[*PG1167]CONSTITUTIONAL LIMITS ON THE LIABILITY OF CHURCHES FOR NEGLIGENT SUPERVISION AND BREACH OF FIDUCIARY DUTY
Abstract: In the many suits against churches during the past several years for alleged misbehavior of clergy, a wide variety of tort theories have been put forward as possible bases for recovery. Among these are breach of fiduciary duty owed to church members, negligence in hiring, supervision and retention of clergy, intentional or negligent infliction of mental distress and vicarious liability for torts committed by individual clergy. This Article explores possible federal constitutional barriers to these tort actions, focusing mostly on the torts of negligent supervision and breach of fiduciary duty.
In the many suits against churches during the past several years for alleged misbehavior of clergy, a wide variety of tort theories have been put forward as possible bases for recovery. Among these are breach of fiduciary duty owed to church members, negligence in hiring, supervision and retention of clergy, intentional or negligent infliction of mental distress and vicarious liability for torts committed by individual clergy.1 In the present Article, I explore possible federal constitutional barriers to these tort actions, focusing mostly on the torts of negligent supervision and breach of fiduciary duty.
The first principle that must be accepted as established is that the outcome of a tort suit against a church may not be based upon a gov[*PG1168]ernmental answer—whether given by a legislature, administrative official, court or jury—to a religious question. This principle was established in United States v. Ballard over fifty years ago, and finds its basis in both the Free Exercise Clause and the Establishment Clause of the First Amendment.2 For an agency of the state to make a finding of fact on a religious question would tend to suppress other answers to that question and impair freedom in matters of religion. Of course, determining what constitutes a religious question within this prohibition is not a matter free from doubt. Is a religious question a question whose answer depends upon faith rather than reason? Is a religious question one that pertains to the existence of a spiritual, transcendent or absolute reality?3 In the Ballard case, a prosecution for mail fraud, some of the defendants’ representations clearly were about spiritual matters, but others hovered about the border between the celestial and the terrestrial.4 For present purposes, we need not enter into this problem, nor need we attempt to answer the question: if government may not base its activities upon assertions about spiritual realities, upon what basis may they rest?
The constitutional prohibition against answering religious questions does not end with forbidding findings of fact that require faith or that pertain to the spiritual or transcendent. It also forbids findings regarding temporal realities that make reference to spiritual realities. The U.S. Supreme Court has held that the Constitution prohibits the government from determining what someone thought or intended in regard to a religious matter. Thus the government is forbidden from determining that the settlor of a trust or a party to a contract had a certain idea or intention in regard to a religious matter, even though the fact that he had such an idea or intention is a fact of this world.5 This prohibition was set forth clearly by the Supreme Court in Presbyterian Church v. Mary Elizabeth Blue Hull Memorial Church and in Jones v. Wolf.6 Both cases involved disputes within churches about doctrine and polity, which ultimately affected rights to church property.7 The Court held that in deciding who owned and had the right to control [*PG1169]the property, courts could not determine the nature of a religious idea—such as what was meant by Presbyterianism.8 The reason for this prohibition is not the inability of a court to answer such a question, but the fear that if it does give an answer, no matter how careful it may be in assessing all the evidence, its answer is likely to be influenced or thought to be influenced by its own views about a particular religion or religion in general, or by secular considerations.9 One can appreciate that if a court interprets a very general provision, such as “they shall enjoy the property so long as they are faithful to the teachings of Vatican II,” these fears would not be groundless. At the same time, if a court is asked to determine the meaning and purpose of a nonreligious organization—the Odd Fellows for instance—there is no constitutional bar to the court doing its best to answer the question on the basis of available evidence. At least there is no restraint coming from the religion clauses of the First Amendment. It is only when the question posed pertains to religion, so that the answer may be influenced by religious ideas or religious ideas may be overridden by secular ideas, that the First Amendment forbids courts from answering. Even though a court’s answer may not be an expression of faith, if it would touch upon a matter that is the object of faith for the parties and others, the Supreme Court has held that the First Amendment requires courts to steer clear of the question.10
A consequence of the Blue Hull-Wolf prohibition is that when a case presents a religious question, there must be a special rule for disposing of the case. The religious question must be set aside and the dispute somehow decided upon another ground. The irony of the situation is that although the Blue Hull-Wolf prohibition protects religious freedom in the sense of excluding government ideas about what is desirable, it creates a special difficulty for the implementation of private plans relating to religion by depriving them of a certain kind of government assistance in their fulfillment. In the case of the Odd Fellows, by contrast, courts plunge ahead and do their best to decide, for instance, which of two warring factions are the true Odd Fellows as determined by the intentions of the founders. This consequence must always leave a lingering doubt as to the correctness of the Blue Hull-Wolf prohibition.
[*PG1170] As applied in the context of tort actions against churches, the Blue Hull-Wolf prohibition means that liability may not be based upon a court’s interpretation of the church’s own law, customs or traditions. If it were the case—which it very rarely would be—that there is no doubt that there was a violation of church law—perhaps a church official with undoubted and final authority has so declared it—then the Blue Hull-Wolf principle might not apply. But another question would be presented: assuming that a nonreligious organization may be held liable in tort for a violation of its own rules, itself a matter of doubt, would the religion clauses of the Constitution permit the same approach to be taken towards a religious organization?
There are more subtle inhibitions on church liability in tort flowing from the prohibition against a court answering religious questions. For instance, in the case of a breach of a fiduciary duty claimed to be owed to a church member—the breach perhaps consisting in a failure to supervise clergy or to respond in a particular way to a church member’s complaint of abuse—the existence of such a relationship between the church and the member that would give rise to a duty to act or to act in a particular way may not be based upon an interpretation of the church’s laws, structures or traditions. Even to describe the parties as church and member is to invoke a relationship grounded in church law, and so ultimately to determine the content of a religious idea. Likewise, a duty not to be negligent in supervision, even if the standard of care is entirely a creation of secular law and not based upon church law, may not be imposed on the basis of a finding of a relationship that is a creation of church law, such as the relationship between bishop and priest or between bishop and parishioner.11 To base liability on the existence of such relationships would violate the Blue Hull-Wolf prohibition. The dangers the Supreme Court thought it saw in allowing courts to interpret religious terms in [*PG1171]wills, deeds of trust, church constitutions or contracts also would be present if courts made determinations about relationships grounded in religious beliefs in deciding whether under tort law there was a duty to act or to act in a certain way.
By contrast, it may be possible, although difficult, to base tort liability upon facts identified as existing independently of any norms or relationships created by church law. For example, it may be possible to formulate a rule of civil liability that anyone who directs another to engage in an activity, if he knows or ought to know that the other will act on the direction, must supervise the other person so as to insure that in carrying out the direction, the other person does not inflict harm. The duty thus described may avoid reference to religious norms or relationships and so avoid violating the Blue Hull-Wolf prohibition.12 This would mean, however, only that one had avoided that particular constitutional prohibition, not that all the requirements of the religion clauses had been met. In addition, of course, the law-maker, whether legislature or court, may be unwilling to recognize such a general rule. If the duty is limited to religious organizations, there would in all likelihood be a violation of the Free Exercise Clause.13
Another situation in which the prohibition against answering religious questions may prevent imposing liability on churches involves respondeat superior. The difficulty is already intimated if the church is characterized as a principal, master or employer, and the cleric as an agent, servant or employee. The theory of respondeat superior is that if a certain relationship exists between one person and another, or between an organization and an individual, fairness requires that [*PG1172]liability be imposed upon the first person or the organization for the wrongful acts of the other person. But the prohibition against answering religious questions forbids looking to church law or structure to see if such a relationship exists—whether the bishop retained certain powers, what duties were imposed upon the minister, what the scope of the minister’s assignment was and so forth. Such issues of authority, power, duty and discretion, which are examined routinely in nonreligious cases, may not be inquired into when religious organizations are involved.14 Here again, it may be possible, although difficult, to identify certain facts that exist independently of any consideration of church law—such as that A had given B directions and B acted upon them, the notions of giving directions and acting upon them being entirely secular ideas—as a basis for vicarious liability. But if there is unwillingness to adopt a rule of such scope, then the question would be presented whether it is constitutional to have a rule of respondeat superior specially tailored to religious organizations.
Let us suppose that in a case involving a charge of negligent supervision there is no violation of the Blue Hull-Wolf prohibition. Instead, the rule to be applied is something like the one suggested: whenever A gives directions to B and B is likely to act upon them, A has a duty to watch over B so as to protect others from harm from B’s activities. In an ordinary suit for negligent supervision, account would be taken of the importance of the benefit to be gained by maintaining a certain relationship between A and B: too close a supervision of B might thwart the purpose of the enterprise. A court or a jury would consider the benefit sought to be achieved and weigh it against the seriousness of the harm that could occur and the risk that that harm will occur if there is little supervision. In the case of a tort of negligent supervision by a religious organization, what consideration would be given to the religious benefits that the organization sought to achieve by structuring the relationship between A and B in the way it did?15 [*PG1173]There are two possibilities. One is to attach no value to the benefit, precisely because it is a benefit only from a religious point of view. With no value attached to the benefit, the creation of any risk of the harm—for example, sexual molestation of minors—would not be justified. The very setting on foot of the religious enterprise would be tortious. The other possibility would be to attach a fixed value to the religious benefit sought to be achieved. This value would be determined by the benefit’s religious character as such, rather than by its importance in the scheme of a particular religion, so as to avoid violating the Blue Hull-Wolf prohibition. It would be prohibited, for instance, to determine whether under the beliefs of a particular religion it was thought important not to inform third parties of allegations of misbehavior by clergy. The value attached to the religious benefit—valued simply because it was religious—would be derived from the Free Exercise Clause in the light of its underlying constitutional philosophy. Attaching more than a certain value to religious goals, in view of the risk of a particular harm deemed such from a secular point of view, might violate the Establishment Clause.16
It is necessary now to bring into consideration Employment Division v. Smith, the well-known peyote case that reinterpreted the Free Exercise Clause.17 Only a few tort cases involving churches refer to Smith and almost none discuss it at any length. In some cases Smith is dismissed as not affecting the protection churches are suggested to have under constitutional principles shortly to be discussed, namely the right of church autonomy and the right not to have government excessively entangled with religion.18 But the validity of these principles and their relation to Smith are not at all clear.
The Supreme Court, in Smith, held that the Free Exercise Clause does not prohibit the application of a neutral and generally applicable law to religiously motivated conduct, even though there is no [*PG1174]compelling state interest to justify the law. In Smith, a prohibition against possessing peyote was applied to persons who had used it as part of a religious ceremony. In a subsequent case, however, Church of the Lukumi Babalu Aye, Inc. v. City of Hialeah, the Court made clear that if a law “targeted” religion for the imposition of a burden, it was not governed by the Smith rule.19 In Lukumi, city ordinances prohibited the ritual killing of animals while permitting killing for a wide variety of secular purposes. But the meaning of Lukumi is not clear. For instance, is a rule “neutral and generally applicable” for Smith purposes if religiously motivated conduct is grouped with some secular activities for prohibition, but other secular activities are permitted? What would be the result in Lukumi if nearly all killing of animals was prohibited, including religious killing, but killing for purposes of science was permitted? What is in question here is the constitutionally permissible relative effect of government action on the positions of religion and secular values in society.
Negligence in supervision of a person to whom one has given directions might be proposed as a neutral law of general application within the meaning of Smith. But should it be so regarded if heed is paid to the policy that underlies Smith? Recall what was said earlier regarding the value to be attached to the objective sought to be achieved by A giving a task to B when the objective is religious.20 If no value is attached to the religious objective, then religion probably has been “targeted” for a disadvantage in the Lukumi sense. If under the law of torts a fixed value is attached to a religious objective simply because it is religious, then the question is whether this value accords with the value that must be attached to the practice of religion by virtue of the Free Exercise and Establishment Clauses.
In Employment Division v. Smith the Supreme Court stated that its principle does not apply when there is “individualized governmental assessment of the reasons for the relevant conduct.”21 It was because of this exception that Sherbert v. Verner, a case involving a Seventh-day Adventist’s right to unemployment compensation when she could not get a job because of a religious scruple against working on Saturday, [*PG1175]was saved from overruling.22 When a case falls within the category of “individualized assessment,” there must be a compelling state interest to justify the imposition of a burden or the denial of a benefit, the requirement generally applicable under the Free Exercise Clause before Smith.23 The idea behind this exception to the Smith rule—if it is an exception rather than simply an application of the “neutral and generally applicable” test—may be that when there is individualized assessment, either through the development of the common law or in the course of interpreting broad statutory language, official attitudes toward religion, for instance as a reason for not working, attitudes that do not accord with the Constitution’s standard for the value that must be attached to the practice of religion, may creep in and determine outcomes. So, in such circumstances, in a somewhat awkward response to that danger, the requirement of a compelling state interest is retained.
A requirement of nonnegligence in supervision, assuming it passes the neutral and generally applicable test of Smith, would appear to be an excellent candidate for inclusion in the “individualized assessment” category. Its application is closer to deciding whether a person is unavailable for work without good cause—the question posed in Sherbert—than to deciding whether peyote has been possessed.24 In an ordinary tort suit, as noted earlier, in deciding whether supervision was negligent, a number of factors would be considered: the seriousness of the harm threatened, the risk of that harm occurring, the importance of the good sought to be achieved by the enterprise and the likelihood that the achievement of that good would be threatened by requiring a particular kind of supervision. Characteristically, in the law of negligence, no great degree of certainty of law is ever achieved or even attempted. Under these conditions, the door would be wide open for an administrator or a court or a jury to attach no value to a religious objective or to attach a value that is less than that required by the religion clauses. Thus a finding of negligence could well conflict with the constitutional standard regarding the permissible [*PG1176]relative effect of government action on the positions of religion and nonreligion. In addition, of course, there would always lurk the danger of a violation of the Blue Hull-Wolf prohibition, through a finding of negligent supervision based on the fact that the church had fallen below its own standard.
Assume that a claim of negligent supervision comes under the “individualized assessment” exception to Smith. Assume also that conduct has been identified as a basis for liability that avoids violating the Blue Hull-Wolf prohibition. These conditions might be satisfied either by a rule that applies to all who engage in specified conduct or by a rule specially crafted for religious organizations. In the latter case it may be that even though the rule involves special treatment for religion, it does not violate the fundamental norm of the religion clauses, which, as already stated, dictates the permissible relative effect of government action on the positions of religion and nonreligion. This is still not the end of difficulties, for under the “individualized assessment” exception to the Smith rule, religiously motivated conduct may not be prohibited or burdened unless there is a compelling reason for doing so. Whether there is a compelling reason requires consideration of the gravity of the harm threatened—compare sexual molestation of a minor with sexual relations between a minister and an adult member of the church who sought pastoral counseling—and the risk that the harm would occur. These factors must be weighed against the fixed value that the Constitution attaches to the fact that the kind of supervision the church provided, or the lack of supervision, was based upon religious belief.25 The reason for the church’s supervision or lack of supervision may be analogous to the Amish reason for not sending their children to school past the age of sixteen, a reason that was required to be honored under the Free Exercise Clause in Wisconsin v. Yoder.26
So far I have spoken only of the prohibition laid down by the Blue Hull-Wolf rule against answering religious questions and the Smith rule regarding neutral and generally applicable laws, along with its exception for individualized assessments. But a number of other notions have been deployed by courts in measuring the constitutionality of tort claims of the sort I have been discussing—negligent supervision of clergy and breach of fiduciary duty—against the requirements of the religion clauses. One of these ideas has been expressed as a constitutional right of “church autonomy.” In some instances, talk of a right of church autonomy has been simply a way of referring to the Blue Hull-Wolf prohibition.27 In other cases, however, it is said or implied that the right of church autonomy is a distinct constitutional right and that it provides a bulwark against the Smith rule.28 Thus it is suggested that the right of autonomy protects churches against government regulation even when regulation takes the form of a neutral rule of general applicability and the facts do not bring the case within any of the exceptions stated in Employment Division v. Smith, including the “individualized assessment” exception.29 Some judicial statements of the right of church autonomy go so far as to suggest that it protects churches against regulation even though particular conduct was not religiously motivated, so long as the conduct took place within the context of a religious institution.30 This claim of church autonomy is [*PG1178]sometimes expressed by stating that under the Constitution, civil courts lack jurisdiction over tort suits against churches.31
In other cases, a right not to have government “excessively entangled” with religion or religious institutions is named as the barrier against suits for negligent supervision and breach of fiduciary duty.32 “Excessive entanglement” is a term usually associated with inquiries into determining the validity of programs of government aid to religiously-affiliated institutions under the Establishment Clause, but in some cases it has been invoked as an aspect of Free Exercise protection.33 Sometimes government wished to provide protection for religion, and the Court has cited the avoidance of excessive entanglement as one reason for allowing the protection.34 In other cases, the notion of excessive entanglement was invoked as the basis for a positive right, founded in either the Free Exercise Clause or the Establishment Clause, to be free from regulation even when government wished to impose it.35
The case of NLRB v. Catholic Bishop furnishes some support for a right of nonentanglement based on the Free Exercise Clause.36 In that case, the National Labor Relations Board (“NLRB”) asserted jurisdiction over labor relations between Catholic schools and their lay teachers.37 The Supreme Court interpreted the statute involved, the National Labor Relations Act, not to confer jurisdiction on the NLRB because, as the Court said, if the statute were otherwise interpreted, a grave question of constitutionality under the religion clauses would be presented, especially because of entanglement of government with a religious institution.38 The Court came very close to holding that if the statute conferred jurisdiction, it was unconstitutional.39 The Court was particularly concerned with the danger that administrators would impose their own notions about religion in general or Catholicism in particular or, consciously or unconsciously, intrude secular values [*PG1179]where religious beliefs were entitled to prevail.40 Catholic Bishop, of course, is a pre-Smith case.41
The notion of excessive entanglement may turn out to be indistinguishable from the claimed right of church autonomy and, like it, extend a mantle of protection to religious institutions even against neutral laws of general applicability and even though there is no attempt by government to answer a religious question. A case invoking a broad notion of a church’s right to exclude regulation involved a claim of discrimination on the ground of sex in a ministerial appointment.42 Whether the ideas of church autonomy and nonentanglement can be limited to ministerial appointments and not extended to all aspects of church activity seems doubtful, for no reason presents itself why ministerial appointments should be singled out from other decisions made in pursuit of a religious mission, at least not without embracing a particular idea of what is important in religion. Some courts suggest that the right of church autonomy or nonentanglement is an absolute right, at least in regard to ministerial appointments, and that this right does not give way even in the face of a compelling state interest.43 If the idea of ministerial appointment is confined to asserting that a person holds a particular ecclesiastical office or possesses a certain spiritual power, the suggestion may be unobjectionable. But if the idea of appointment goes beyond that and includes some identifiable secular facts, the suggestion is hard to accept. It is difficult to believe, for example, that the religion clauses prevent application to a church of a rule prohibiting putting a person who has been convicted of a sexual offense against a minor in contact with minors, even though such contact is permitted or even required by religious doctrine.
The seemingly indistinguishable ideas of church autonomy and nonentanglement may provide a barrier to suits based on claims of negligent supervision and breach of fiduciary duty.44 But, as already [*PG1180]suggested, such tort claims may be more easily derailed by one of the principles earlier discussed, that a religious question may not be answered by government and that when a case involves “individualized assessment,” a compelling state interest is required.45 But governmental regulation can be imagined that may escape these other limitations and leave only the claimed right of church autonomy or nonentanglement. Consider, for instance, a statute that requires employers to report to civil authorities accusations of sexual abuse made against employees. Possibly the church and the minister can be identified as employer and employee by virtue of facts of a purely secular character. The rule is broadly applicable to all employers and does not call for any “individualized assessment.” Under these circumstances, assuming Smith to be established law, only invocation of some notions such as church autonomy and nonentanglement can keep open the possibility of requiring a compelling state interest to justify the regulation, or in their more absolutist form, to sustain a claim of immunity even in the face of a compelling state interest.