The REGENTS OF the UNIVERSITY OF CALIFORNIA et al., Petitioners, v. The SUPERIOR COURT of Los Angeles County, Respondent; Katherine Rosen, Real Party in Interest.
After he enrolled in the University of California at Los Angeles (UCLA), Damon Thompson experienced auditory hallucinations. He believed other students in the classroom and dormitory were criticizing him. School administrators eventually learned of Thompson's delusions and attempted to provide mental health treatment. However, one morning Thompson stabbed fellow student Katherine Rosen during a chemistry lab. Rosen sued the university and several of its employees for negligence, arguing they failed to protect her from Thompson's foreseeable violent conduct.
This case involves whether, and under what circumstances, a college or university 1 owes a duty of care to protect students like Rosen from harm. Considering the unique features of the collegiate environment, we hold that universities have a special relationship with their students and a duty to protect them from foreseeable violence during curricular activities. Because the Court of Appeal reached a different conclusion, we reverse its decision and remand for further proceedings.
A. Thompson's Behavior Preceding the Assault
Damon Thompson transferred to UCLA in the fall of 2008. He soon began experiencing problems with other students in both classroom and residence hall settings.
At the end of fall quarter, Thompson emailed his history professor that he was “angered” by “offensive” remarks from other students during the final examination and “outrage[d]” because their comments had affected his performance. Thompson also complained he had heard the professor calling him “ ‘troubled’ and ‘crazy’ among other things.” When the professor forwarded Thompson's messages to his department chair, he was advised to calm Thompson and encourage him to visit the school's counseling services if he appeared “genuinely paranoid or a potential threat.”
Thompson next complained about mistreatment by fellow dormitory residents. In a three-page letter to the Dean of Students, Thompson alleged a female resident had repeatedly made “unwelcomed verbal sexual advances” toward him, and others had spread rumors and “accusations of a sexual nature about [him] ․ throughout the entire student body.” He claimed the residents frequently disrupted his sleep, called him “ ‘stupid,’ ” and eavesdropped on his phone calls. Not only had he been “made the ‘target’ ” of the residents' “teasing,” but he also “receive[d] an immense amount of unwanted attention” around campus. Thompson warned that if the university failed to discipline the responsible parties, the matter would likely “escalate into a more serious situation,” and he would “end up acting in a manner that will incur undesirable consequences.” A week later, the school moved Thompson to a new dormitory.
In late January 2009, Thompson complained to three professors and teaching assistant Jenny Hernandez that students had been trying to distract him with offensive comments. Hernandez told her supervising professor she had never observed this behavior but Thompson himself acted oddly, frequently talking to himself. She believed he was displaying signs of schizophrenia and should be referred to the university's Counseling and Psychological Services (CAPS). Hernandez and the professor met with Thompson and urged him to use these services, but Thompson denied “ ‘hearing things' ” or “ ‘making this up.’ ” Another professor forwarded Thompson's complaints to Assistant Dean of Students Cary Porter, who contacted the university's “Consultation and Response Team” (Response Team). The Response Team advises campus members who have concerns about the well-being of particular students. Dean Porter also met with Thompson and encouraged him to seek medical help at CAPS.
Thompson's dormitory problems escalated in February. He told resident director Janelle Rahyns there were “voices coming through the walls calling him an idiot.” He heard a clicking noise above his room that sounded like a gun, and he believed the other residents were planning to shoot him. Thompson told Rahyns he had telephoned his father and was advised to “hurt the other residents.” While admitting he had “thought about it,” Thompson said he decided not to hurt anyone. Campus police arrived and searched the premises but found no weapon. They concluded Thompson needed a psychiatric evaluation and escorted him to the emergency room for that purpose. During the examination, Thompson reported a history of depression and complained of auditory hallucinations and paranoid thinking. For several months, he had heard people talking about him and insulting him, even when “ ‘there's no one there.’ ” He denied suicidal or homicidal thinking. The examiner diagnosed Thompson with possible schizophrenia and major depressive disorder. Thompson agreed to take a low-dose antipsychotic medication and begin outpatient treatment at CAPS. Dean Porter and the Response Team were informed about the incident and Thompson's mental evaluation. The Response Team began discussing Thompson at its weekly meetings.
In March 2009, Thompson began sessions with CAPS psychologist Nicole Green. Although he denied wanting to hurt himself or others, he continued to report auditory hallucinations and paranoid thoughts. He had thrown away the prescribed antipsychotic medication. Green diagnosed schizophrenia and urged Thompson to see a CAPS psychiatrist. Thompson refused to consider medication until he could determine whether the voices were real. He expressed frustration that nobody believed him and said he would try to record the voices. Around this time, Rahyns notified CAPS that Thompson was “still having trouble” in the dormitory. The Response Team decided to move him to a single room and explore possibilities for transitioning him into different housing.
Later in March, Thompson told Green he was still hearing voices and being harassed by other students. He was now amenable to psychiatric evaluation. Later that day, at a session with CAPS psychiatrist Charles McDaniel, Thompson admitted thinking about harming others, although he had no identified victim or plan. He heard numerous distinct voices in his dormitory and classrooms. He wanted to harm the people insulting him but could not attribute the voices to specific individuals. McDaniel strongly urged Thompson to submit to voluntary hospitalization. He refused but agreed to take medication. While CAPS staff agreed Thompson did not meet the criteria for an involuntary hold, McDaniel recommended involuntary hospitalization if his thoughts of harming others worsened. Thompson attended additional CAPS sessions in April and continued to report auditory hallucinations. Although angered by this perceived harassment, Thompson said he did not intend to harm his tormentors. He withdrew from treatment in late April.
On June 3, 2009, campus police responded to an incident at Thompson's dormitory. A resident reported that Thompson had knocked on his door, accused him of making too much noise, and pushed him. When the resident denied making noise, Thompson pushed him again, saying this was his “ ‘last warning.’ ” As a result of the incident, Thompson was expelled from university housing and ordered to return to CAPS at the beginning of fall quarter. After he moved to an apartment, Thompson twice called the police to complain neighbors were yelling at him through the floor.
Meanwhile, Thompson continued to experience auditory hallucinations in the classroom. During the summer, he complained to two faculty members about insults and harassment in his chemistry laboratory. After fall quarter started, Thompson emailed professor Alfred Bacher that the disruptive behavior of other students was interfering with his experiments. The next day, September 30, Thompson told CAPS psychologist Tanya Brown he still “occasionally” heard “voices of other students having ‘malice’ toward him and making critical and racist comments.” Nevertheless, he denied an intent to harm anyone, including those criticizing him. Brown noted that Thompson displayed a guarded attitude, slowed speech, delusional thought processes, and impaired insight. CAPS psychiatrist Charles McDaniel met with Thompson the same day and made similar observations. Due to Thompson's behavior, McDaniel was unsure whether he was reporting his symptoms accurately. Thompson agreed to start treatment at the university's behavioral health clinic.
B. The Assault
On October 6th, teaching assistant Adam Goetz emailed Professor Bacher describing “another incident” with Thompson in that day's chemistry lab. Shortly after the professor left the room, Thompson accused another student of calling him stupid. He insisted on learning the student's name. After Goetz gave him the name, Thompson “calmed down” and “seemed fine.” But Goetz remained worried that Thompson's behavior was becoming a weekly “routine.” Goetz later testified that Thompson frequently identified Katherine Rosen, who worked “right next to” Thompson in the lab, as one of the students calling him stupid.
The following day, another teaching assistant told Professor Bacher that Thompson had come into his chemistry lab from a different section and accused students of verbally harassing him. Although Thompson did not know the students' names, he did identify a specific student, other than Rosen, as one of his tormentors. The teaching assistant saw no harassment and was skeptical of Thompson's claims.
Bacher forwarded Goetz's email to Dean Porter on the morning of October 7th, seeking advice on how to handle the situation. Porter contacted Karen Minero of the Response Team, who expressed concern that Thompson had identified a specific student. Minero forwarded Porter's email to other Response Team members and CAPS personnel. The CAPS director contacted Green, suggesting Thompson “may need urgent outreach,” and members of the Response Team tried to schedule a meeting to discuss Thompson. Thompson did not appear for a scheduled session with Green that afternoon. The next morning, Porter and Minero discussed Thompson and decided to investigate whether he was having similar difficulties in other classes.
Around noon on October 8th, Thompson was doing classwork in Professor Bacher's chemistry laboratory. Suddenly, without warning or provocation, he stabbed fellow student Katherine Rosen in the chest and neck with a kitchen knife. Rosen had been kneeling down, placing items in her lab drawer, when Thompson attacked her from behind. She was taken to the hospital with life-threatening injuries but ultimately survived. When campus police arrived, Thompson admitted he had stabbed someone and explained that the other students had been teasing him. Thompson ultimately pleaded not guilty by reason of insanity to a charge of attempted murder. (Pen. Code, §§ 187, subd. (a), 664, 1026.) He was admitted to Patton State Hospital and diagnosed with paranoid schizophrenia.
C. Procedural History
Rosen sued Thompson, the Regents of the University of California, and several UCLA employees, including Alfred Bacher, Cary Porter, Robert Naples, and CAPS psychologist Nicole Green. The complaint alleged a single cause of action against the UCLA defendants 2 for negligence. Rosen alleged UCLA had a special relationship with her as an enrolled student, which entailed a duty “to take reasonable protective measures to ensure her safety against violent attacks and otherwise protect her from reasonable foreseeable criminal conduct, to warn her as to such reasonable foreseeable criminal conduct on its campus and in its buildings, and/or to control the reasonably foreseeable wrongful acts of third parties/other students.” She alleged UCLA breached this duty because, although aware of Thompson's “dangerous propensities,” it failed to warn or protect her or to control Thompson's foreseeably violent conduct.
UCLA moved for summary judgment on three alternative grounds: (1) colleges have no duty to protect their adult students from criminal acts; (2) if a duty does exist, UCLA did not breach it in this case; and (3) UCLA and Green were immune from liability under certain Government Code provisions. In opposing the motion, Rosen argued UCLA owed her a duty of care because colleges have a special relationship with students in the classroom, based on their supervisory duties and the students' status as business invitees. Rosen also claimed UCLA assumed a duty of care by undertaking to provide campus-wide security.
The trial court denied the motion. The court concluded a duty could exist under each of the grounds Rosen identified, triable issues of fact remained as to breach of duty, and the immunity statutes did not apply.
UCLA challenged this order in a petition for writ of mandate. A divided panel of the Court of Appeal granted the petition. The majority held that UCLA owed no duty to protect Rosen based on her status as a student or business invitee, or based on the negligent undertaking doctrine. It also rejected Rosen's new theories of duty based on implied-in-fact contract and labor laws regarding violence in the workplace. The dissenting justice would have held that colleges have a special relationship with their enrolled students, “at least when the student is in a classroom under the direct supervision of an instructor,” and have a corresponding duty to protect against foreseeable threats of violence in the classroom. We granted review.
A. Standard of Review
“On review of an order granting or denying summary judgment, we examine the facts presented to the trial court and determine their effect as a matter of law.” (Parsons v. Crown Disposal Co. (1997) 15 Cal.4th 456, 464 (Parsons).) We review the entire record, “considering all the evidence set forth in the moving and opposition papers except that to which objections have been made and sustained.” (Guz v. Bechtel National, Inc. (2000) 24 Cal.4th 317, 334.) Evidence presented in opposition to summary judgment is liberally construed, with any doubts about the evidence resolved in favor of the party opposing the motion. (Yanowitz v. L'Oreal USA, Inc. (2005) 36 Cal.4th 1028, 1037.)
Summary judgment is appropriate only “where no triable issue of material fact exists and the moving party is entitled to judgment as a matter of law.” (Merrill v. Navegar, Inc. (2001) 26 Cal.4th 465, 476.) A defendant seeking summary judgment must show that the plaintiff cannot establish at least one element of the cause of action. (Aguilar v. Atlantic Richfield Co. (2001) 25 Cal.4th 826, 853.) Rosen's negligence suit required her to prove duty, breach, causation, and damages. (Conroy v. Regents of University of California (2009) 45 Cal.4th 1244, 1250.) The Court of Appeal determined summary judgment should have been granted because Rosen could not establish duty. “Duty, being a question of law, is particularly amenable to resolution by summary judgment.” (Parsons, supra, 15 Cal.4th at p. 465.) Contrary to the Court of Appeal, however, we conclude universities do have a legal duty, under certain circumstances, to protect or warn 3 their students from foreseeable violence in the classroom or during curricular activities. The trial court properly denied summary judgment on this ground.
B. A College's Duty to Protect Students from Foreseeable Harm
Because UCLA is a public entity, its exposure to tort liability is nominally defined by statute. (Gov. Code, § 815, subd. (a); C.A. v. William S. Hart Union High School Dist. (2012) 53 Cal.4th 861, 868 (William S. Hart).) However, the Tort Claims Act provides that public employees are liable for their acts and omissions “to the same extent as a private person” (Gov. Code, § 820, subd. (a)), and public entity employers are vicariously liable for employees' negligent acts within the scope of their employment to the same extent as private employers (Gov. Code, § 815.2, subd. (a); William S. Hart, at p. 868). Because it is undisputed that all university employees here were acting within the scope of their employment, UCLA's potential liability therefore “turns on ordinary and general principles of tort law.” (Lugtu v. California Highway Patrol (2001) 26 Cal.4th 703, 716.)
In general, each person has a duty to act with reasonable care under the circumstances. (Cabral v. Ralphs Grocery Co. (2011) 51 Cal.4th 764, 771 (Cabral); see Civ. Code, § 1714, subd. (a).) However, “one owes no duty to control the conduct of another, nor to warn those endangered by such conduct.” (Davidson v. City of Westminster (1982) 32 Cal.3d 197, 203 (Davidson).) “A person who has not created a peril is not liable in tort merely for failure to take affirmative action to assist or protect another unless there is some relationship between them which gives rise to a duty to act.” (Williams v. State of California (1983) 34 Cal.3d 18, 23; see Cabral, at p. 771.)
A duty to control, warn, or protect may be based on the defendant's relationship with “either the person whose conduct needs to be controlled or [with] ․ the foreseeable victim of that conduct.” (Tarasoff, supra, 17 Cal.3d at p. 435; see Davidson, supra, 32 Cal.3d at p. 203.) Specifically, a duty to control may arise if the defendant has a special relationship with the foreseeably dangerous person that entails an ability to control that person's conduct. (Rest.3d Torts, Liability for Physical and Emotional Harm, § 41.) The parent-child relationship is an example of a special relationship giving rise to a duty to control. (See id., § 41, subd. (b)(1); Smith v. Freund (2011) 192 Cal.App.4th 466, 473.) Similarly, a duty to warn or protect may be found if the defendant has a special relationship with the potential victim that gives the victim a right to expect protection. (See Zelig v. County of Los Angeles (2002) 27 Cal.4th 1112, 1129; Davidson, at p. 203.) The relationships between common carriers and their passengers, or innkeepers and their guests, are classic examples of this type of special relationship. (See Rest.3d Torts, Liability for Physical and Emotional Harm, § 40, subd. (b)(1)-(2).)
Rosen's complaint alleges UCLA had separate duties to protect her and to “control the reasonably foreseeable wrongful acts of third parties/other students.” Here, we have focused on the university's duty to protect students from foreseeable violence. Having concluded UCLA had a duty to protect Rosen under the circumstances alleged, we need not decide whether the school had a separate duty to control Thompson's behavior to prevent the harm.
1. College-Student Special Relationship Supports a Limited Duty
Whether UCLA was negligent in failing to prevent Thompson's attack depends first on whether a university has a special relationship with its students that supports a duty to warn or protect them from foreseeable harm. The determination whether a particular relationship supports a duty of care rests on policy and is a question of law. (Rest.3d Torts, Liability for Physical and Emotional Harm, § 40, coms. e & h, pp. 41-42.)
a. Features of a Special Relationship
The Restatement Third of Torts identifies several special relationships that may support a duty to protect against foreseeable risks. In addition to the common carrier and innkeeper relationships previously mentioned, the list includes a business or landowner with invited guests, a landlord with tenants, a guard with those in custody, an employer with its employees, and “a school with its students.” (Rest.3d Torts, Liability for Physical and Emotional Harm, § 40, subd. (b).) The Restatement does not exclude colleges from the school-student special relationship. However, the drafters observe that reasonable care varies in different school environments, with substantially different supervision being appropriate in elementary schools as opposed to colleges. (Id., § 40, com. l, p. 45.) State courts have reached different conclusions about whether colleges owe a special relationship-based duty to their students. (Id., § 40, com. l, reporter's notes, p. 57.) We have not previously addressed the question.
Relationships that have been recognized as “special” share a few common features. Generally, the relationship has an aspect of dependency in which one party relies to some degree on the other for protection. (See Baldwin v. Zoradi (1981) 123 Cal.App.3d 275, 283 (Baldwin); Mann v. State of California (1977) 70 Cal.App.3d 773, 779-780.) The Restatement authors observed over 50 years ago that the law has been “working slowly toward a recognition of the duty to aid or protect in any relation of dependence or of mutual dependence.” (Rest.2d Torts, § 314A, com. b, p. 119.)
The corollary of dependence in a special relationship is control. Whereas one party is dependent, the other has superior control over the means of protection. “[A] typical setting for the recognition of a special relationship is where ‘the plaintiff is particularly vulnerable and dependent upon the defendant who, correspondingly, has some control over the plaintiff's welfare.’ [Citations.]” (Giraldo v. Department of Corrections & Rehabilitation (2008) 168 Cal.App.4th 231, 245-246.) One court observed that “the epitome” of such a special relationship exists between a jailer and prisoner. (Id. at pp. 250-251.) Common carriers and their passengers present another quintessential example. In Lopez v. Southern Cal. Rapid Transit Dist. (1985) 40 Cal.3d 780, 789, we held this special relationship gives common carriers a duty to protect passengers from onboard violence, noting that passengers are sealed together in a moving vehicle, with the means of entry and exit under the exclusive control of the driver. “Thus, passengers have no control over who is admitted on the bus and, if trouble arises, are wholly dependent upon the bus driver to summon help or provide a means of escape.” (Ibid.)
Special relationships also have defined boundaries. They create a duty of care owed to a limited community, not the public at large. We have held that police officers are not in a special relationship with the citizens in their jurisdiction (see Williams v. State of California, supra, 34 Cal.3d at pp. 27-28), even when officers are aware of risks to a specific potential victim (see Davidson, supra, 32 Cal.3d at pp. 208-209). Nor is a government entity in a special relationship with all citizens who use its facilities. (Zelig v. County of Los Angeles, supra, 27 Cal.4th at p. 1130.) In declining to find a duty of care owed to courthouse visitors, we observed that a “county, ‘as with all public entities,’ has the responsibility to ‘exercise reasonable care to protect all of its citizens' [citation], but does not thereby become liable to each individual for all foreseeable harm.” (Id. at p. 1131.) Because a special relationship is limited to specific individuals, the defendant's duty is less burdensome and more justifiable than a broad-ranging duty would be. (See Rest.3d Torts, Liability for Physical and Emotional Harm, § 40, com. h, p. 43.)
Finally, although relationships often have advantages for both participants, many special relationships especially benefit the party charged with a duty of care. (Rest.3d Torts, Liability for Physical and Emotional Harm, § 40, com. h, p. 43.) Retail stores or hotels could not successfully operate, for example, without visits from their customers and guests.
b. The College Environment
The legal significance of the college-student relationship has changed with shifting cultural attitudes. Before the 1960s, colleges stood in loco parentis to students, who were viewed as being under their custody and institutional control. (Sokolow et al., College and University Liability for Violent Campus Attacks (2008) 34 J.C. & U.L. 319, 321 (hereafter Sokolow).) Although the role of parental stand-in may have given colleges some obligation to protect students (see Bradshaw v. Rawlings (3d Cir. 1979) 612 F.2d 135, 139 (Bradshaw)), the era also recognized parental immunity. Parents were largely immune from suit by their children, and colleges often enjoyed the same immunity, at least with respect to disciplining or regulating student conduct. (Lake, The Rise of Duty and the Fall of In Loco Parentis and Other Protective Tort Doctrines in Higher Education Law (1999) 64 Mo. L.Rev. 1, 5 (hereafter Lake).)
When rigid immunity defenses gave way to more flexible doctrines during the 1970s and 1980s, the view that colleges stood in loco parentis shifted to what Professor Peter Lake calls the “bystander” era in university liability. (See Lake, supra, 64 Mo. L.Rev. at pp. 11, 16.) Dramatic social changes of that time expanded the privacy and autonomy rights of adult students and, correspondingly, reduced the authority of college administrators to control student behavior. (Bradshaw, supra, 612 F.2d at p. 140.) Courts generally reacted to these changes by treating colleges like businesses. (Lake, at p. 12.) While the university might owe a duty as a landowner to maintain a safe premises, courts typically resisted finding a broader duty based on a special relationship with students. (See, e.g., Nero v. Kansas State University (1993) 253 Kan. 567, 580, 583-584 [861 P.2d 768, 778-780].) This was particularly so when injuries resulted from alcohol consumption or fraternity activity. (See Lake, at p. 12.) 4
California appellate decisions followed this trend. In Baldwin, supra, 123 Cal.App.3d 275, a student sued the California State University system after she was injured in a drunken, highway drag racing contest. Citing secondary school cases, the court assumed colleges “owe a duty to students who are on school grounds to supervise them and to enforce rules and regulations necessary for their protection.” (Id. at p. 281, italics added.) However, the more “difficult question” was whether colleges have a special relationship-based duty to protect students in other environments. (Id. at p. 283.) The court examined relevant policy factors and concluded the demise of colleges' in loco parentis role weighed against finding a duty of care. (Id. at p. 287.) Distinguishing special relationships recognized in other contexts, the court concluded the university lacked sufficient control over student behavior to justify imposing a duty to prevent on-campus drinking. (Id. at pp. 285-287, 290-291.)
Another California State student sued the university after a fellow student assaulted him at a dormitory “keg party.” (Crow v. State of California (1990) 222 Cal.App.3d 192, 197.) Under the facts of the case, the Court of Appeal rejected the argument that plaintiff's student status put him in a special relationship with the university. (Id. at p. 208.) The court had previously held that high schools have a duty to protect students from assault on campus. (See Leger v. Stockton Unified School Dist. (1988) 202 Cal.App.3d 1448.) However, it distinguished that case because high school attendance is mandatory, high schools “are directly in charge” of students on campus, and the attack in the Leger case was foreseeable. (Crow, at p. 208.) “Here, in contrast, plaintiff was an adult college student voluntarily participating in drinking beer at the dormitory. No claim was made that the [university] knew or should have known of any particular risk of harm at this particular dormitory.” (Ibid.) The court drew heavily from Baldwin's reasoning that the in loco parentis obligations of colleges have declined and modern universities cannot reasonably be expected to safeguard students from all the risks associated with alcohol use. (Crow, at p. 209; see Baldwin, supra, 123 Cal.App.3d at p. 287.)
In a third case from this era, a University of California student was raped by fellow students after a dormitory party. (Tanja H. v. Regents of University of California (1991) 228 Cal.App.3d 434.) The court relied on Baldwin and Crow to reject her claim against the school. Sensitive to the fact that both the plaintiff and her attackers had ingested significant amounts of alcohol, the court stressed that a duty to prevent alcohol-related crimes would require colleges to “impose onerous conditions on the freedom and privacy of resident students,” contrary to the modern view that adult students are generally responsible for their own welfare. (Tanja H., at p. 438.) Such a duty could not be recognized without resurrecting the university's former in loco parentis role. (Id. at pp. 438-439; see id. at p. 446 (conc. opn. of Kline, P. J.).)
When the particular problem of alcohol-related injuries is not involved, our cases have taken a somewhat broader view of a university's duties toward its students. Peterson v. San Francisco Community College Dist. (1984) 36 Cal.3d 799, 804 (Peterson), considered “whether a community college district and its agents have a duty to exercise due care to protect students from reasonably foreseeable assaults on the campus.” There, a student was injured during an attempted rape in a parking structure's stairway. (Id. at pp. 804-805.) The defendants were aware of similar assaults in the same area. (Id. at p. 805.) We held that, while the community college district was immune from liability for failing to provide adequate police protection, it did have a duty “to warn its students of known dangers posed by criminals on the campus.” (Id. at p. 804.) This holding was based on the district's status as a landowner, however. We discussed the general principle that a duty to warn or protect does not exist absent a special relationship (id. at p. 806) but ultimately concluded the district owed plaintiff a duty because she was on the premises as a business invitee (id. at pp. 808-809). Although a public entity's liability is generally limited to injuries resulting from a dangerous condition of the property itself, suit was not barred because the plaintiff had alleged the university's poor property maintenance increased the risk of criminal activity. (Id. at p. 812.)
In Avila v. Citrus Community College Dist. (2006) 38 Cal.4th 148 (Avila), a college baseball player was injured by a pitch to the head, and we examined a university's duty to students participating in intercollegiate sports. We noted that athletic competition is often an important part of the college environment, benefiting both the students who participate and the schools they represent. (Id. at p. 162.) Given these benefits, we held that a school hosting an athletic event owes a duty to student-players “to, at a minimum, not increase the risks inherent in the sport.” (Ibid.) While acknowledging and professing “no quarrel with” the Court of Appeal cases holding colleges have no general duty to ensure student welfare (ibid.), we concluded that recognizing a duty to students in school-sponsored athletic events was “plainly warranted by the relationship of the host school to all the student participants in the competitions it sponsors” (id. at p. 163).
This court has not addressed the college-student relationship since Avila. However, we recently discussed the relationship between students and a high school in William S. Hart, supra, 53 Cal.4th 861, where the plaintiff alleged sexual harassment by a high school guidance counselor. We explained that a school district has a special relationship with students “arising from the mandatory character of school attendance and the comprehensive control over students exercised by school personnel, ‘analogous in many ways to the relationship between parents and their children.’ ” (Id. at p. 869.) This special relationship gives secondary school personnel “the duty to use reasonable measures to protect students from foreseeable injury at the hands of third parties acting negligently or intentionally.” (Id. at p. 870.)
We must now decide whether a similar special relationship should be recognized in the college setting. Considering the unique features of the college environment, we conclude postsecondary schools do have a special relationship with students while they are engaged in activities that are part of the school's curriculum or closely related to its delivery of educational services.
Although comparisons can be made, the college environment is unlike any other. Colleges provide academic courses in exchange for a fee, but a college is far more to its students than a business. Residential colleges provide living spaces, but they are more than mere landlords. Along with educational services, colleges provide students social, athletic, and cultural opportunities. Regardless of the campus layout, colleges provide a discrete community for their students. For many students, college is the first time they have lived away from home. Although college students may no longer be minors under the law, they may still be learning how to navigate the world as adults. They are dependent on their college communities to provide structure, guidance, and a safe learning environment. “In the closed environment of a school campus where students pay tuition and other fees in exchange for using the facilities, where they spend a significant portion of their time and may in fact live, they can reasonably expect that the premises will be free from physical defects and that school authorities will also exercise reasonable care to keep the campus free from conditions which increase the risk of crime.” (Peterson, supra, 36 Cal.3d at p. 813.)
Colleges, in turn, have superior control over the environment and the ability to protect students. Colleges impose a variety of rules and restrictions, both in the classroom and across campus, to maintain a safe and orderly environment. They often employ resident advisers, mental health counselors, and campus police. They can monitor and discipline students when necessary. “While its primary function is to foster intellectual development through an academic curriculum, the institution is involved in all aspects of student life. Through its providing of food, housing, security, and a range of extracurricular activities the modern university provides a setting in which every aspect of student life is, to some degree, university guided.” (Furek v. University of Delaware (Del. 1991) 594 A.2d 506, 516.) Finally, in a broader sense, college administrators and educators “have the power to influence [students'] values, their consciousness, their relationships, and their behaviors.” (de Haven, The Elephant in the Ivory Tower: Rampages in Higher Education and the Case for Institutional Liability (2009) 35 J.C. & U.L. 503, 611 (hereafter de Haven).)
The college-student relationship thus fits within the paradigm of a special relationship. Students are comparatively vulnerable and dependent on their colleges for a safe environment. Colleges have a superior ability to provide that safety with respect to activities they sponsor or facilities they control. Moreover, this relationship is bounded by the student's enrollment status. Colleges do not have a special relationship with the world at large, but only with their enrolled students. The population is limited, as is the relationship's duration.
Of course, many aspects of a modern college student's life are, quite properly, beyond the institution's control. Colleges generally have little say in how students behave off campus, or in their social activities unrelated to school. It would be unrealistic for students to rely on their college for protection in these settings, and the college would often be unable to provide it. This is another appropriate boundary of the college-student relationship: Colleges are in a special relationship with their enrolled students only in the context of school-sponsored activities over which the college has some measure of control. (Cf. Avila, supra, 38 Cal.4th at p. 163 [“school-supervised” athletic events]; Patterson v. Sacramento City Unified School Dist. (2007) 155 Cal.App.4th 821, 830 [“ ‘school-sponsored’ ” community service project].) As commentators have observed, while there is an “emerging trend” of courts recognizing a special relationship between colleges and their students (Sokolow, supra, 34 J.C. & U.L. at p. 323), this relationship supports a duty of care only with respect to “risks that arise within the scope of the school-student relationship.” (Id. at pp. 323-324; see Peters, Protecting the Millennial College Student (2007) 16 S. Cal. Rev. L. & Soc. Just. 431, 459-460; Dall, Determining Duty in Collegiate Tort Litigation: Shifting the Paradigms of the College-Student Relationship (2003) 29 J.C. & U.L. 485, 485-487.)
Our recognition of a special relationship is consistent with decisions from other states. The Supreme Judicial Court of Massachusetts was one of the first to hold that colleges have a duty to protect their students against criminal attacks. In Mullins v. Pine Manor College (1983) 389 Mass. 47, 49-50 [449 N.E.2d 331, 334], a student was raped by an intruder to her dorm room. The court observed that colleges customarily take steps to protect their students from crime on campus, recognizing they had some obligation to do so. (Id. at p. 335.) In addition, colleges are the party best situated to implement safety measures. (Ibid.) Rejecting the argument that a duty to protect would be inconsistent with colleges' retreat from an in loco parentis role, the court reasoned: “The fact that a college need not police the morals of its resident students ․ does not entitle it to abandon any effort to ensure their physical safety. Parents, students, and the general community still have a reasonable expectation, fostered in part by colleges themselves, that reasonable care will be exercised to protect resident students from foreseeable harm.” (Id. at pp. 335-336.) Similarly, Florida's Supreme Court observed that the recognition of a special relationship between schools and minor students, based on in loco parentis principles, does not preclude finding that a different special relationship also exists between universities and their adult students. (Nova Southeastern University v. Gross (Fla. 2000) 758 So.2d 86, 89.) The Florida high court reasoned that a duty to protect arises from the relationship but is limited by the university's degree of control over student conduct in a given setting. (Ibid.) In addressing a university's liability for injuries arising from a fraternity hazing incident, the Supreme Court of Delaware also rejected the university-as-bystander rationale of cases like Baldwin and focused instead on “the uniqueness of the student-university relationship.” (Furek v. University of Delaware, supra, 594 A.2d at p. 518.) Delaware's court ultimately held that, while “[t]he university is not an insurer of the safety of its students nor a policeman of student morality, nonetheless, it has a duty to regulate and supervise foreseeable dangerous activities occurring on its property,” including “the negligent or intentional activities of third persons.” (Id. at p. 522.) This duty was limited, however, to instances where the university exercised some control over the environment or student behavior. (Ibid.) 5
The special relationship we now recognize is similarly limited. It extends to activities that are tied to the school's curriculum but not to student behavior over which the university has no significant degree of control. The incident here occurred in a chemistry laboratory while class was in session. Education is at the core of a college's mission, and the classroom is the quintessential setting for curricular activities. Perhaps more than any other place on campus, colleges can be expected to retain a measure of control over the classroom environment. Although collegiate class attendance may not be as strictly monitored as in secondary school, this distinction is not especially significant. All college students who hope to obtain a degree must attend classes and required laboratory sessions. It is reasonable for them to expect that their schools will provide some measure of safety in the classroom.
2. Policy Considerations Support Recognizing a Limited Duty
As discussed, there is generally no duty to protect others from the conduct of third parties. The “special relationship” doctrine is an exception to this general rule. (Delgado v. Trax Bar & Grill (2005) 36 Cal.4th 224, 235; Tarasoff, supra, 17 Cal.3d at pp. 435-436; Rest.3d Torts, Liability for Physical and Emotional Harm, § 40.) Accordingly, as a consequence of the special relationship recognized here, colleges generally owe a duty to use reasonable care to protect their students from foreseeable acts of violence in the classroom or during curricular activities.
Whether a new duty should be imposed in any particular context is essentially a question of public policy. “The existence of ‘ “ ‘[d]uty’ is not an immutable fact of nature ‘ “but only an expression of the sum total of those considerations of policy which lead the law to say that the particular plaintiff is entitled to protection.” ’ ” ' (Parsons[, supra,] 15 Cal.4th [at p.] 472.)” (Avila, supra, 38 Cal.4th at pp. 160-161.) As relevant here, the California Constitution “declare[s] that the right to public safety extends to public and private primary, elementary, junior high, and senior high school, and community college, California State University, University of California, and private college and university campuses, where students and staff have the right to be safe and secure in their persons.” (Cal. Const., art. I, § 28, subd. (a)(7).) College students “have the inalienable right to attend campuses which are safe, secure and peaceful.” (Id., § 28, subd. (f)(1).) Even assuming the constitutional provision is not self-executing (see Clausing v. San Francisco Unified School Dist. (1990) 221 Cal.App.3d 1224, 1236-1237), it clearly expresses a “fundamental public policy favoring measures to ensure the safety of California's public school students.” (William S. Hart, supra, 53 Cal.4th at p. 870, fn. 3.) 6
The court may depart from the general rule of duty, however, if other policy considerations clearly require an exception. (Kesner v. Superior Court (2016) 1 Cal.5th 1132, 1143 (Kesner); Cabral, supra, 51 Cal.4th at p. 771; see also Verdugo v. Target Corp. (2014) 59 Cal.4th 312, 344 (conc. opn. of Werdegar, J.).) We have identified several factors that may, on balance, justify excusing or limiting a defendant's duty of care. These include: “the foreseeability of harm to the plaintiff, the degree of certainty that the plaintiff suffered injury, the closeness of the connection between the defendant's conduct and the injury suffered, the moral blame attached to the defendant's conduct, the policy of preventing future harm, the extent of the burden to the defendant and consequences to the community of imposing a duty to exercise care with resulting liability for breach, and the availability, cost, and prevalence of insurance for the risk involved.” (Rowland v. Christian (1968) 69 Cal.2d 108, 113 (Rowland).) These factors must be “evaluated at a relatively broad level of factual generality.” (Cabral, at p. 772.) In considering them, we determine “not whether they support an exception to the general duty of reasonable care on the facts of the particular case before us, but whether carving out an entire category of cases from that general duty rule is justified by clear considerations of policy.” (Ibid.) In other words, the duty analysis is categorical, not case-specific. (See Kesner, at p. 1144.)
The Rowland factors fall into two categories. The first group involves foreseeability and the related concepts of certainty and the connection between plaintiff and defendant. The second embraces the public policy concerns of moral blame, preventing future harm, burden, and insurance availability. The policy analysis evaluates whether certain kinds of plaintiffs or injuries should be excluded from relief. (Kesner, supra, 1 Cal.5th at p. 1145.) We conclude that violence against students in the classroom or during curricular activities, while rare, is a foreseeable occurrence, and considerations of public policy do not justify categorically barring an injured student's claims against the university.
a. Foreseeability Factors
(1) “The most important factor to consider in determining whether to create an exception to the general duty to exercise ordinary care ․ is whether the injury in question was foreseeable.” (Kesner, supra, 1 Cal.5th at p. 1145, italics added; see Tarasoff, supra, 17 Cal.3d at p. 434.) In examining foreseeability, “the court's task ․ ‘is not to decide whether a particular plaintiff's injury was reasonably foreseeable in light of a particular defendant's conduct, but rather to evaluate more generally whether the category of negligent conduct at issue is sufficiently likely to result in the kind of harm experienced that liability may appropriately be imposed ․’ ” (Cabral, supra, 51 Cal.4th at p. 772; accord Parsons, supra, 15 Cal.4th at p. 476.)
Phrased at the appropriate level of generality, then, the question here is not whether UCLA could predict that Damon Thompson would stab Katherine Rosen in the chemistry lab. It is whether a reasonable university could foresee that its negligent failure to control a potentially violent student, or to warn students who were foreseeable targets of his ire, could result in harm to one of those students. Violent unprovoked attacks by and against college students, while still relatively uncommon, are happening more frequently. (See de Haven, supra, 35 J.C. & U.L. at p. 510.) One example occurred on April 16, 2007 at Virginia Polytechnic Institute and State University (Virginia Tech), when an emotionally disturbed underclassman barred the doors to a classroom building, then walked the halls shooting people, killing five professors and 24 students. (See id. at pp. 554-566.) He left over a dozen more wounded before taking his own life. (Id. at p. 566.) Although mass shootings on college campuses had occurred before, the record demonstrates that the Virginia Tech tragedy prompted schools to reexamine their campus security policies. A January 2008 report of the University of California Campus Security Task Force recommended several improvements in student mental health services, emergency communications, preparedness, and hazard mitigation across all campuses. In April 2008, almost exactly one year after the Virginia Tech shootings, a special review task force of the International Association of Campus Law Enforcement Administrators published a “Blueprint for Safer Campuses,” with several recommendations for assessing and responding to potential threats. Colleges across the country, including the public universities of California, created threat assessment protocols and multidisciplinary teams to identify and prevent campus violence. Thus, particularly after the Virginia Tech shootings focused national attention on the issue, colleges have been alert to the possibility that students, particularly those with mental health issues, may lash out violently against those around them. Even a comparatively rare classroom attack is a foreseeable occurrence that colleges have been equipping themselves to address for at least the past decade.
Whether a university was, or should have been, on notice that a particular student posed a foreseeable risk of violence is a case-specific question, to be examined in light of all the surrounding circumstances. Any prior threats or acts of violence by the student would be relevant, particularly if targeted at an identifiable victim. (See Mullins v. Pine Manor College, supra, 449 N.E.2d at p. 337.) Other relevant facts could include the opinions of examining mental health professionals, or observations of students, faculty, family members, and others in the university community. Such case-specific foreseeability questions are relevant in determining the applicable standard of care or breach in a particular case. They do not, however, inform our threshold determination that a duty exists.
(2) The second factor, “the degree of certainty that the plaintiff suffered injury” (Rowland, supra, 69 Cal.2d at p. 113, italics added), may come into play when the plaintiff's claim involves intangible harm, such as emotional distress. (Kesner, supra, 1 Cal.5th at p. 1148.) Here, however, we are addressing claims for physical injuries that are capable of identification. (See ibid.)
(3) The third factor is “the closeness of the connection between the defendant's conduct and the injury suffered.” (Rowland, supra, 69 Cal.2d at p. 113, italics added.) “Generally speaking, where the injury suffered is connected only distantly and indirectly to the defendant's negligent act, the risk of that type of injury from the category of negligent conduct at issue is likely to be deemed unforeseeable. Conversely, a closely connected type of injury is likely to be deemed foreseeable.” (Cabral, supra, 51 Cal.4th at p. 779.) The negligence alleged here is the failure to prevent a classroom assault, either by controlling the perpetrator or warning the potential victim. Although the immediate cause of injury in such cases will be the perpetrator's violent outburst, we have explained that the existence of an intervening act does not necessarily attenuate a defendant's negligence. Rather, “the touchstone of the analysis is the foreseeability of that intervening conduct.” (Kesner, supra, 1 Cal.5th at p. 1148.) When circumstances put a school on notice that a student is at risk to commit violence against other students, the school's failure to take appropriate steps to warn or protect foreseeable victims can be causally connected to injuries the victims suffer as a result of that violence. Although a criminal act is always shocking to some degree, it is not completely unpredictable if a defendant is aware of the risk. (See, e.g., Randi W. v. Murdoc Joint Unified School Dist. (1997) 14 Cal.4th 1066, 1077-1078; cf. Thompson v. County of Alameda (1980) 27 Cal.3d 741, 758 [crime committed after release of a potentially dangerous offender “is statistically foreseeable”].)
b. Policy Factors
Although Rowland's foreseeability factors weigh in favor of recognizing a duty of care, we must also consider whether public policy requires a different result. (See Kesner, supra, 1 Cal.5th at pp. 1149-1150; Cabral, supra, 51 Cal.4th at p. 781.) “A duty of care will not be held to exist even as to foreseeable injuries ․ where the social utility of the activity concerned is so great, and avoidance of the injuries so burdensome to society, as to outweigh the compensatory and cost-internalization values of negligence liability.” (Merrill v. Navegar, Inc., supra, 26 Cal.4th at p. 502; see Parsons, supra, 15 Cal.4th at p. 476.)
(1) Some measure of moral blame does attach to a university's negligent failure to prevent violence against its students. “We have previously assigned moral blame, and we have relied in part on that blame in finding a duty, in instances where the plaintiffs are particularly powerless or unsophisticated compared to the defendants or where the defendants exercised greater control over the risks at issue.” (Kesner, supra, 1 Cal.5th at p. 1151.) With the decline of colleges' in loco parentis role, adult students can no longer be considered particularly powerless or unsophisticated. “While the college will take a lead role in campus security and safety issues, ‘babysitting’ would defeat the proper role of most colleges in most instances. Most often the proper student/college relationship is one of shared responsibility.” (Lake, supra, 64 Mo. L.Rev. at p. 26.) Nevertheless, compared to students, colleges will typically have access to more information about potential threats and a superior ability to control the environment and prevent harm. This disparity in knowledge and control tips the balance slightly in favor of duty.
(2) “The overall policy of preventing future harm is ordinarily served, in tort law, by imposing the costs of negligent conduct upon those responsible. The policy question is whether that consideration is outweighed, for a category of negligent conduct, by laws or mores indicating approval of the conduct or by the undesirable consequences of allowing potential liability.” (Cabral, supra, 51 Cal.4th at pp. 781-782, italics added.) UCLA argues imposing a duty of care would discourage colleges from offering comprehensive mental health and crisis management services. Rather than become engaged in the treatment of their mentally ill students, colleges would have an incentive to expel anyone who might pose a remote threat to others. We understand that the recognition of a duty of care will force schools to balance competing goals and make sometimes difficult decisions. The existence of a duty may give some schools a marginal incentive to suspend or expel students who display a potential for violence. It might make schools reluctant to admit certain students, or to offer mental health treatment. But colleges' decisions in this area are restricted to some extent by laws such as the Americans with Disabilities Act (42 U.S.C. § 12101 et seq.). In addition, the market forces that drove colleges across the country to adopt sophisticated violence prevention protocols in the wake of the Virginia Tech incident would likely weigh against the dismantling of these protections. Colleges and universities also may have options short of expelling or denying admission to deal with potentially violent students. What constitutes reasonable care will vary with the circumstances of each case. On the whole, however, if such steps can avert violent episodes like the one that occurred here, recognizing a duty serves the policy of preventing future harm.
UCLA also predicts that legal recognition of a duty might deter students from seeking mental health treatment, or being candid with treatment providers, for fear that their confidences would be disclosed. To a large extent, however, the conditions that might influence student perceptions about confidentiality already exist. Psychotherapists' duty to warn about patient threats is well established in California. Indeed, despite fears that this duty would deter people from seeking treatment and irreparably damage the psychotherapist-patient relationship (see, e.g., Tarasoff, supra, 17 Cal.3d at pp. 458-460 (dis. opn. of Clark, J.)), empirical studies have produced “no evidence thus far that patients have been discouraged from coming to therapy, or discouraged from speaking freely once there, for fear that their confidentiality will be breached” (Buckner & Firestone, “Where the Public Peril Begins” 25 Years After Tarasoff, 21 J. Legal Med. 187, 221). Moreover, as the record in this case demonstrates, threat assessment and violence prevention protocols are already prevalent on university campuses. Recognizing that the university owes its students a duty of care under certain circumstances is unlikely to appreciably change this landscape.
(3) Which leads to the next policy factor: the burden that recognizing a tort duty would impose on the defendant and the community. (See Rowland, supra, 69 Cal.2d at p. 113.) UCLA and some amici curiae place considerable weight on this factor, arguing it would be prohibitively expensive and impractical to make university professors and administrators the “insurers” of student safety. But the record shows that UCLA, like other colleges across the country, has already developed sophisticated strategies for identifying and defusing potential threats to student safety. The school created multidisciplinary teams of trained staff members and professionals for this very purpose. Indeed, one of these teams was closely monitoring Thompson's behavior. UCLA also expressly marketed itself to prospective students, and their parents, as “one of the safest campuses in the country.” These enhanced safety measures came at a price, but students paid the bill. In 2007, schools in the University of California system raised mandatory registration fees 3 percent to improve student mental health services, and they planned further increases to implement all of the violence prevention measures recommended by the Campus Security Task Force. Because the record reflects that colleges have already focused considerable attention on identifying and responding to potential threats, and have funding sources available for these efforts, it does not appear that recognizing a legal duty to protect students from foreseeable threats would pose an unmanageable burden.
The duty we recognize here is owed not to the public at large but is limited to enrolled students who are at foreseeable risk of being harmed in a violent attack while participating in curricular activities at the school. Moreover, universities are not charged with a broad duty to prevent violence against their students. Such a duty could be impossible to discharge in many circumstances. Rather, the school's duty is to take reasonable steps to protect students when it becomes aware of a foreseeable threat to their safety. The reasonableness of a school's actions in response to a potential threat is a question of breach.
(4) The final policy factor in a duty analysis is the availability of insurance for the risk involved. (Rowland, supra, 69 Cal.2d at p. 113.) While not addressing this issue specifically, UCLA has offered no reason to doubt colleges' ability to obtain coverage for the negligence liability under consideration.
Accordingly, an examination of the Rowland factors does not persuade us to depart from our decision to recognize a tort duty arising from the special relationship between colleges and their enrolled students. Specifically, we hold that colleges have a duty to use reasonable care to protect their students from foreseeable violence during curricular activities.7
We emphasize that a duty of care is not the equivalent of liability. Nor should our holding be read to create an impossible requirement that colleges prevent violence on their campuses. Colleges are not the ultimate insurers of all student safety. We simply hold that they have a duty to act with reasonable care when aware of a foreseeable threat of violence in a curricular setting. Reasonable care will vary under the circumstances of each case. Moreover, some assaults may be unavoidable despite a college's best efforts to prevent them. Courts and juries should be cautioned to avoid judging liability based on hindsight.
Our conclusion that universities owe a duty to protect students from foreseeable violence during curricular activities does not end the matter, however.8 UCLA's petition for writ of mandate argued summary judgment should have been granted for three reasons. First, UCLA claimed it owed Rosen no duty of care; second, it did not negligently breach any duty to Rosen; and third, various immunity statutes shielded the school and individual defendants from liability.9 The Court of Appeal majority agreed that UCLA owed no duty of care and did not reach the other arguments. Thus, while we conclude UCLA did owe a duty to protect Rosen, we will remand for the Court of Appeal to decide whether triable issues of material fact remain on the questions of breach and immunity. In regard to breach, we note that the appropriate standard of care for judging the reasonableness of the university's actions remains an open question, which the parties are free to litigate on remand. UCLA's argument that there was little more it reasonably could have done to prevent the assault may be relevant to this determination.
Finally, apart from their diverging views on duty, the majority and dissenting justices below agreed that Rosen had failed to plead or support a claim against UCLA psychologist Nicole Green under Civil Code section 43.92. That statute provides that a psychotherapist is not liable for failing to protect against a patient's violent behavior unless the patient has told the therapist about a serious threat of physical violence against a reasonably identifiable victim. (Civ. Code, § 43.92, subd. (a).) Because Rosen's petition for review was limited to the issue of duty and did not challenge the Court of Appeal's conclusion regarding section 43.92, we decline her invitation to revisit the ruling now.
The decision of the Court of Appeal is reversed. The case is remanded for further proceedings consistent with this opinion.
CONCURRING OPINION BY CHIN, J.
I agree with the majority that universities have a duty to warn or protect their students from foreseeable acts of violence “in the classroom.” (Maj. opn., ante, at p. 9.) However, for several reasons, I do not join the majority opinion insofar as it would extend this duty beyond the classroom, to encompass more broadly “curricular activities” (ibid.) and activities “closely related to [the] delivery of educational services” (id. at p. 17).
First, we need not decide whether the duty extends beyond the classroom, because the attack in this case occurred in a classroom and, as the majority states, “[t]he negligence alleged here is the failure to prevent a classroom assault.” (Maj. opn., ante., at p. 25, italics added.) Notably, the majority rightly declines to decide several other issues that we need not resolve in order to dispose of this case, i.e., whether universities have a duty to control the behavior of students (id. at p. 11) and alternate theories of duty based on an implied-in-fact contract or the negligent undertaking doctrine (id. at p. 30, fn. 8). In my view, we should exercise similar restraint in addressing a university's duty to protect or warn, and should confine our consideration of the issue to what is necessary to decide this case.
Second, in terms of the various factors courts apply to determine whether to impose a duty as a matter of public policy, activities outside the classroom differ in potentially significant ways from activities inside the classroom. As the majority explains, among the relevant factors is the extent of the defendant's control in the particular setting over the environment and third party behavior. (Maj. opn., ante, at pp. 18-20, 26.) As the majority also explains, “[p]erhaps more than any other place on campus, colleges can be expected to retain a measure of control over the classroom environment.” (Id. at p. 20.) Implicit in this statement is recognition that the extent of a university's control over the environment and student behavior is likely to be considerably less outside of the classroom. Indeed, the extent of a university's control in a nonclassroom setting varies considerably depending on the particular activity and the particular setting. It may be that, as to any given nonclassroom activity, a university's control is sufficient, from a public policy perspective, to impose a duty to protect or warn. But I would leave that question for a case that presents the issue on concrete facts, rather than broadly conclude, in a case involving classroom activity, that a university's control in nonclassroom settings is sufficient to impose a duty to protect or to warn.
Finally, the majority's conclusion seems likely to create confusion, because the majority offers no guidance as to which nonclassroom activities qualify as either “curricular” (maj. opn., ante, at p. 9) or “closely related to [the] delivery of educational services” (id. at p. 17), or what factors are relevant to this determination. This omission no doubt results from the circumstance, as already noted, that this case involves classroom activity, and that the majority is thus deciding the duty question as to nonclassroom activities in the abstract, without any concrete facts to guide its analysis. For this reason, and the others mentioned above, although I concur in the judgment, I do not join the majority's conclusion that a university's duty to warn or protect extends beyond the classroom, to encompass more broadly “curricular activities” (id. at p. 9) and activities “closely related to [the] delivery of educational services” (id. at p. 17).
1. We use the terms “college” and “university” interchangeably to refer to all schools that provide postsecondary education to enrolled students.
2. We refer to these defendants collectively, and the school itself, by the acronym UCLA.
3. We speak here of a university's duty “to protect” its students from foreseeable harm. However, in an appropriate case, this duty may be fully discharged if adequate warnings are conveyed to the students at risk. (Cf. Tarasoff v. Regents of University of California (1976) 17 Cal.3d 425, 439-440 (Tarasoff) [discussing therapist's duty to protect discharged through warnings].)
4. “These cases have become known as the ‘bystander’ cases because in each of them the university was cast in the role of a legal bystander to ‘uncontrollable’ student actions and drinking.” (Lake, supra, 64 Mo. L.Rev. at p. 16.)
5. Some state courts, unwilling to recognize a special relationship between colleges and their adult students, have nevertheless imposed a duty to protect under landlord-invitee principles. (See, e.g., Nero v. Kansas State University, supra, 861 P.2d at p. 780; Johnson v. State (1995) 77 Wn.App. 934, 941 [894 P.2d 1366, 1370].)
6. The Court of Appeal majority suggested recognizing a duty of care based on the special relationship doctrine would undermine the liability limits of Government Code section 835. Under that statute, a public landowner's liability to business invitees is limited to injuries caused by a physical defect in the property. (See Peterson, supra, 36 Cal.3d at p. 810.) Generally, third party criminal conduct unrelated to the property's physical condition does not constitute a “dangerous condition” for which a public entity may be liable under section 835. (Zelig v. County of Los Angeles, supra, 27 Cal.4th at pp. 1134-1135; Peterson, at pp. 810-811.) However, UCLA offers no persuasive reason why statutory limits on the scope of an entity's duties under one type of special relationship, such as the landowner-invitee relationship, must restrict the entity's duties in the context of other relationships. On the contrary, despite limits on premises liability, we have explained that “another body of law establishes that public schools and universities owe certain non-property-based duties to their students.” (Avila, supra, 38 Cal.4th at p. 158.)
7. To the extent they are inconsistent with our holdings regarding the special relationship between colleges and students, or colleges' duty of care, we disapprove Baldwin v. Zoradi, supra, 123 Cal.App.3d 275, Crow v. State of California, supra, 222 Cal.App.3d 192, Tanja H. v. Regents of University of California, supra, 228 Cal.App.3d 434, Ochoa v. California State University (1999) 72 Cal.App.4th 1300, and Stockinger v. Feather River Community College (2003) 111 Cal.App.4th 1014.
8. Because we decide the university had a duty arising out of its special relationship with Rosen, we do not address Rosen's alternate theories of duty based on an implied-in-fact contract or the negligent undertaking doctrine.
9. Specifically, the school relied on Government Code section 856, which immunizes public entities' decisions about involuntary confinement, and section 820.2, which immunizes public employees' discretionary acts.
WE CONCUR: CANTIL-SAKAUYE, C. J. LIU, J. CUÉLLAR, J. KRUGER, J. RICHMAN, J.*