exerts meaningful control (e.g., supervised events, courses, official travel, team activities).
Pleading standards after Chen: Plaintiffs must plead concrete facts indicating the institution’s knowledge of a specific assailant’s dangerousness to a specific victim (prior complaints, warnings, threatening conduct, observable indicators of imminent harm) to survive a motion to dismiss on duty/foreseeability.
IIED claims in campus investigations: The bar remains high. Insensitivity, delays, or imperfect enforcement, without more, likely fail. However, the dissent signals that allegations of bullying, threats, and victim-blaming by officials wielding institutional authority, especially against vulnerable complainants, may suffice in some cases. Expect litigants to sharpen factual allegations around authority abuse, victim vulnerability, and persistent, deliberate mistreatment.
ELCRA disparate impact suits: General social science or national statistics about sexual assault are not enough. Plaintiffs will need school-level data and a causal link between the policy and a sex-based disparity in outcomes (e.g., reporting rates, discipline rates, no-contact orders, time-to-resolution). Absent such evidence, ELCRA impact claims will likely be dismissed at the pleading stage.
Title IX versus state law: Institutions that do not accept federal funding avoid Title IX’s statutory obligations but face state-law claims. Chen underscores that Title IX case law does not map neatly onto state negligence and IIED claims.
Faith-based institutions and mission statements: Plaintiffs cannot rely on a school’s religious views on sexuality as a stand-in for discriminatory impact. Courts will look for concrete policy mechanisms and evidence of disparate effects.
Certification strategy: Litigants should identify unsettled state-law questions early and seek certification in the district court. Post-judgment certification requests are disfavored.

Complex Concepts Simplified

Duty to protect against third-party crime: Generally, a person (or institution) is not legally obligated to prevent criminal acts by others. Exceptions exist only where the law recognizes a “special relationship” (e.g., common carrier–passenger) that justifies imposing a protective duty.
Special/dependent relationship in the campus context: This usually requires the school to be in a position of control over the setting and participants (e.g., during official practices, classes, or supervised travel). Merely being a student does not create a custodial relationship with the college for all places and times.
Foreseeability: To trigger a duty even within a special relationship, the risk must be specific, imminent, and directed at an identifiable victim (not generalized concerns or campus climate critiques).
In loco parentis: Historically, colleges acted like parents to students. Modern courts reject that model for adult higher education; colleges are not insurers of student safety.
IIED “extreme and outrageous” standard: Conduct must be truly egregious—beyond all possible bounds of decency. Authority figures who exploit their power over particularly vulnerable individuals can meet this standard, but courts demand concrete, severe facts.
Disparate treatment vs. disparate impact (ELCRA):
- Disparate treatment requires plausible facts that similarly situated individuals of different sexes were treated differently because of sex.
- Disparate impact challenges facially neutral policies that disproportionately harm one sex; plaintiffs must show a causal link between the policy and the disparity with school-specific facts or data.
Certification to state supreme courts: Federal courts may ask a state supreme court to resolve novel state-law questions. But parties should seek certification early; it is generally disfavored when sought only after losing in federal court.

Conclusion

Chen v. Hillsdale College is a significant, published decision that tightens the boundaries of institutional liability for student-on-student assaults under Michigan law. The court squarely holds that, absent control over the setting and specific, imminent foreseeability of harm to an identifiable student, a college owes no duty to protect against peer criminal acts. It also reaffirms the demanding threshold for IIED in the university-investigation context, while a partial dissent highlights circumstances where abuse of institutional authority and victim vulnerability could state a claim. On ELCRA, the opinion clarifies that plaintiffs must allege school-specific comparators or data linking a policy to a sex-based disparity; national statistics and generalized critiques will not do.

Practically, Chen directs plaintiffs to plead specific pre-assault notice and control facts and to develop robust, institution-level evidence for ELCRA claims. For colleges—especially those outside Title IX’s ambit—the case provides a framework for evaluating risk and for structuring investigation protocols that are both fair to the accused and responsive to complainants. Finally, the panel’s certification discussion is a reminder to identify unsettled state-law issues early and to make timely certification requests.