OPINION

ARGUELLES, J. [*]

John R., then a 14-year-old junior high school student, allegedly was sexually molested by his mathematics teacher while he was at the teacher's apartment participating in an officially sanctioned, extracurricular program. The principal question before us is whether the school district that employed the teacher can be held vicariously liable for the teacher's acts under the doctrine of respondeat superior. We hold that the doctrine is not applicable in these circumstances and that while the school district may be liable if its own direct negligence is established, it cannot be held vicariously liable for its employee's torts.

At the time of the incidents giving rise to this case, John R. was a ninth grade student at a junior high school in the Oakland Unified School District (district). His mathematics teacher, who had also taught John in the seventh grade, asked John to participate in the school's instructional, work-experience program, under which students received both school credit and monetary payments for assisting teachers by, for example, helping to correct other students' papers. The nature of the tasks would suggest that the program was aimed mainly at high-performing students. John had a history of poor grades in mathematics, but his marks in this teacher's class reflected what his attorney, no doubt ironically, termed "a remarkable increase in his ability to do math...."

Whether legitimately or through artificially inflated grades, John was allowed to participate in the program. Performance of the required work by students at teachers' homes was an option authorized by the district, and the teacher either encouraged or required John to come to his apartment for this purpose. Over the course of many sessions at the teacher's apartment, the teacher sought to develop a close relationship with John as the boy's tutor and counselor, and ultimately endeavored to seduce him. The teacher attempted to convince John that engaging in sex acts with him would be a constructive part of their relationship and, at times, threatened to give John failing grades if John would not go along with his desires and said he would tell people that John had solicited sex from him. On one occasion in February of 1981, the teacher succeeded in pressuring John into sexual acts, including oral copulation and anal intercourse.

When John protested and told the teacher he would report the incidents to his parents, the teacher threatened to retaliate against him if he revealed what had taken place. As a result of these threats, and his embarrassment and shame at what had happened, John did not disclose the incidents to anyone for a number of months. John finally told his father about the molestation 10 months later in December 1981.

John's mother reported the incident to the district that same month, speaking to the vice-principal of John's school and a district community relations representative and asking them how she should proceed. She was advised to put the matter in the hands of the police, who were then told of the molestation by the district representative. John's mother also contacted an attorney and was advised by him to wait for the criminal investigation to substantiate John's charges before she pursued any civil remedy. [2]

John's parents, on behalf of their son and on their own behalf, brought suit against the teacher and the district, alleging that the district was vicariously liable for the teacher's acts and directly liable for its own negligence. After two rounds of demurrers and amended complaints, the district's demurrer to plaintiffs' third amended complaint was sustained without leave to amend as to the four causes of action on which plaintiffs sought to hold the district indirectly liable for the teacher's acts under the doctrine of respondeat superior. The case then proceeded to trial against the teacher on all causes of action and against the district limited to those causes of action premised upon its direct liability for negligent hiring and supervision of the teacher. At the outset of trial, the district's motion for nonsuit as to those remaining claims was granted on grounds unrelated to the merits ( post, p. 444), and judgment was entered in the district's favor on all causes of action against it.

The Court of Appeal reversed both the grant of nonsuit and the earlier order sustaining the district's demurrer to those causes of action against it premised on a theory of vicarious liability, reasoning that the facts as pleaded by plaintiffs could allow the trier of fact to find the district responsible for the tort of its employee because the teacher's misconduct, although not within or contemplated by his official duties, was made possible by his use, and abuse, of the official, job-created authority he was given over the boy. We granted review to determine whether the Court of Appeal correctly resolved this unsettled and significant question.

(1a) Before we turn to the vicarious liability issue, we must first address a threshold question whether plaintiffs complied in timely fashion with the requirements of the California Tort Claims Act (Gov. Code, § 900 et seq.) for if we were to conclude they did not, all of their claims against the district would be barred on that ground (Gov. Code, § 945.4; see Whitfield v. Roth (1974) 10 Cal.3d 874, 883 [ 112 Cal. Rptr. 540 , 519 P.2d 588 ]), and we would have no occasion to consider whether the district could be held vicariously liable for the tort of its employee. The question arises here because plaintiffs did not present a written claim to the district within 100 days of the accrual of their causes of action measured from the date that John was molested as then required by Government Code section 911.2, nor did they present an application for leave to file a late claim within 1 year of that time, as required by Government Code section 911.4, subdivision (b). [3]

Although the trial court initially found that plaintiffs should be excused from the statutory claim requirement, it later granted a nonsuit in favor of the district on this issue. The Court of Appeal, however, reversed the trial court on this point, holding that plaintiffs' late-claim application, presented to the district in May 1982 some 15 months after the assault, was made within the allowable 1-year period because, under the "delayed discovery" doctrine (see Neel v. Magana, Olney, Levy, Cathcart & Gelfand (1971) 6 Cal.3d 176, 187 [ 98 Cal. Rptr. 837 , 491 P.2d 421 ]), plaintiffs' causes of action should not be deemed to have accrued until John told his parents about the incidents in December 1981. The district contends the Court of Appeal erred in finding the claim timely on that basis.

Our initial review of this issue raised a serious question in our minds whether the relevant authorities supported application of a delayed-discovery theory of accrual on the facts of this case, but it appeared that, under the reasoning of a number of recent Court of Appeal decisions (see, e.g., Snyder v. Boy Scouts of America, Inc. (1988) 205 Cal. App.3d 1318, 1324 [ 253 Cal. Rptr. 156 ]; DeRose v. Carswell (1987) 196 Cal. App.3d 1011, 1026 [ 242 Cal. Rptr. 368 ]), the facts alleged in the complaint, if proven, might well demonstrate that the claim was timely filed under a theory of equitable estoppel. We accordingly requested the parties to file supplemental briefs on the question and, having fully considered the issue, now conclude it is appropriate to remand the timeliness question for a factual determination on the applicability of equitable estoppel. [4]

(2a) It is well settled that a public entity may be estopped from asserting the limitations of the claims statute where its agents or employees have prevented or deterred the filing of a timely claim by some affirmative act. (See, e.g., Fredrichsen v. City of Lakewood (1971) 6 Cal.3d 353, 357-359 [ 99 Cal. Rptr. 13 , 491 P.2d 805 ]; Rand v. Andreatta (1964) 60 Cal.2d 846, 850 [ 36 Cal. Rptr. 846 , 389 P.2d 382 ]; Bruce v. Jefferson Union High Sch. Dist. (1962) 210 Cal. App.2d 632, 635 [ 26 Cal. Rptr. 762 ].) Estoppel most commonly results from misleading statements about the need for or advisability of a claim; actual fraud or the intent to mislead is not essential. (See Industrial Indem. Co. v. Ind. Acc. Com. (1953) 115 Cal. App.2d 684, 689-690 [ 252 P.2d 649 ].) A fortiori, estoppel may certainly be invoked when there are acts of violence or intimidation that are intended to prevent the filing of a claim. (See, e.g., DeRose v. Carswell, supra, 196 Cal. App.3d at p. 1026 ; Longo v. Pittsburgh and Lake Erie Railroad Co., N.Y.C. Sys. (3d Cir.1966) 355 F.2d 443, 444 .) (1b) And here, the teacher's threats to retaliate against John if the boy reported the incidents of sexual molestation allegedly did just that.

Although the teacher's alleged threats in this case were no doubt motivated largely by self-interest, rather than to prevent John from filing a claim against the district, it would clearly be inconsistent with the equitable underpinnings of the estoppel doctrine to permit the district to benefit to plaintiffs' detriment by such threats. (Cf. Dettamanti v. Lompoc Union School Dist. (1956) 143 Cal. App.2d 715, 722-723 [ 300 P.2d 78 ].) Putting aside for the moment the substantive question whether the district may be held vicariously liable for the teacher's alleged molestation of John, we have no hesitation in concluding that the teacher's threats may be taken into account in resolving the procedural status of plaintiffs' claims against the district. [5] Plaintiffs alleged that the district's own negligent conduct e.g., the failure to make an adequate investigation of the teacher's background before hiring him and the failure to guarantee that the extracurricular program was properly supervised renders it liable for their injuries. Assuming plaintiffs can establish their case, it would plainly be inequitable to permit the district to escape liability only because the teacher's threats succeeded in preventing his victim from disclosing the molestation until the time for filing a claim against the district had elapsed. We conclude that, for purposes of applying equitable estoppel, the time for filing a claim against the district was tolled during the period that the teacher's threats prevented plaintiffs from pursuing their claims.

We do not determine as a matter of law that the district is estopped from asserting as a defense plaintiffs' failure to comply with the claims statutes. That is a question of fact for the trial court on remand. ( Henry v. City of Los Angeles (1962) 201 Cal. App.2d 299, 306 [ 20 Cal. Rptr. 440 ].) Because the trial court did not analyze the timeliness of plaintiffs' claims in this light, it made no findings on any of the factual issues relevant to the equitable estoppel doctrine. It did not determine (1) whether any threats were in fact made by the teacher, (2) when the effect of any such threats ceased, or (3) whether plaintiffs acted within a reasonable time after the coercive effect of the threats had ended. (See DeRose v. Carswell, supra, 196 Cal. App.3d at pp. 1026-1027; Bertorelli v. City of Tulare (1986) 180 Cal. App.3d 432, 440-442 [ 225 Cal. Rptr. 582 ].) In the absence of an adverse finding on any of these points, it was error to have granted a nonsuit in favor of the district on the timeliness question; on remand, the trial court must resolve these matters to determine whether the action may go forward. [6]

Our decision that a remand is required on the timeliness question brings to the fore the more pressing question in this case, as we must now decide whether plaintiffs should not only have the opportunity to pursue their claims against the district premised upon its direct liability for negligent hiring and supervision of the teacher, but whether they may also seek to hold the district vicariously liable for the acts of its employee under the doctrine of respondeat superior.

The principles governing application of the doctrine of respondeat superior to make an employer responsible for the torts of an employee are familiar and easily stated. (3) "[A]n employer's liability extends to torts of an employee committed within the scope of his employment. [Citation.] This includes willful and malicious torts as well as negligence. [Citation.]" ( Martinez v. Hagopian (1986) 182 Cal. App.3d 1223, 1227 [ 227 Cal. Rptr. 763 ].) Whether a tort was committed within the scope of employment is ordinarily a question of fact; it becomes a question of law, however, where the undisputed facts would not support an inference that the employee was acting within the scope of his employment. ( Alma W. v. Oakland Unified School Dist. (1981) 123 Cal. App.3d 133, 138 [ 176 Cal. Rptr. 287 ].)

Scope of employment is viewed broadly in this context. (4) "The fact that an employee is not engaged in the ultimate object of his employment at the time of his wrongful act does not preclude attribution of liability to an employer. [Citation.]" ( Alma W., supra, 123 Cal. App.3d at p. 139 .) The employer is not liable if the employee substantially departs from his duties for purely personal reasons ( ibid. ), but "where the employee is combining his own business with that of his employer, or attending to both at substantially the same time, no nice inquiry will be made as to which business he was actually engaged in at the time of injury, unless it clearly appears that neither directly nor indirectly could he have been serving his employer." ( Lockheed Aircraft Corp. v. Ind. Acc. Com. (1946) 28 Cal.2d 756, 758-759 [ 172 P.2d 1 ], quoted with approval in Perez v. Van Groningen & Sons, Inc. (1986) 41 Cal.3d 962, 970 [ 227 Cal. Rptr. 106 , 719 P.2d 676 ].)

(5a) The question before us here is whether an employer (specifically, a school district) can be held liable for a sexual assault committed by an employee (here, a teacher) on another person (particularly, on a student committed to that teacher's supervision). The natural, initial reaction is "No! Of course not!" A more personal escapade less related to an employer's interests is difficult to imagine. But the question is not so easily disposed of. It is closer than might appear upon first examination, but we ultimately conclude the Court of Appeal erred in reversing the trial court on this point.

The courts of other jurisdictions and our own Courts of Appeal have struggled in recent years over whether and how to apply the respondeat superior doctrine to the sexual assaults or misconduct of employees. The historical and perhaps still prevailing point of view declines to impose vicarious liability in such circumstances. [7] But the other school of thought has its adherents as well, [8] and what we must decide here is if such liability should be imposed in light of the law of this state and the purposes of the doctrine.

Plaintiffs urge us to take the approach adopted by the Court of Appeal, which looked not to whether a teacher's sexual abuse of a student is foreseeable in the sense that it is characteristic of the job or broadly incidental to a school district's activities (see Perez, supra, 41 Cal.3d at p. 968 ), but rather to the nature of the teacher-student relationship. The essence of their argument is that vicarious liability is appropriate when the tort is a consequence of the employer's conferring of official authority on the employee. That is, plaintiffs would have us impose liability on the employer if: (1) there is an official, job-created, hierarchical relationship by which the employee is given authority over a certain, and possibly limited, class of persons; and (2) there is a sufficient nexus between the exercise of that authority and the commission of the tort that it was foreseeable the authority so conferred might be abused to the detriment of the victim.

We recognize that this theory is not without substance. It serves to explain certain of the decisions imposing liability on employers for the sexual misconduct of their employees and to distinguish those cases declining to hold the employer responsible, and finds some support in the language of both lines of cases.

In Alma W., supra, 123 Cal. App.3d 133 , for example, holding a school district not liable for a janitor's rape of a student, the court noted: "In each of the[] cases [in which an employer was held vicariously liable for an employee's tortious assault on a third party], a work-related dispute preceded the assault.... [I]t is clear from this line of cases ... that the acts leading up to the tort must bear some relation to the employee's duties." ( Id., at p. 141.) Unlike the present case, there was plainly no work-related basis for the janitor's assault, no authority given him over the student, and no more than "an independent, self-serving pursuit wholly unrelated to his custodial duties." ( Ibid. )

In White, supra, 166 Cal. App.3d 566 , holding a county subject to vicarious liability for the threats made by a deputy sheriff to rape and murder a motorist he had stopped, the court employed substantially the reasoning plaintiffs would have us adopt here: "A police officer is entrusted with a great deal of authority. This authority distinguishes the situation here from the facts of Alma W. Unlike a school custodian, the police officer carries the authority of the law with him into the community. The officer is supplied with a conspicuous automobile, a badge and a gun to ensure immediate compliance with his directions. The officer's method of dealing with this authority is certainly incidental to his duties; indeed, it is an integral part of them.... [T]he wrongful acts flowed from the very exercise of this authority. [¶] It follows that the employer/government must be responsible for acts done during the exercise of this authority." ( Id., at p. 571.)

And in Jeffrey E., supra, 197 Cal. App.3d 718 , holding a church not responsible for a Sunday school teacher's molestation of a minor, the same Court of Appeal that decided White explained its earlier decision: "The distinguishing feature in White is that the errant conduct arose out of an abuse of the employee's official authority. By virtue of the exercise of this authority, the police officer was able to perpetrate his assault. The focus is not on whether the police officer's activity is either characteristic or foreseeable, but rather on whether the assault arose out of the exercise of job-created authority over the plaintiff." ( Id., at p. 723.)

Plaintiffs urge us to adopt this rationale and to find the authority given a teacher over a student tantamount to that given the police officer over the community at large. We agree that in the eyes of a child, a teacher's authority can be very great. And here the complaint alleged the teacher used his authority to obtain John's participation in the extracurricular program and thereby obtained the boy's presence at the teacher's home away from other eyes. The teacher told John that sexual conduct was part of a teacher-student relationship and was intended to help John with his problems. The extensive control teachers are authorized to exercise over their students supports the analogy to White, supra, 166 Cal. App.3d 566 , and lends some credence to the proposition that a school district should not be immune from liability where a teacher's molestation of a student directly flows from the exercise of that job-created authority.

But although the facts of this case can be made to fit a version of the respondeat superior doctrine, we are unpersuaded that they should be or that the doctrine is appropriately invoked here. We draw our decision not from the various factual scenarios in which vicarious liability has or has not been imposed on employers for the torts of their employees, but instead from the underlying rationale for the respondeat superior doctrine. [9]

(6) "The principal justification for the application of the doctrine of respondeat superior in any case is the fact that the employer may spread the risk through insurance and carry the cost thereof as part of his costs of doing business." ( Johnston v. Long (1947) 30 Cal.2d 54, 64 [ 181 P.2d 645 ].) "Although earlier authorities sought to justify the respondeat superior doctrine on such theories as `control' by the master of the servant, the master's `privilege' in being permitted to employ another, the third party's innocence in comparison to the master's selection of the servant, or the master's `deep pocket' to pay for the loss, `the modern justification for vicarious liability is a rule of policy, a deliberate allocation of a risk. The losses caused by the torts of employees, which as a practical matter are sure to occur in the conduct of the employer's enterprise, are placed upon that enterprise itself, as a required cost of doing business.'" ( Hinman v. Westinghouse Elec. Co. (1970) 2 Cal.3d 956, 959-960 [ 88 Cal. Rptr. 188 , 471 P.2d 988 ], quoting Prosser, Law of Torts (3d ed. 1964) p. 471.) "Three reasons have been suggested for imposing liability on an enterprise for the risks incident to the enterprise: `(1) [I]t tends to provide a spur toward accident prevention; (2) it tends to provide greater assurance of compensation for accident victims[;] and (3) at the same time it tends to provide reasonable assurance that, like other costs, accident losses will be broadly and equitably distributed among the beneficiaries of the enterprises that entail them.'" ( Perez, supra, 41 Cal.3d at p. 967 , quoting 5 Harper et al., The Law of Torts (2d ed. 1986) § 26.5, p. 21, fns. omitted.)

The first of these three considerations just noted plays little role in the allocation of responsibility for the sexual misconduct of employees generally, and with respect to the unique situation of teachers, indicates that untoward consequences could flow from imposing vicarious liability on school districts. Although it is unquestionably important to encourage both the careful selection of these employees and the close monitoring of their conduct, such concerns are, we think, better addressed by holding school districts to the exercise of due care in such matters and subjecting them to liability only for their own direct negligence in that regard. Applying the doctrine of respondeat superior to impose, in effect, strict liability in this context would be far too likely to deter districts from encouraging, or even authorizing, extracurricular and/or one-on-one contacts between teachers and students or to induce districts to impose such rigorous controls on activities of this nature that the educational process would be negatively affected. [10]

Nor is the second consideration the assurance of compensation for accident victims appropriately invoked here. The acts here differ from the normal range of risks for which costs can be spread and insurance sought. (See Alma W., supra, 123 Cal. App.3d at p. 144 .) The imposition of vicarious liability on school districts for the sexual torts of their employees would tend to make insurance, already a scarce resource, even harder to obtain, and could lead to the diversion of needed funds from the classroom to cover claims.

The only element of the analysis that might point in favor of vicarious liability here is the propriety of spreading the risk of loss among the beneficiaries of the enterprise. School districts and the community at large benefit from the authority placed in teachers to carry out the educational mission, and it can be argued that the consequences of an abuse of that authority should be shared on an equally broad basis. But the connection between the authority conferred on teachers to carry out their instructional duties and the abuse of that authority to indulge in personal, sexual misconduct is simply too attenuated to deem a sexual assault as falling within the range of risks allocable to a teacher's employer. It is not a cost this particular enterprise should bear, and the consequences of imposing liability are unacceptable.

(5b) In sum, we believe the Court of Appeal erred in looking mainly to the factual similarities between this case and White, supra, 166 Cal. App.3d 566 , and in failing to consider whether the underlying justifications for the respondeat superior doctrine would be served by imposing vicarious liability here. We need not and do not decide whether White itself was properly decided or whether the job-created authority theory has any validity in evaluating vicarious liability for the torts of police officers. It suffices here to note that the authority of a police officer over a motorist bolstered most immediately by his uniform, badge and firearm, and only slightly less so by the prospect of criminal sanctions for disobedience plainly surpasses that of a teacher over a student. The teacher's authority is different in both degree and kind, and it is simply not great enough to persuade us that vicarious liability should attach here for the teacher's tort. Furthermore, invoking respondeat superior here would raise an entirely different specter of untoward consequences, or interference with the purposes for which the authority was conferred in the first place, than might result from the imposition of vicarious liability in the limited context of a police officer's abuse of authority. We doubt that police departments would deprive their officers of weapons or preclude them from enforcing the laws, but we see a significant and unacceptable risk that school districts would be dissuaded from permitting teachers to interact with their students on any but the most formal and supervised basis.

The judgment of the Court of Appeal is affirmed insofar as it reversed the trial court's grant of nonsuit based on the untimeliness of plaintiffs' late-claim application under the Tort Claims Act thus allowing plaintiffs an opportunity to show, in conformity with the views expressed in this opinion, that their application was timely under principles of equitable estoppel. The judgment of the Court of Appeal is reversed insofar as it reversed the trial court's order sustaining the district's demurrer to those claims premised on a theory of vicarious liability under the respondeat superior doctrine and the Court of Appeal is directed to enter judgment affirming the trial court's order thus leaving plaintiffs free to pursue only their claims against the district premised on its own direct negligence in hiring and supervising the teacher.

Broussard, J., concurred.

MOSK, J.

(1c), (2b) I concur in the views of the majority on the late-claim application under the Tort Claims Act. However, I dissent from their conclusion on the respondeat superior doctrine.

It is my view that the Court of Appeal was entirely correct in its analysis of plaintiffs' claim against the school district. Therefore I adopt as my own the respondeat superior discussion in the opinion by Presiding Justice Racanelli for the Court of Appeal: "Under the respondeat superior doctrine, an employer's liability extends to torts of an employee committed within the scope of his employment. [Citation.] This includes willful and malicious torts as well as negligence. [Citation.]" ( Martinez v. Hagopian (1986) 182 Cal. App.3d 1223, 1227 [ 227 Cal. Rptr. 763 ].)

"The determination as to whether an employee committed a tort during the course of his employment turns on whether `1) the act performed was either required or "incident to his duties" [citation], or 2) the employee's misconduct could be reasonably foreseen by the employer in any event [citations].' [Citation.] The employee's actions need only fall within the range of actions covered by either part of this two-pronged test for the employer to be held liable." ( Martinez, supra, 182 Cal. App.3d at p. 1228 .)

Relying on White v. County of Orange (1985) 166 Cal. App.3d 566 [ 212 Cal. Rptr. 493 ], appellants argue that the teacher's misconduct occurring within the course and scope of his employment imposes vicarious liability upon the public entity for such misconduct. In contrast, respondent relies on Alma W. v. Oakland Unified School Dist. (1981) 123 Cal. App.3d 133 [ 176 Cal. Rptr. 287 ] and Rita M. v. Roman Catholic Archbishop (1986) 187 Cal. App.3d 1453 [ 232 Cal. Rptr. 685 ], to argue that sexual molestation of students is not a part of a teacher's duties so as to fasten liability upon the employer district.

White involved an action by a woman against a deputy sheriff who kidnapped her while he was on patrol in uniform in a marked patrol car. In reversing a summary judgment in favor of the county, the court stated that the employer is responsible for acts done during the exercise of the employee's authority. It is noteworthy that the plaintiff in White claimed she would not have stopped her car had it not been for the deputy's use of his apparent authority by activating the flashing lights of his patrol car. ( White v. County of Orange, supra, 166 Cal. App.3d at pp. 571-572.)

Alma W. involved the sexual molestation of an 11-year-old child by a school custodian in the latter's office. The court affirmed a judgment on a demurrer in favor of the employer, stating that the employee's actions were neither foreseeable nor an incident of his employment. White distinguished Alma W. on the basis that the wrongful act did not flow from the exercise of the employee's duties notwithstanding that the assault took place during working hours in the custodian's office. ( White v. County of Orange, supra, 166 Cal. App.3d at p. 571 .)

Rita M. involved a complaint by a 16-year-old girl who was seduced by several of her parish priests. Citing Alma W., the court stated: "It would defy every notion of logic and fairness to say that sexual activity between a priest and a parishioner is characteristic of the Archbishop of the Roman Catholic Church." ( Rita M. v. Roman Catholic Archbishop, supra, 187 Cal. App.3d at p. 1461 .) The court concluded that sexual activity with a parishioner was not incidental to priestly duties or reasonably foreseeable as an outgrowth of these duties. However, the opinion failed to discuss the salient fact that the priests obtained the child's silence by informing her the sex acts were ethically and religiously permissible. ( Id., at pp. 1456-1457.)

[I] conclude Rita M. and Alma W. are factually dissimilar from the instant case. There, neither the custodian nor the priests had any actual authority over their victims and did not accomplish the assaults through the official exercise of their job-related duties. The distinguishing feature in White is that the assault arose out of the deputy's abuse of his official authority and was, therefore, incident to his duties. In Alma W., the assault was completely unrelated to the performance of the custodian's duties. Although Rita M. falls somewhere between White and Alma W., [I] conclude that the sexual assaults in Rita M. did not arise out of the priests' exercise of job-related authority over the plaintiff.

The case at bench is in analytical symmetry with White . Here, under the charging allegations of the third amended complaint which we must accept as true for purposes of review ( Molien v. Kaiser Foundation Hospitals (1980) 27 Cal.3d 916, 919 [ 167 Cal. Rptr. 831 , 616 P.2d 813 , 16 A.L.R.4th 518 ] it is alleged that the teacher engaged the child in an officially sanctioned instructional work experience program which involved extracurricular work at the teacher's home. It was there, while the minor was participating in the program by correcting papers, that the alleged sexual conduct took place. The teacher explained to the child that the sexual conduct was a part of his role as teacher and was designed to help the student with his problems. The complaint also alleged that the acts took place solely because of the existence of the relationship of teacher and pupil.

As in White , the teacher, by virtue of the exercise of his official authority, was able to perpetrate the sexual assault. That is, through the use of his authority to administer grades, to assign extracurricular work projects, and, significantly, by utilizing the school-approved work experience program, the teacher procured the student's presence in his home facilitating the opportunity for the assault. In concluding that such pleaded facts would be clearly incident to the teacher's exercise of his official duties, we [should] focus not on whether the school teacher's sexual activity with a student is either "characteristic" or foreseeable, but rather on whether the assault arose out of the exercise of job-created authority over the plaintiff student. It was, therefore, error to sustain the demurrer as to appellants' respondeat superior theory. [End of Court of Appeal opinion, parallel citations omitted, fn. omitted.]

I would affirm the Court of Appeal judgment in its entirety.

EAGLESON, J.

(5c) I concur in the majority's holding that the defendant school district cannot be held vicariously liable for its teacher's alleged sexual misconduct. I respectfully dissent, however, from the majority's additional holding that plaintiffs must be allowed an opportunity to establish the elements of equitable estoppel so as to avoid the claim-filing periods under Government Code sections 911.2 and 911.4, subdivision (b).

I find no merit in the majority's effort to reconcile its holding that there is no vicarious liability with its holding that the teacher's alleged threats can be imputed to the district. The two holdings are legally inconsistent. If the district is not vicariously liable for the underlying tort, the district should not be held vicariously liable for threats related to the tort.

Before discussing the inconsistency, it is helpful to explain why the majority imputes the alleged threat to the district. In deciding whether a public entity should be estopped from relying on a claims statute, we have identified four elements that must be shown for estoppel to arise: "(1) the party to be estopped must be apprised of the facts; (2) he must intend that his conduct shall be acted upon, or must so act that the party asserting the estoppel had a right to believe it was so intended; (3) the other party must be ignorant of the true state of facts; and (4) he must rely upon the conduct to his injury." ( Driscoll v. City of Los Angeles (1967) 67 Cal.2d 297, 305 [61 Cal. Rptr. 661, 431 P.2d 245 ], italics added.) The first element is absent in this case. There is no evidence that the district the party to be estopped knew of the alleged molestation or threats, and plaintiffs do not contend otherwise. This element can be shown only if the teacher's alleged threats are imputed to the district. The alleged threats, however, cannot be properly imputed to the district.

In rejecting vicarious liability for the teacher's alleged sexual misconduct, the majority correctly concludes that, "the connection between the authority conferred on teachers to carry out their instructional duties and the abuse of that authority to indulge in personal, sexual misconduct is simply too attenuated to deem a sexual assault as falling within the range of risks allocable to a teacher's employer." (Lead opn. at p. 452, ante, italics added.) If the sexual misconduct itself is too attenuated to impose liability on the district, a threat relating to that misconduct is even more attenuated. The majority's view, stated more clearly, is that the teacher's sexual misconduct was so beyond the bounds of his employment that it would be unfair to impute his misconduct to the district, but that his threats relating to the misconduct are sufficiently related to his employment and should be imputed to the district. That conclusion is oxymoronic.

The majority's implicit justification for its inconsistent holdings is that the district may be independently liable for its own misconduct. That is correct but irrelevant. A threat is conduct. By imputing the teacher's alleged threat to the district, the majority in effect is holding the district responsible for the teacher's conduct, not for the district's own conduct. The flaw in the majority's analysis is that in order to reach the issue of whether the district is responsible for its own conduct the majority holds the district responsible for the teacher's conduct. Doing so squarely conflicts with the holding of no vicarious liability.

None of the cases cited by the majority support its reasoning on this issue. Those cases stand for the proposition that a public entity may be estopped from relying on the limitations period of a claims statute when the entity's agents or employees have prevented the filing of a timely claim. In each of the cited cases, however, there was no suggestion that the agent or employee was not acting in the course and scope of his employment. To the contrary, it is clear from the facts of each case that the employee or agent was acting on behalf of the public entity. ( Fredrichsen v. City of Lakewood (1971) 6 Cal.3d 353, 355-356 [ 99 Cal. Rptr. 13 , 491 P.2d 805 ] [city clerk misrepresented city's responsibility for accident]; Rand v. Andreatta (1964) 60 Cal.2d 846, 850 [ 36 Cal. Rptr. 846 , 389 P.2d 382 ] [claims agent advised injured party not to retain counsel; parties stipulated that county driver who caused accident was acting in course and scope of her employment]; Bruce v. Jefferson Union High Sch. Dist. (1962) 210 Cal. App.2d 632, 633 [ 26 Cal. Rptr. 762 ] [school nurse misinformed student as to claims procedure]; Industrial Indem. Co. v. Ind. Acc. Com. (1953) 115 Cal. App.2d 684, 686 [ 252 P.2d 649 ] [insurer's representative misled accident victim as to status of claim]; Dettamanti v. Lompoc Union School Dist. (1956) 143 Cal. App.2d 715, 722-723 [ 300 P.2d 78 ] [deputy district attorney induced delay in filing claim]; Bertorelli v. City of Tulare (1986) 180 Cal. App.3d 432, 436-437 [ 225 Cal. Rptr. 582 ] [city's insurance adjuster allegedly caused delay]; see also Driscoll v. City of Los Angeles, supra, 67 Cal.2d 297, 303 [public entity's employees misadvised pension claimants]; Farrell v. County of Placer (1944) 23 Cal.2d 624, 627 [ 145 P.2d 570 , 153 A.L.R. 323 ] [agent of public entities misled accident victim].)

I have been unable to find a single decision in which a public entity was estopped based on conduct of an employee who was found not to be acting in the course and scope of his employment. The majority's conclusion that the teacher's threats can be imputed to the district is a radical departure from well-established law.

A party seeking to rely on estoppel must show that he was ignorant of the true state of facts. ( Driscoll v. City of Los Angeles, supra, 67 Cal.2d 297, 305 .) It is undisputed in this case that John was aware of the molestation and knew that it was wrongful when it occurred. (He was 14 years old.) For that reason alone, it is questionable whether plaintiffs can rely on estoppel based on the teacher's alleged threats.

The majority does not even refer to this problem, but attempts to avoid it, claiming that "... estoppel may certainly be invoked when there are acts of violence or intimidation that are intended to prevent the filing of a claim." (Lead opn. at p. 445, ante, italics in opinion.) This contention is somewhat of a red herring. There are no allegations in this case of violence or any form of physical intimidation. More important, the majority treats as a well-established rule what is at best an unsettled question. In DeRose v. Carswell (1987) 196 Cal. App.3d 1011 [ 242 Cal. Rptr. 368 ], the only California case cited by the majority, the Court of Appeal explained, "... the usual context for an estoppel argument is a history of negotiations between the parties that lead one side to believe that claims will be settled. Basing an estoppel on threats appears to be a novel but plausible application of the estoppel rule." ( Id. at p. 1026, italics added.) The majority characterizes as a clear rule what the DeRose court found to be an unresolved question. Moreover, the DeRose court held the plaintiff in that case could not rely on estoppel to avoid the statute of limitations. DeRose is also inapposite because the person that the plaintiff sought to estop was the alleged perpetrator of sexual misconduct, not a vicariously liable defendant.

In the only other case cited by the majority on this point, it was clear that the plaintiff, who was seeking to rely on estoppel to avoid a statute of limitations, was unaware of the statute. ( Longo v. Pittsburgh and Lake Erie Railroad Co., N.Y.C. Sys. (3d Cir.1966) 355 F.2d 443, 444 .) That case does not stand for the proposition that a threat is sufficient to give rise to estoppel even if the plaintiff was aware of the true state of facts.

Perhaps under certain circumstances there should be a rule that a threat is sufficient by itself to create an estoppel. No such rule, however, has yet been established in this state. If the majority proposes to create such a rule, the majority should duly consider the question and announce forthrightly that it is establishing a new rule. The majority should not attempt to justify its result by mischaracterizing California law as already being as the majority wishes. [1]

The majority premises its discussion of estoppel on the alleged threats by John's teacher. Even if those threats are found to estop the district, plaintiffs' claims were untimely. Under section 911.2, as it was then worded, plaintiffs had to present their claims to the district within 100 days of the accrual of the cause of action. [2] Plaintiffs admit that John told his father on December 17, 1981, of the alleged molestation. The threats ceased to cause any delay as of that date, and estoppel was no longer applicable. Plaintiffs therefore had 100 days after December 17, 1981, to file their claims. They did not do so until May 13, 1982 147 days later. Their claim was untimely under section 911.2 even if the alleged threats delayed John in telling his parents of the molestation.

Under section 911.4, plaintiffs had a reasonable time not to exceed one year after the cause of action accrued in which to apply for leave to file a late claim. In light of the majority's admitted reservations as to the Court of Appeal's finding that accrual was delayed until John informed his parents (lead opn. at p. 444, ante ), I presume the majority believes the cause of action accrued on the date of molestation. If so, plaintiffs' application under section 911.4 was also untimely. The molestation occurred no later than February 1981, but the application was not filed until May 1982 more than one year later.

The alleged threat is relevant only if the majority's unstated premise is that estoppel based on the threat delayed accrual of plaintiffs' causes of action until John informed them of the molestation. If that is the premise, the majority should explain why it has reservations as to the Court of Appeal's finding of delayed discovery because, as applied by the majority, estoppel is delayed discovery under a different label. It is therefore inconsistent and misleading for the majority to purport to disapprove of delayed discovery while at the same time molding the doctrine of estoppel to delay accrual until discovery. "Rose is a rose is a rose is a rose." (Gertrude Stein, Sacred Emily (1913).)

Because the majority uses estoppel to delay accrual of plaintiffs' causes of action, I believe it necessary to explain briefly some of the reasons why delayed discovery is inappropriate.

1. A delayed-discovery rule is unnecessary in most cases to protect a minor's interest. The delayed-discovery issue arises in this case only because the minor is seeking recovery from a governmental entity under the Tort Claims Act. In cases in which the defendants are not protected by the Tort Claims Act, the minor has the benefit of Code of Civil Procedure section 352, which provides that the period of limitation on a minor's claim is tolled during his minority. If the period of limitation, as in this case, is one year, the minor has until the age of nineteen to file suit.

2. Even in the narrow context of molestation claims against governmental entities, the delayed-discovery rule is largely unnecessary to protect the minor's interest. A majority of the court agrees there should be no vicarious liability imposed on the district because the teacher was not acting in the course and scope of his employment. The logical consequence of that conclusion is that the teacher is not subject to the protection of the Tort Claims Act. As to him, the general statute of limitation applies, and the minor had until one year after reaching the age of majority to file suit.

Moreover, under the Tort Claims Act, a claimant has a reasonable time, not to exceed one year after accrual of his cause of action, to file an application to present a late claim. Thus, even if we define the date of accrual as the date of injury, a minor has a year thereafter to file a claim if he acts reasonably. (A minor is almost always found to have acted reasonably in filing a claim.) This seems to be a sufficient period of time. Indeed, in this case John's parents learned of the alleged molestation well within the one-year period.

3. A key purpose of the prompt claim-filing requirement is to allow a public entity the opportunity to investigate and to respond. A delayed-discovery rule in molestation cases is fundamentally inconsistent with that purpose. Under the majority's estoppel theory, a claim could be filed 10 or 15 years after an alleged molestation. It will be much more difficult to investigate at that late date than if the claim is filed within the period specified in the Tort Claims Act.

The Court of Appeal relied on the delayed-discovery rule as applied in malpractice cases against public hospitals. As recently explained by the Washington Supreme Court in Tyson v. Tyson (1986) 107 Wn.2d 72 [ 727 P.2d 226 ], there are sound reasons why the rule in malpractice cases should not be extended to sexual molestation cases. Unlike in medical malpractice cases, the evidence of injury and damages in a molestation case is often not objectively verifiable. For example, if a surgeon leaves a sponge in a patient's abdomen and the sponge is not discovered for 15 years, the defendant is not much prejudiced by the passage of time because the injury (leaving the sponge) and the damages (the current illness) are objectively verifiable. This may not be true in every malpractice case, but it is likely the general situation. There will also be medical records and often witnesses, e.g., nurses and other physicians.

In almost all molestation cases, however, there will be no physical evidence of molestation, especially after considerable time has passed; there will be no documentary evidence; and there will be only two witnesses the child and the defendant. It will be very difficult for a public entity to investigate the allegations against it, much more so than in the typical medical malpractice case. Thus, the delayed-discovery rule in molestation cases is inconsistent with the purpose of the prompt claim-filing requirement.

I also note the dynamics of a belated molestation case. A defendant charged with a molestation that allegedly occurred five or ten years ago is at a terrible disadvantage. The plaintiff, perhaps still a child, understandably has the jury's sympathy. The defendant has only his denial. Even worse, the child often "remembers" the alleged molestation several years later only as a result of counseling in which a therapist "discovers" what happened.

There is also an evidentiary problem that may prejudice a defendant. As evidenced by John's conflicting testimony in the criminal proceeding against the teacher, minors often do not remember when the alleged molestation occurred. This makes it very difficult for the innocent defendant to establish an alibi defense. A defendant charged with other crimes, e.g., murder or robbery, can attempt to show that he was elsewhere when the crime was committed. Consider, however, the trial of an action for molestation many years after the event. The child will likely not remember with any specificity when the alleged molestation occurred. His recollection may be something like "a few years ago or the summer two or three years ago." An innocent defendant will be hard pressed to assert an alibi after passage of such a long and vague period of time.

I note the foregoing factors not out of sympathy or tolerance for child molesters. Sexual abuse of a child is tragic. Society's justifiable repugnance toward such misbehavior, however, is the reason why a falsely accused defendant can be gravely harmed. The potential for prejudice is exacerbated where the defendant is not the molester himself and knew nothing about the molestation until after it occurred, for example, a school district. I doubt that a school district can adequately investigate and defend a claim 5, 10, or 15 years after an alleged molestation.

4. The only reason for applying the delayed-discovery rule in this case is to provide the minor with a deep financial pocket for recovery. I do not believe such goal, though beneficent, is a sufficient reason to undercut the policies that require prompt claim filing against governmental entities. The court's proper function is not to search for deep pockets. Moreover, even if one concludes that in a perfectly fair world the minor should have the same length of time to file a claim against the district as against a private defendant, that determination should be made by the Legislature, not this court. It is unquestioned that the Legislature had the authority to provide less time for claims against the government than against private defendants. Every judicial decision that operates to lengthen the claim period is contrary to the underlying legislative goal of requiring prompt claims.

On a more general level, I believe statutes of limitation, including the Tort Claims Act, serve a valuable purpose for society. It seems increasingly fashionable to view the statutes as obstacles that must always be overcome to allow compensation to an injured party. That the statutes were enacted, however, indicates a legislative decision that, despite tortious injury, there are compelling reasons to deny recovery after the passage of time. All statutes of limitation are arguably unfair when they operate to preclude recovery for wrongfully caused injuries, but society has decided this unfairness is outweighed by the unfairness that results from the assertion of stale claims.

5. I believe there should be some symmetry between the civil and criminal statutes of limitation. For sex crimes and most others, the period of limitation begins to run at the commission of the offense rather than when the offense is discovered. From a societal perspective, I believe the state's interest in prosecuting a child molester is equal to the interest of a molestation victim in receiving compensation in a civil action. The defendant's perspective is also similar. He will be gravely harmed by false charges if they are brought in a penal or civil action. If the Legislature has determined that for penal purposes the statute begins to run at the time of the offense, it seems reasonable that the civil statute should operate likewise, especially when, as in this case, the real purpose of applying the delayed-discovery rule is to allow recovery from a party other than the molester.

In summary, the equitable estoppel rule approved by the majority benefits plaintiffs only if it operates to delay the accrual of their causes of action until the date John's parents learned of the alleged molestation. There are numerous reasons, however, why the delayed-discovery rule should not apply, and the majority itself expresses reservations as to the rule's application in this case. Those same reservations should cause the majority to refrain from relying on equitable estoppel to preserve plaintiffs' claims.

I would reverse the Court of Appeal's decision in its entirety with directions to remand this action to the trial court for entry of judgment in favor of defendant school district.

Lucas, C.J., and Panelli, J., concurred.

KAUFMAN, J.

(1d) I concur in the majority's holding that the district may be estopped from asserting the statute of limitations as a bar to the action. I dissent from its additional holding that the district may not, as a matter of law, be held vicariously liable for the intentional torts of its employee in this case. I would agree it is the rare case in which a sexual assault by a teacher against a student should give rise to liability on the part of his employing school district. Nevertheless, as I shall explain below, this is such a case.

To recover under respondeat superior, plaintiff bears the burden of proof to demonstrate that the employee's tortious acts were committed within the scope of his employment. ( Perez v. Van Groningen & Sons, Inc. (1986) 41 Cal.3d 962, 968 [ 227 Cal. Rptr. 106 , 719 P.2d 676 ].) Generally, the issue of scope of employment is a question of fact. ( Ducey v. Argo Sales Co. (1979) 25 Cal.3d 707, 722 [ 159 Cal. Rptr. 835 , 602 P.2d 755 ].) The issue only becomes a matter of law when the facts are undisputed and no conflicting inferences are possible. ( Hinman v. Westinghouse Elec. Co. (1970) 2 Cal.3d 956, 963 [ 88 Cal. Rptr. 188 , 471 P.2d 988 ].)

The determination as to whether an employee committed a tort in the course of his employment turns upon whether or not the act was required by or incident to his duties, or could reasonably be foreseen by the employer in any event. ( Martinez v. Hagopian (1986) 182 Cal. App.3d 1223, 1228 [ 227 Cal. Rptr. 763 ]; Avila v. Standard Oil Co. (1985) 167 Cal. App.3d 441, 447 [ 213 Cal. Rptr. 314 ]; Clark Equipment Co. v. Wheat (1979) 92 Cal. App.3d 503, 520 [ 154 Cal. Rptr. 874 ].)

"Foreseeability" in this context must be distinguished from foreseeability as a test for negligence. "In the latter sense `foreseeable' means a level of probability which would lead a prudent person to take effective precautions whereas `foreseeability' as a test for respondeat superior merely means that in the context of the particular enterprise an employee's conduct is not so unusual or startling that it would seem unfair to include the loss resulting from it among other costs of the employer's business." ( Rodgers v. Kemper Constr. Co. (1975) 50 Cal. App.3d 608, 618-619 [ 124 Cal. Rptr. 143 ], quoted with approval in Perez v. Van Groningen & Sons, Inc., supra, 41 Cal.3d at p. 968 , italics added.) "Accordingly, the employer's liability extends beyond his actual or possible control of the employee to include risks inherent in or created by the enterprise." ( Perez v. Van Groningen & Sons, Inc., supra, 41 Cal.3d at p. 968 .)

In keeping with these principles, courts have long held that an employer may be vicariously liable for the intentional as well as the negligent torts of its employees. (See, e.g., Perez v. Van Groningen & Sons, Inc., supra, 41 Cal.3d at p. 969 ; Carr v. Wm. C. Crowell Co. (1946) 28 Cal.2d 652, 654-656 [ 171 P.2d 5 ].) As this court explained in Perez v. Van Groningen & Sons, Inc., supra, 41 Cal.3d at page 969 : "There is no requirement that an employee's act benefit an employer for respondeat superior to apply. In fact, an employer can be liable for his employee's unauthorized intentional torts committed within the scope of employment despite lack of benefit to the employer." (Original italics.) If an assault is motivated by personal malice and is totally unrelated to the employment, the act is outside the scope of employment and the employer is not vicariously liable; otherwise, liability may be found if the assault results from circumstances generally "arising out of the employment" ( Carr v. Wm. C. Crowell Co., supra, 28 Cal.2d at p. 654) or as an "outgrowth of the employment." ( Martinez v. Hagopian, supra, 182 Cal. App.3d at p. 1229 ; see also Rodgers v. Kemper Const. Co, supra, 50 Cal. App.3d at pp. 618-621.)

A sexual assault by a teacher against a student, indeed a sexual assault by anyone under any circumstances, strikes the normal sensibility as so extreme and abhorrent that it would indeed be "unfair to include the loss resulting from it among other costs of the employer's business." ( Rodgers v. Kemper Constr. Co., supra, 50 Cal. App.3d at pp. 618-619.) As noted earlier, the courts of this state have long acknowledged that intentional torts, including assault and battery ( Ruppe v. Los Angeles (1921) 186 Cal. 400 [ 199 P. 496 ]), harassment and coercion ( Ramos v. County of Madera (1971) 4 Cal.3d 685 [ 94 Cal. Rptr. 421 , 484 P.2d 93 ]) and false imprisonment ( Allison v. County of Ventura (1977) 68 Cal. App.3d 689 [ 137 Cal. Rptr. 542 ]), are unfortunate but not altogether unexpected occurrences. Yet there remains an understandable reluctance to recognize that sexual assaults may, in certain situations, also reasonably be "expected."

Sadly, however, we have learned that sexual harassment and assaults in the home as well as the workplace are not uncommon occurrences. This is a hard truth to accept. But putting our collective heads in the sand will not make it go away. And clinging to a less "pessimistic" view of human nature (lead opn., p. 450, fn. 9) will not compensate the victims of such outrages. On the contrary, indulging such illusions merely deepens and perpetuates the injustice.

Thus, the courts of this and other jurisdictions have recognized that there are circumstances where an employee's sexual misconduct cannot, in all candor, be deemed so "unusual or startling" that it would be unfair to impose vicarious liability upon the employer. (See, e.g., White v. County of Orange (1985) 166 Cal. App.3d 566, 571-572 [ 212 Cal. Rptr. 493 ] [liability of county for deputy sheriff's threats to rape motorist]; Simmons v. United States (9th Cir.1986) 805 F.2d 1363, 1369 [liability of federal agency for mental health counselor's sexual involvement with client]; Turner v. State (La. App. 1986) 494 So.2d 1292, 1295-1296 [liability of state for National Guard recruiter's sexual misconduct with applicants]; Marston v. Minneapolis Clinic of Psychiatry (Minn. 1982) 329 N.W.2d 306, 310-311 [liability of clinic for therapists' sexual relations with patient]; Lyon v. Carey (D.C. Cir.1976) 533 F.2d 649, 651 [174 App.D.C 422] [liability of employer for deliveryman's rape of customer].)

The case before us falls within this narrow category. The sexual assault by the teacher against the minor occurred in the teacher's home, in the course of an official educational program known as Instructional Work Experience (IWE). The district affirmatively sanctioned IWE work at teachers' homes as an acceptable feature of the program. The assault was further facilitated by the teacher's representations to the minor that a sexual relationship would enhance his educational experience in the IWE program.

The district did not require that a student obtain the written permission of his parents to participate in the IWE program at the teacher's home, nor did it require that other students or adults be present during the home instruction. In effect, the district-sanctioned IWE program virtually guaranteed that the teacher could act with impunity, free from the fear of interruption or discovery, fully assured of complete privacy and secrecy. Under these circumstances, it is not so "unusual or startling" that a teacher might seize the opportunity created by the program to make improper sexual advances toward one of his students. Under these circumstances, it is unjust not to require that the district share in the liability for the injuries which resulted.

Rather than address these compelling facts head on, however, the majority cites certain "policy" reasons which appear, in its view, to trump both law and equity. [1] It is suggested, for example, that the imposition of vicarious liability would "deter" school districts from sponsoring extracurricular school activities. I do not believe that such would result from a decision adverse to the district in this case. Respondeat superior is a fact-specific determination; a holding adverse to the district would necessarily be limited to the uniquely compelling facts of this case. Indeed, as noted earlier, vicarious liability for sexual assaults should be recognized as the exception, not the rule.

The IWE home-instruction program contained none of the usual safeguards incident to most normal extracurricular activities, i.e., a public setting (as opposed to the private seclusion of the teacher's home) and the presence or or knowledge of other persons (in contrast to the isolation and secrecy inherent in the IWE program). Such reasonable safeguards would normally act to deter such misconduct, or, failing that, to limit the district's exposure to claims based on vicarious liability.

Indeed, "public policy" militates strongly in favor of vicarious liability in this case. One of the "policy" bases of respondeat superior is said to be its tendency to act as a "spur toward accident prevention." ( Perez v. Van Groningen & Sons, Inc., supra, 41 Cal.3d at p. 967 .) If that is the case, and incidents of the nature at issue here are, as I believe, a foreseeable result of such ill-advised programs, then the imposition of vicarious liability in this case might ultimately prove to be an inducement, not a deterrent, to well planned and properly executed extracurricular school programs.

It is further suggested that the underlying purposes of respondeat superior the assurance of fair compensation for tort victims by spreading the risk of losses through insurance carried by the responsible enterprise as a cost of doing business ( Hinman v. Westinghouse Elec. Co., supra, 2 Cal.3d at pp. 959-960) would not be furthered in this case. [2] On the contrary, these considerations amply justify the imposition of vicarious liability. The sexual assault against the minor was the direct result of the home-instruction program conceived and sponsored by the district. The district's employee persuaded the minor that sexual relations would legitimately contribute to his educational experience in the IWE program, and capitalized on the secrecy and seclusion created by that program to perpetrate the assault without fear of interruption or discovery. There is a clear and direct nexus between the teacher's misconduct and the risks inherent in the district-sanctioned program. It is fair and just and entirely consistent with the underlying purposes of the doctrine of respondeat superior that the district be required to bear the cost of the resulting losses.

For these reasons, I would allow plaintiffs to proceed against the district on its claims based on vicarious liability. I would affirm the Court of Appeal's decision to reverse the trial court's order sustaining the district's demurrer to these claims.

Respondent's petition for a rehearing was denied June 14, 1989.

[*] Retired Associate Justice of the Supreme Court sitting under assignment by the Chairperson of the Judicial Council.

[1] As this case comes before us following an appeal from a judgment of nonsuit at the outset of trial and a prior order sustaining a demurrer to the third amended complaint, we accept as true the facts alleged by plaintiffs. ( Sullivan v. County of Los Angeles (1974) 12 Cal.3d 710, 714, fn. 3 [ 117 Cal. Rptr. 241 , 527 P.2d 865 ]; Smith v. Roach (1975) 53 Cal. App.3d 893, 897 [ 126 Cal. Rptr. 29 ].)

[2] Criminal charges were filed against the teacher on the theory, based on John's statements at the time, that the molestation occurred in May of 1981. John so testified at the teacher's trial in August of 1982. After both sides had rested in that trial, the court dismissed the charges. The parties before us agreed in the Court of Appeal that the case was dismissed because the evidence showed, and the judge stated, that the incident could not have occurred any later than February of 1981.

[3] A cause of action accrues for purposes of the filing requirements of the Tort Claims Act on the same date a similar action against a nonpublic entity would be deemed to accrue for purposes of applying the relevant statute of limitations. (Gov. Code, § 901.) For minors, however, the time of accrual is generally more significant in the context of the claims statute. With certain exceptions, the statute of limitations does not run during the time a potential plaintiff is a minor, and such a party accordingly has up to a year after attaining the age of majority to bring suit on a cause of action for personal injury. (See Code Civ. Proc., § 352, subd. (a).) This respite does not apply to a claim against a public entity under the Tort Claims Act, however (Code Civ. Proc., § 352, subd. (b)), and Government Code section 911.4, subdivision (b) expressly provides that the "time during which the person who sustained the alleged injury, damage, or loss is a minor shall be counted" in determining whether a late-claim application was timely filed within one year of the accrual of the cause of action.

[4] This is not, as Justice Eagleson's concurring and dissenting opinion would have it (conc. and dis. opn., post, p. 459), delayed discovery in different guise. Quite simply, equitable estoppel may prove to preclude the district from asserting that plaintiffs' claims were not timely made; it cannot, as might delayed accrual of the causes of action, lead to the conclusion that the claims were in fact timely. As a result, equitable estoppel imposes requirements on a claimant that a delayed-discovery theory of accrual would not. (See, e.g., Regus v. Schartkoff (1957) 156 Cal. App.2d 382, 387 [ 319 P.2d 721 ].) That the two doctrines may produce similar results in some circumstances does not make them equivalents, and it is misleading to assert that the majority does so.

[5] The concurring and dissenting opinion of Justice Eagleson apparently thinks it contradictory for us to conclude that the teacher's threats may be the basis of an estoppel against the district although we conclude the district cannot be held vicariously liable for the teacher's underlying conduct. (See conc. and dis. opn., post, pp. 455-457.) Not so. There is no inconsistency in distinguishing the acts supporting a claim of estoppel here (the teacher's threats) from those on which plaintiffs premise their causes of action (the sexual molestation), for estoppel rests on the conduct which precluded the filing of a claim, conduct separate and apart from that which caused the injury on which the claim is based. First, such threats might be made by an employee in an effort to conceal any number of underlying torts, some of which would be within the scope of employment and others not. Whether an employer would be vicariously liable on the underlying claim has no bearing on whether the threats would justify an estoppel, allowing that claim to be brought and the question of vicarious liability resolved on the merits. Second, even if, as here, the conduct which caused the injury falls outside the scope of employment and fails to provide a basis for imposing vicarious liability, the employer may be independently liable for its own conduct. And if the employee succeeds by threats or duress in coercing the victim not to file a claim within the relevant period, the employer may be estopped from relying on that separate and independent misconduct to absolve itself of any responsibility for its own acts.

[6] In determining when the estoppel ceased to operate here, the trial court must take into account the nature of the conduct on which the estoppel is based conduct that here lacks a clear ending point, particularly in light of the allegations that the teacher's threats against John continued even after the boy had told his parents of the molestation and the incidents had been reported to the district and the police. The finding on that point will not only affect the inquiry into whether plaintiffs acted within a reasonable time after the estoppel expired, but will also dictate whether plaintiffs must show the delay in filing their claims was excused by mistake or excusable neglect (see Gov. Code, § 946.6, subd. (c)) and whether the district should have the opportunity to demonstrate prejudice in defending against the claims as a result of the delay. (See Viles v. State of California (1967) 66 Cal.2d 24, 32 [ 56 Cal. Rptr. 666 , 423 P.2d 818 ].)

[7] See Jeffrey E. v. Central Baptist Church (1988) 197 Cal. App.3d 718, 722-724 [ 243 Cal. Rptr. 128 ] (church not liable for sexual abuse of minor by Sunday school teacher); Rita M. v. Roman Catholic Archbishop (1986) 187 Cal. App.3d 1453, 1461 [ 232 Cal. Rptr. 685 ] (archbishop not liable for sexual relations between seven priests and minor parishioner); Alma W., supra, 123 Cal. App.3d at pages 143-144 (school district not liable for rape of student by janitor); Boykin v. District of Columbia (D.C.App. 1984) 484 A.2d 560, 562-563 (school district not liable for sexual assault on student by teacher); Bozarth v. Harper Creek Bd. of Ed. (1979) 94 Mich. App. 351 [ 288 N.W.2d 424, 425 ] (same); Gambling v. Cornish (N.D.Ill. 1977) 426 F. Supp. 1153, 1155 (municipality not liable for abduction and rape by police officers).

[8] See Richard H. v. Larry D. (1988) 198 Cal. App.3d 591, 596 [ 243 Cal. Rptr. 807 ] (liability of clinic where psychotherapist consulted by married couple had sexual relations with the wife); White v. County of Orange (1985) 166 Cal. App.3d 566, 571-572 [ 212 Cal. Rptr. 493 ] (liability of county for deputy sheriff's threats to rape motorist); Simmons v. United States (9th Cir.1986) 805 F.2d 1363, 1368-1371 (liability of federal agency for mental health counselor's sexual involvement with client); Turner v. State (La. App. 1986) 494 So.2d 1292, 1295-1296 (liability of state for National Guard recruiter's sexual misconduct with applicants); Marston v. Minneapolis Clinic of Psychiatry (Minn. 1982) 329 N.W.2d 306, 310-311 (liability of clinic for therapist's sexual relations with patient); Applewhite v. City of Baton Rouge (La. App. 1979) 380 So.2d 119, 121-122 (liability of city for policeman's rape of detainee); Lyon v. Carey (D.C. Cir.1976) 533 F.2d 649, 651 [ 174 App.D.C. 422 ] (liability of employer for deliveryman's rape of customer).

[9] We thus cannot agree with the view espoused by Justice Kaufman in his concurring and dissenting opinion, that the facts of this case bring it within some exception to what he concedes should ordinarily be a rule against the imposition of vicarious liability on an employer for a sexual assault committed by an employee. Although Justice Kaufman acknowledges the distinction between foreseeability as an element of respondeat superior and foreseeability as a test for negligence, we think his analysis in fact confuses the two. The factors he points to in urging we recognize vicarious liability here are matters relevant to whether the district itself acted negligently, not to whether it should be vicariously liable for its employee's acts regardless of its own fault. Justice Kaufman would have us prejudge plaintiffs' direct claims against the district for its own negligence and would deny the district any opportunity to show that it acted reasonably. We cannot accept such an approach.

We must also observe that the opinion seems to reflect an unduly pessimistic view of human nature, for, if we read the opinion correctly, it seems to suggest that sexual misconduct is foreseeable any time a minor and an adult are alone in a room together, at least if not constrained by the possibility of being interrupted. (See conc. and dis. opn., post, p. 465.) Given the facts of this case and the benefit of hindsight, all would have to agree that the prospect of such misconduct is conceivable, but that is a far cry from foreseeability, even under the broad meaning that concept is given in the respondeat superior context. Simply stated, for the reasons we discuss ( post, pp. 450-452), we think the teacher's acts here can only be characterized as "so unusual or startling" ( Rodgers v. Kemper Constr. Co. (1975) 50 Cal. App.3d 608, 619 [ 124 Cal. Rptr. 143 ]) that vicarious liability cannot fairly be imposed on the district.

[10] Thus, to the extent the extracurricular program involved in this case may have lacked certain prudent safeguards, such as requiring the written permission of a child's parents before the child could participate or requiring the presence of other children or adults if the activities were to occur in a private location away from the school, we think these factors relevant mainly to plaintiffs' claims against the district based on its own alleged negligence and not to the policy question (see Hinman v. Westinghouse Elec. Co., supra, 2 Cal.3d at p. 959 ) of the district's vicarious liability for the acts of its employee. (See also ante, p. 448, fn. 8.) And, Justice Kaufman's perplexing comments to the contrary notwithstanding (conc. and dis. opn., post, p. 465, fn. 1), plaintiffs' direct claims against the district are before us. (Cal. Rules of Court, rule 29(a); see ante, p. 443; post, p. 452.)

[1] The majority is potentially misleading on one other point. In a footnote, the majority indicates that the trial court must determine when the alleged threats by the teacher ceased. (Lead opn. at p. 446 fn. 6, ante. ) Apparently, the majority means to suggest that the estoppel continued even after John informed his parents of the alleged molestation. I disagree. The majority provides no guidance as to when the estoppel period ceased. Such a rule creates a potentially infinite claims period. Moreover, the majority also suggests that plaintiffs' claims could be found to be timely even if they were filed long after the expiration of the claim-filing period. The majority's implicit reasoning is that the claim period under the Tort Claims Act does not begin to run until the estoppel ceases. In other words, if the circumstances giving rise to the estoppel continued for 10 years, the claim period would not even begin to run until after the 10-year period of estoppel. According to the majority's reasoning, if the claimant then filed his claim within this delayed claim period, his claim would be timely, and he would not have to show any reason why his claim was not filed within the statutory period, and the public entity would not be allowed to show prejudice from the lengthy delay. This result is clearly contrary to Government Code section 946.6, subdivision (c)(1), which provides that a court shall grant relief from the claim-filing requirement if "The failure to present the claim was through mistake, inadvertence, surprise or excusable neglect unless the public entity establishes that it would be prejudiced in the defense of the claim...."

[2] All statutory references are to the Government Code unless otherwise indicated.

[1] The only evidence that the majority has considered these compelling facts is a footnote observation that the absence of "prudent safeguards" in the IWE program might be "relevant" to plaintiffs' direct claim against the district for negligent supervision. (Lead opn. at p. 451, fn. 10.) Unless the majority is implying that facts may be applied to no more than one legal theory of recovery in any given case (a patently nonsensical proposition), I am at a loss to understand the pertinence of this observation. Plaintiffs' direct claims against the district are simply not before us.

[2] As to the availability of insurance to cover such claims, it is well settled that Insurance Code section 533 (insurance for willful acts void as against public policy) does not preclude coverage for liability based upon respondeat superior for the intentional torts of an employee. ( Arenson v. Nat. Automobile & Cas. Ins. Co. (1955) 45 Cal.2d 81, 84 [ 286 P.2d 816 ]; Fireman's Fund Ins. Co. v. City of Turlock (1985) 170 Cal. App.3d 988, 1000-1001 [ 216 Cal. Rptr. 796 ].) In the rare case where vicarious liability is appropriate, the distribution of costs is clearly consistent with the policy underlying the doctrine of respondeat superior. Should the minor or his parents have to bear the expense of psychiatric or psychological treatment in addition to the trauma resulting from the sexual assault?