April 19, 2003

Copyright (c) 2000 School of Law, Santa Clara University
Santa Clara Law Review


40 Santa Clara L. Rev. 991

LENGTH: 12221 words


Marc A. Franklin* & Matthew Ploeger**

* Frederick I. Richman Professor, Stanford Law School. LL.B., Cornell Law School; A.B., Cornell University. A version of this essay was presented at the Santa Clara Law Review Symposium, Law, Ethics, and the Good Samaritan, held at Santa Clara University School of Law on March 24, 2000.
** J.D. candidate, Stanford Law School; Ph.D., University of Texas at Austin; M.A., University of Texas at Austin; B.A., University of Texas at Austin.

  ... Since no civil court has given content to a civil duty to rescue, Vermont's well-known and typical criminal statute will be used for purposes of discussion. ... It is important to recognize that although tort law has long used the violation of a criminal statute to help define the content of due care - it has done so only after a duty to exercise care has been independently imposed. ...  


I. Introduction
This essay explores whether a civil duty to rescue n1 should be imposed on a person who has the apparent ability to save another person or to prevent that person from entering a position of peril. n2 It also examines the related question of  [*992]  whether a civil duty to report child abuse should be imposed. A government that can achieve a result by offering incentives and encouragement should take that route rather than mandating action. If that course does not achieve the desired result, then the government must face the question of whether a command is appropriate. In that spirit, Part II of this essay reviews the existing incentives in the rescue situation. Part III turns to the arguments for and against imposing a civil duty of easy rescue. It concludes that, given the existing incentives and the arguments against the civil duty, no such duty should be imposed. Part IV then asks whether the analysis retains its force in states that adopt a criminal duty to rescue. It also considers the impact of legislation that requires those who are at the scene of a crime or who believe that child abuse has occurred to report this information to police or other authorities. It concludes, finally, that the case for a civil duty of easy rescue has not been established, but that the case for a civil duty to report child abuse is a much closer question.

II. Civil Incentives to Rescue or Aid
The story of the Good Samaritan is one that resonates deeply in the United States. In recognition, state and federal governments have created a broad array of incentives to encourage rescues. This section identifies the most important of these incentives in an attempt to gauge whether the cluster of incentives is adequate to the situation - whether taken together these incentives might render unnecessary the creation of a duty to rescue. These incentives fall into two major categories: (1) alleviating the loss to rescuers who are injured while attempting rescues, and (2) protecting those who are sued because of events occurring during their attempts to rescue or render aid. Governments offering incentives have not limited either type of incentive to the easy rescue situation.

A. Encouraging Conduct by Aiding Those Hurt While Trying to Help Others

1. Legislation
Perhaps the most common statutes in this category are  [*993]  found in the universal state legislative schemes that compensate victims of violent crime. n3 These statutes generally extend coverage to rescuers who are injured while trying to protect or rescue crime victims. n4 Although some of these statutes limit recovery to victims who demonstrate need or serious financial hardship, n5 rescuers need not meet that condition. n6 These statutes generally treat rescuers more generously than they do victims. n7 This may be seen as an effort to encourage rescue in a statute that is not otherwise concerned with inducing conduct. A related approach singles out rescuers for compensation if they are injured while trying to rescue. n8

2. Common Law
Common law doctrines treat rescuers more favorably than they treat victims. The courts often establish a tortfeasor's duty not only to the victim, but also to rescuers attempting to aid the victim. The most famous example is  [*994]  Wagner v. International Railway Co., n9 in which the plaintiff was injured while searching in the dark for his cousin who had fallen from defendant's train due to the negligence of the crew. Judge Cardozo responded to the argument that the defendant owed no duty of care to the rescuer with the famous passage: "Danger invites rescue.... The wrong that imperils life is a wrong to the imperiled victim, it is a wrong also to his rescuer.... The wrongdoer may not have foreseen the coming of a deliverer. He is accountable as if he had." n10 Moreover, courts extended this notion by eliminating the requirement that the danger imperil a third party. A defendant who is careless for his or her own safety will be liable if that conduct leads to the injury of a rescuer at the scene. n11

The desirability of rescues has also been recognized in a line of cases imposing tort liability against those who negligently n12 or intentionally n13 interfere with attempted rescues.

 [*995]  In addition to helping rescuers establish duty, courts have been reluctant to find rescuers contributorily negligent for trying to rescue or for how they attempted the rescue. During the era when contributory negligence was a complete bar to recovery, n14 most courts protected rescuers by holding that the conduct of a plaintiff rescuer could not be considered contributorily negligent unless it reached the level of recklessness. n15

The emergence of comparative negligence has led to arguments that allegedly negligent rescuers no longer need the benefits of the earlier rule barring consideration of a rescuer's possibly negligent conduct. Opponents of the earlier rule argue that a rescuer's negligence should be straightforwardly compared with the negligence that led to the need for rescue. Although some courts have agreed, n16 at least one court has asserted that society still needs to encourage rescuers, and that this goal may be furthered by retaining the earlier rule that there was no contributory negligence to compare unless the rescuer acted rashly or recklessly. n17


3. Non-Governmental
Incentives to rescue may come from non-governmental sources as well. The media provides extensive coverage of heroic acts, n18 human caring, and, as in the Kitty Genovese case, accounts of the failure of others to help someone in peril. n19 In addition, wealthy citizens and communities often arrange to help caring people, heroes, or their survivors, for example by creating college funds for their children. n20

B. Encouraging Conduct by Protecting Those Sued After Trying to Help Others
In addition to helping persons hurt during their efforts to rescue, the government also helps rescuers who are sued for alleged misconduct during the rescue. Even though the government has not compelled the rescue, n21 it nonetheless encourages such an effort by protecting rescuers against some  [*997]  tort liability. Although some of these examples may not involve imminent danger, they all involve situations in which life and health are at stake.

The most widely known protection against lawsuits reflects the medical profession's successful efforts to obtain greater protection when sued over the rendition of emergency services in so-called "Good Samaritan" situations - those in which no prior doctor-patient relationship exists and no remuneration is anticipated. Physicians claimed that fear of being sued for less-than-perfect results in such situations inhibited them from stopping at roadside accidents and similar emergencies to render aid. In response, legislation extending such tort protection has been enacted in every jurisdiction. n22 The resulting legislation protects physicians, and in some cases other personnel, n23 from liability unless their conduct is found to be grossly negligent or worse. n24 The lobbying effort for tort protection of the medical profession took an unexpected turn in Vermont to produce the nation's first general duty-to-rescue legislation. n25

Closely related to the statutes protecting the medical profession are those that protect non-medical personnel who administer emergency assistance. These protections may be based on the employment status of the person attempting the aid, n26 his or her relationship to the victim, n27 or explicitly on the training the rescuer has had. n28

 [*998]  An obscure but important example involves federal and state efforts to encourage the donation of outdated, but still edible, foods and crops. Americans annually discard approximately "[fourteen] billion pounds of food, a good portion of which is still edible." n29 Commercial sources of such food - restaurants and processors - were apparently reluctant to donate leftovers for fear of being sued if the food caused illness. Congress and some states responded by extending protection against tort suits. Congress provided that such suppliers "shall not be subject to civil or criminal liability arising from the nature, age, packaging, or condition of apparently wholesome food or an apparently fit grocery product that the [supplier] donates in good faith to a nonprofit organization for ultimate distribution to needy individuals." n30 Coordinated protection is extended to the nonprofit organizations distributing the food. n31

Finally, protection against liability for defamation provides another example of how the government encourages prosocial behavior. Before much of libel law was constitutionalized, it was a system of strict liability relieved by qualified (and occasionally absolute) privileges in situations in which the state encouraged persons to speak. What resulted was a series of qualified privileges for, among other things, n32 reporting suspected crimes to law enforcement agencies, n33 internal discussions among partners about the operation of a business, n34 and responses of present or former employers to requests from prospective employers for information about job  [*999]  applicants. n35 Without protection in these situations, essential communications would have been inhibited. One striking example is the report of a commercial airplane crash attributed to a pilot who had recently worked for another airline. The pilot "had resigned from [the first] airline to avoid being fired for failing a critical flight test." n36 The second airline did not obtain that information from the first airline until after the accident, and admitted that it would not share such information about its own present or former pilots. Asserting fear of libel suits, the second airline asked Congress for immunity for airlines that share such information. n37

In sum, the government provides an impressive array of incentives to rescue. Still, some may conclude that incentives are inherently insufficient and that tort law should mandate rescue. Part III addresses that contention.


III. Should a State Impose a Tort Duty to Rescue?
Tort liability should not be imposed unless that imposition can be justified on tort criteria. Even when a defendant's undoubtedly careless conduct has harmed another, tort law asks whether the defendant owed the plaintiff any duty not to act in that manner. n38 Courts face concerns that include unbounded and disproportionate liability, the impact of judgments on defendants and society, the need to deter dangerous conduct, the administrability of new claims in terms of numbers of suits and the nature of proof, and whether a new damage claim can be effectively controlled. The oft-repeated functions of tort law - compensation and deterrence - do not apply in the duty-to-rescue area because they are usually invoked in cases in which a defendant's conduct caused the victim's plight. As tort law currently operates, courts must initially focus on whether the defendant's conduct warrants the imposition of liability. Compensation, which focuses on the plaintiff, comes into play only after a court concludes the defendant's conduct warrants the imposition of a duty. Admittedly, an emphasis on compensation would lead to more duties than an emphasis on deterrence. If, however,  [*1001]  compensation were the main focus, tort law would not subordinate compensation nearly as often as it does.

This is not the occasion to reconsider the underlying nature of duty in tort law. The point here is that tort law is already concerned with exposing defendants to liability in cases in which their careless behavior has hurt others. Courts will be even more uneasy about imposing duty in a case in which the defendant has done absolutely nothing to create the plaintiff's danger. Is there a need for rescues that justifies the state telling its citizens not only to avoid carelessly hurting others, but also that they must affirmatively act in certain situations or risk extensive liability? n39

No civil court has ever demanded that one stranger rescue another. This essay does not suggest that tort law is incapable of development. n40 Rather, it is important to identify and make explicit the reasons underlying this failure to require rescue. Several justifications require discussion. n41 First, the legislatures and courts may have concluded that even where the incentives are inadequate, personal freedom is more important than saving lives by mandating rescue. Second, theory aside, governments may have concluded that there is no need to mandate action - the incentives are working adequately. Third, tort law's sanctions are potentially much more severe than those of the criminal law. n42 Fourth, there is the complication of how to proceed when several people fail to take the required action - a serious concern since lack of action is most likely to occur with groups. n43 Can the victim sue all those who failed to rescue him? Have they all "caused" the death or injury of the victim by their inaction? How does the court apportion any resulting liability? Fifth,  [*1002]  since tort law necessarily addresses the conduct of plaintiffs as well as that of defendants, how the plaintiff got into the perilous situation may affect the defendant's liability for not rescuing. Sixth, even if a court were otherwise persuaded to impose a duty to rescue, it might find retroactive imposition of such a duty quite unfair. Finally, perhaps a criminal mandate, with its exclusive focus on defendants, will be more effective in obtaining compliance.

Since no civil court has given content to a civil duty to rescue, Vermont's well-known and typical criminal statute will be used for purposes of discussion. The Vermont provision provides:

A person who knows that another is exposed to grave physical harm shall, to the extent that the same can be rendered without danger or peril to himself or without interference with important duties owed to others, give reasonable assistance to the exposed person unless that assistance or care is being provided by others. n44
Any court considering imposing a civil duty to rescue would need to address the issues that this provision addresses and would not impose more onerous civil obligations. Since every proposal of a duty to rescue (whether criminal or civil) is limited to mandating only "easy" rescues, n45 the Vermont  [*1003]  statute - with its exemption for danger to the rescuer and for duties owed others - seems a reasonable standard to hypothesize. With this statute as a backdrop, this essay now considers the arguments against a tort law duty to rescue.

A. Personal Freedom
For some, personal freedom is a core issue. Some argue that the government should never force citizens to take steps to protect each other. n46 Communitarians disagree. n47 Apart from disagreements over the nature of government and society, some dispute whether a mandatory duty would stifle or undermine the notion of altruism - and whether that is important for society. n48 However important this philosophical  [*1004]  point may be, we are unwilling to rest our argument on it. Although we reject arguments that rely entirely on personal freedom, n49 we nonetheless reject the imposition of a civil duty to rescue for other reasons.

B. Is There a Need?
There is no evidence that citizens are refusing to make easy rescues. The oft-raised paradigm here is the baby who crawls into a few inches of water in a puddle and might drown. If citizens en masse do not rescue drowning babies, one might well argue that some legal mandate is appropriate. It appears, however, that drowning babies are rescued. Despite the lack of data, one can surmise that when the need for rescue is clear and the situation safe, rescues do occur. n50 Those who propose a mandated duty to rescue must show that failures to attempt easy rescues are a problem in our society.

If the argument is that today's perils create a greater need for intervention than in the past, this is an argument in favor of reducing unjustifiable perils and educating citizens about how to minimize or avoid these perils rather than requiring rescue. Most contemporary assertions of a need for a duty to rescue arise in the context of criminal conduct. n51 Dangerous situations involving ongoing criminal conduct raise much more difficult questions for proponents of mandated  [*1005]  easy rescues. n52 Even such proponents recognize that today's society presents physical risks that are much more common and deadly than those existing when James Barr Ames proposed his duty of easy rescue in 1908. n53 In today's society, it may be unrealistic to demand that citizens stop at an accident scene in an unpopulated area, even if someone appears to be seriously hurt and the situation has no obvious dangers. n54

C. Civil Liability for Damages
Those states that impose a criminal duty to rescue impose only minimal penalties. n55 Tort law, on the other hand, might award damages in death or injury-aggravation cases in the tens or hundreds of thousands of dollars. This potential liability suggests that tort law may be far more threatening and burdensome for many persons than criminal law. The law abounds in situations in which civil liability may be far more onerous from a financial perspective than criminal liability. n56 For judgment-proof defendants, however, tort law is  [*1006]  essentially irrelevant, though the bite of the criminal law remains. n57

D. Complexity of Civil Litigation
When multiple parties are involved, civil cases present difficulties beyond those in criminal cases. Among these is the determination and apportionment of liability among several defendants. Again, the drowning baby paradigm offers no complications. But a variation of the Sherrice Iverson case would involve many complexities. n58 First, there is possible negligence by a father who lets his young daughter wander off and play with strangers. In a survival action for the child's emotional distress before her death, n59 some states would judge the father's conduct by the reasonable parent formula. n60 Second, there is the battery committed by Jeremy Strohmeyer. Tort law has recently focused on apportioning liability among negligent and intentional tortfeasors. n61 Third, if the casino had guards responsible for monitoring the area, their negligent failure to protect an invitee must be considered. Suffice it to say here that there is confusion and disagreement over whether to apportion percentages of liability and, if so, how to do it.

Only when the next defendant, David Cash, is added to the mix, does the crucial question arise. Assuming a civil  [*1007]  version of the Vermont statute is in effect, and that Cash clearly violated it, n62 how should liability be apportioned among the four defendants? How much liability can be apportioned to the non-rescuer as opposed to the person who carelessly allowed the child to get into a position of peril, guards monitoring events in the casino, and the murderer? Since most states have abolished joint-and-several liability, n63 the apportionment issue is important. Depending on the percentage of fault allotted to Cash, he would be liable in some states only for that percentage of liability. In California, Cash would be jointly and severally liable for any economic harm, but only severally for non-pecuniary harm. n64 In a situation as complex as this, it is doubtful that a lawsuit provides enough return for society or the plaintiff to be worth adding the non-rescuer to the mix. n65

E. Defenses
Whether the victim is an adult or a child, tort law must address claims about the victim's carelessness or recklessness in getting into a situation that required or appeared to require rescue. For example, what about a plaintiff who dives into a river knowing that he does not know how to swim but alerting those on the banks to his impending dive? n66 Is this to be compared with the defendant's failure to rescue or clumsy rescue effort? Again, the courts are in disarray on this issue. n67 This may add still another party's conduct to the fault mix.

F. Retroactivity
Even if a court were inclined to impose a civil duty to rescue,  [*1008]  it might be constrained by the retroactive operation of the judicial process. A defendant who understood the law to permit failure to rescue would now be informed that the court has decided to impose a duty and that the defendant must pay the plaintiff thousands of dollars. Although this can be avoided by the use of prospective rulings, n68 courts are understandably uneasy about acting like legislatures. n69 Retroactivity is unattractive in such a situation. n70 Despite occasional judicial threats to impose affirmative duties to act based solely on foreseeability, n71 the courts seem generally and genuinely uneasy about expanding the range of special relationships that warrant such action, n72 and can be expected to be even more wary about imposing such a duty on a complete stranger.

G. The Relative Advantages of Criminal Law
How much is gained by creating a new tort with these probable complications? Although this essay addresses only questions of civil liability, and takes no position on criminal liability, it is impossible to discuss civil liability without recognizing  [*1009]  that most commentators focus on imposing a criminal duty to rescue. Aside from compensation, criminal law can do virtually everything, and maybe more in the case of judgment-proof defendants, while avoiding questions of apportionment, victim fault, and unfair surprise. Moreover, prosecutors, exercising their discretion, are far more likely to screen cases and decline to pursue difficult or close cases than are victims and their survivors.

It is no answer to say that the tort complications pale beside the social need to encourage the desired relationship among citizens. They may pale beside the need for some legal recognition of a duty to rescue. But if society wants to impress upon the public the importance of helping others, criminal law is preferable to tort law. Tort law can cause financial ruin to a defendant who has correctly or incorrectly been found cowardly, or who negligently misjudged the level of risk. If symbolism were the issue, the focus of criminal law would be entirely on the person whose conduct society seeks to change. This would leave aside the conduct of others, including the victim - considerations that tort law cannot avoid. It is a large step - one that this essay does not encourage or support - from finding that a non-rescuer violated social norms (and perhaps deserves punishment) to finding that the violator should compensate the victim or survivors where the non-rescuer played no part in getting the victim into the peril and bore no pre-existing duty to help the victim. n73

There is no reason to believe that the cluster of incentives reviewed in Part II does not produce enough rescues to argue against the need for civil liability. Among the more important of these incentives are compensation for injured rescuers; tort law's solicitude for injured rescuers, including tort liability on those who interfere with rescues; and the protection of rescuers against liability. If needed, increased or new incentives might produce more rescues. n74 The law aside, public  [*1010]  recognition and media adulation of helpful citizens, including heroes, may provide compelling incentives for many citizens.

The foregoing analysis has assumed that the state had no relevant legislation. As seen earlier, however, several states have in fact enacted criminal obligations to rescue. Part IV addresses the question of whether such legislation alters the analysis of the tort duty to rescue.

IV. The Impact of a Criminal Duty to Rescue or Report
Common law is, of course, affected by relevant legislation. This is particularly relevant if the common law question is whether to adopt a duty in an area in which the legislature has created a criminal obligation. This Part explores the impact of legislation creating a criminal duty to rescue and the adoption of legislation in two related fields - the reporting of crimes and the reporting of child abuse.

A. Statutes Requiring Rescue in Certain Situations

1. Arguments For and Against Extending Criminal Duty-to-Rescue Statutes to Create a Civil Duty
A handful of states have adopted criminal statutes that  [*1011]  explicitly require rescue. n75 If that statute is silent about civil liability, n76 the courts must decide whether the criminal duty to attempt "easy" rescues n77 changes the arguments for imposition of a civil duty. This section explores the effect of such statutes on a civil duty. n78

First, proponents may legitimately argue that the statute establishes the state's public policy favoring rescue. This removes the argument that state policy prefers personal freedom to the saving of life. The legal vacuum that requires a court to make decisions without legislative guidance does not exist here because the legislature has expressed its preference for rescue by enacting the criminal duty to rescue statute. n79

Second, proponents may correctly observe that the criminal statute removes any judicial concern that the court might be imposing an impractical duty - assuming the court were to frame the civil duty along the lines articulated by the criminal statute. The legislature already concluded that it is feasible to criminalize failure to rescue. Furthermore, the passage of criminal legislation does not normally carry the implication that the legislature rejected analogous civil liability. In the absence of some reason to think that tort liability  [*1012]  would interfere with apparent legislative goals, the court must consider whether to impose a civil duty as it would in any other case. n80

Third, a plaintiff may also correctly argue that if the court were to adopt the statutory elements in framing a new tort, no defendant could legitimately claim surprise because the obligation was already made clear by the criminal statute. n81 The usual notice problem that makes civil courts understandably reluctant to impose retroactive affirmative obligations is absent here.

But what do these admittedly valid points show? At most, the above arguments remove some impediments to recognizing a tort duty to rescue, but do not necessarily make an affirmative case that the court should take that step. n82 It is important to recognize that although tort law has long used the violation of a criminal statute to help define the content of due care - it has done so only after a duty to exercise care has  [*1013]  been independently imposed. n83

A more general argument against allowing the existence of the criminal sanction to encourage the adoption of a civil duty is the public-private distinction between criminal and tort law. Some conduct is thought to affront the public interest generally, and thus criminal sanctions are proper. Other conduct is thought to improperly harm individual members of society and properly be subject to damage actions. Of course, some conduct implicates both areas. n84 One state that has imposed a criminal obligation to attempt to obtain aid for victims in certain situations has explicitly barred holding an offender civilly liable. n85 That position also has scholarly support in the rescue context. n86 Professor D'Amato justifies this result by identifying the inaction as a "public wrong" subject to sanction by society through criminal statutes: the victim alone should be responsible for his or her injuries. The basis for criminal liability was the anti-social act of not rescuing - an obligation that does not depend on how or why the victim got into the position of needing help. n87

After recommending criminal liability, n88 Professor D'Amato rejects tort liability for three reasons: n89 (1) the problem of judgment-proof defendants; (2) victims who behave carelessly because they know they must be rescued; n90 and (3) "more generally, imposing tort liability gives a monetary reward  [*1014]  to risk takers and penalizes risk avoiders; such personality traits are not role reversible, and hence the argument that [rescuers] might be in [the victim's] position someday is not persuasive." n91 On the procedural side, Professor D'Amato is willing to trust prosecutorial discretion to decide whether to charge a crime, but is (understandably) unwilling to count on similar restraint on the part of the harmed victim or that victim's survivors. n92

2. Examples of Refusing to Hold Criminals Liable in Tort
Criminal law is frequently used to deter serious misconduct even though civil damages are barred to those hurt by that conduct. The examples of perjury, attorney behavior toward non-clients, and antisocial plaintiffs highlight the differences between the two regimes.

a. Perjury
Perjury is a classic and non-controversial crime in the legal system. Whether or not another person was harmed by the perjury (for example, losing a valid lawsuit or being denied proper medical treatment), however, the courts are virtually unanimous n93 in denying the victim civil relief. This is not because the harm is too trivial for civil action or that the conduct was not seriously wrong. Rather, the articulated justification for not permitting civil relief for perjury is that courts want those who testify under oath to do so free from any impediment to their total candor and honesty. Witnesses should not be tempted to trim their testimony due to a fear of being sued by those hurt by the testimony. n94 This kind of absolute privilege is a drastic device in civil cases, but it is also  [*1015]  employed in several situations to ensure freedom of speech n95 and freedom of action. n96

b. Attorney Behavior Toward Non-Clients
In most states, one party's attorney owes no duty of care to the other party. n97 In these states, an attorney who lies to a non-client or withholds crucial discovery materials causing a non-client to lose a case unfairly faces no civil liability. As one court stated in a case of an attorney who allegedly withheld an essential document during discovery, "if [the attorney] violated his professional responsibility, the remedy is public rather than private." n98

c. Antisocial Plaintiffs
Despite the emergence of comparative fault, situations remain in which the plaintiff behaved so badly that civil claims are barred against the defendant who nonetheless remains liable for criminal punishment. For example, a person hurt while playing with explosives may be barred from suing others who may also be partially to blame. The perceived seriousness of the plaintiff's fault bars recovery despite the possibly criminal nature of the defendant's conduct. n99 This result is even clearer in states that bar a reckless plaintiff from suing a reckless defendant. n100 The law's deterrent function is manifested through the criminal law, and an "undeserving" plaintiff is denied compensation. n101

 [*1016]  Although these three examples are different, they suggest some of the important differences between criminal and civil law. n102 Although the existence of criminal liability may help the plaintiff overcome some of the barriers to tort liability, it does not automatically justify the creation of a civil action for anyone who suffered from that criminal conduct. Sometimes the reason is to be found in distinctions between the "public" and "private" nature of the wrongs, though at other times a civil action may run counter to the tort goal of deterring dangerous conduct. n103

3. Criminal Statutes that Explicitly Anticipate Civil Actions
The analysis is not altered if the criminal statute anticipates civil actions by setting the standard of care for law-abiders. In Vermont, for example, the rescue legislation protects rescuers trying to comply with the criminal law by shielding them from tort liability unless their acts constituted "gross negligence." n104 One might argue that Vermont included this bar to tort liability to encourage those on the fence about rescuing even after the criminal law enactment. n105 It is more likely, however, that the legislature was responding to the perceived unfairness of demanding that a person act and then judging that action by traditional tort "reasonable person" standards. n106 The problem with requiring reasonable care from coerced rescuers is precisely that the law has taken from that person the freedom to decide whether or not to undertake the activity in question. Tort law's normal approach to "reasonable conduct" assumes that those who harm others, because they are clumsy or of subnormal intelligence, may be negligent even if they acted to the best of their ability. Therefore, although jurisdictions vary regarding their statutory protections, none asks more of would-be-compliers than a  [*1017]  "good faith" attempt to rescue. n107

There is no basis for assuming that the Vermont legislature implicitly imposed a civil duty by stating the level of protection in a civil case. The civil section simply protects those who comply with the criminal legislation against what the legislature assumed would otherwise be the obligation of due care. This part of the statute does not address the civil liability of a person who does not attempt to comply with the criminal law. Since the Vermont legislature was fully aware of the uniqueness of its enactment and the lack of a civil counterpart, n108 it seems unlikely that it would impose a civil duty implicitly. Rather, it is more plausible to read the statute as leaving the duty question to the court. n109 There appears to be, however, no reason why Vermont should more readily create a tort duty than a state whose statute is solely criminal. n110


B. Statutes Requiring Reporting of Violent Crimes
In addition to the states with general duty-to-aid statutes, n111 some states impose a more limited duty to summon aid for victims of violent crime. n112 Although some reporting statutes appear to focus on helping the victim, n113 not one requires  [*1019]  that witnesses intervene directly, even where doing so would pose no danger. n114 Rather, the statutes require only that witnesses call for help. n115

Hawaii's crime reporting statute is remarkably similar to Vermont's duty-to-rescue statute. The Hawaii statute is, of course, limited to crimes and requires only reporting - otherwise, the two statutes contain remarkably similar language. Unlike the Vermont statute, however, the Hawaii statute explicitly prohibits civil liability for nonfeasance. n116

Since most crime reporting statutes do not mention civil liability, n117 courts must decide whether to adopt a corresponding civil duty to report crimes. As in the duty-to-rescue context, the existence of a reporting statute alleviates judicial concerns about notice and provides an articulated legislative policy preference for reporting over personal freedom. Moreover, plaintiffs can argue that the duty imposed by reporting statutes is less onerous than that imposed by duty to rescue statutes for two reasons: (1) the reporting statutes apply  [*1020]  only to situations involving victims of certain crimes, not all situations of peril; and (2) the witness need not directly intervene or rescue, but rather only report the crime.

Even considering these differences, however, the existence of a reporting statute does not make a convincing case for recognizing a civil duty to report. As in the duty-to-rescue context, there is no demonstrated need for civil liability for failure to report crime. n118 The oft-cited examples of Kitty Genovese and Sherrice Iverson are so shocking in part because they are so rare.

In addition to a lack of a need for duty to report civil liability, proving causation may be even more difficult here than in the duty-to-rescue context. To prove civil liability for failing to report a crime the plaintiff must show what would have happened if the witness had reported the situation to the appropriate authorities "as soon as practicable." n119 In addition to showing that a report could have been made, and that "as soon as practicable" would have been soon enough, the plaintiff must also show how the police (or medical personnel) would have handled the report, how quickly such personnel would have responded, and to what effect. n120 Given the many uncertainties inherent in this chain of causation, it would be quite difficult to establish what harms, if any, could have been avoided by promptly reporting the crime. Again, it is unfair to expose a person to potentially ruinous civil liability for damages resulting from a situation that he or she did not create.

C. Statutes Requiring Reporting of Child Abuse
The command to report child abuse recognizes an underlying problem that is far more common than failure to rescue and far more serious than previously acknowledged. Every  [*1021]  state now has some form of compulsory reporting obligation. n121 Generally, these statutes require those who have knowledge of, or reason to suspect, child abuse to report it. Some statutes require "any person" with such knowledge to report it; n122 others apply only to health care workers. n123 Most statutes protect those who report child abuse in good faith from civil liability, others grant absolute immunity. n124

Although there are no direct holdings addressing a civil duty to rescue in states with criminal legislation, several state legislatures and courts have confronted that very issue in the context of reporting child abuse. Although some child abuse reporting statutes explicitly provide for civil liability, n125 most statutes are silent with regard to whether someone may be held civilly liable for failing to report abuse. Unlike the situation in the duty-to-rescue context, a few courts have found a civil duty to report child abuse. n126 In J.S. v. R.T.H., n127  [*1022]  the New Jersey Supreme Court held that when the defendant knows or should know that a child has been abused, there is a duty to take reasonable steps - such as warning others, if not reporting - to prevent further abuse. n128 Although statutory reporting requirements do not themselves create a private right of action, as the court noted, they do not preclude it. n129 In the court's view, these legislative attempts to combat the problem of child abuse through criminal law evince a strong desire to ameliorate the problem. n130 The court concluded that criminal sanctions were insufficient to "stem the tide" n131 of sexual abuse of children and that "civil remedies will complement statutory protections and further the legislative efforts to enhance the protection of children." n132 Most courts, however, have declined to find a civil duty to report child abuse, whether based on the reporting statute or common law. n133

 [*1023]  The child abuse reporting situation may be sufficiently different from both the rescue and the crime reporting situations that a different civil result may be warranted. First, the need for child abuse reporting seems to be far more pressing than the need for easy rescue. Child abuse is more pervasive than failures to make easy rescues. It is done in secret and is much harder to ferret out. n134 As a result, much child abuse goes unreported and third party reporting is essential to addressing the problem. n135 Second, victims of child abuse are often unable to articulate their harm or even to contact police or other officials. The children are innocent by definition.

Third, a duty to report child abuse - or just keep the child away from an abusive person - infringes less on personal freedom than does a duty of easy rescue. The child abuse reporting duty is limited to situations involving child abuse, whereas the duty to rescue applies to any person in peril. Further, the duty requires only reporting and not direct intervention. This limitation makes it easier both to define the limits of the duty (i.e., what it would take to discharge the duty) and to comply with the duty.

Fourth, unlike duty-to-rescue or crime reporting statutes, child abuse statutes do not demand that people meet possibly difficult standards of reasonableness. The act of reporting child abuse is not difficult to carry out and is not likely to harm the child or the reporter, in contrast to the fears of some potential rescuers and some potential crime reporters. Nonetheless, it is noteworthy that all seven states that mandate a civil duty to report child abuse also protect reporters against being held liable for unreasonable conduct. n136

 [*1024]  Fifth, the harm occurs slowly and over a period of time, allowing the legal system a better chance to measure the harm done by the delay. The rescue debate, conversely, focuses on a victim who is unharmed at one moment, but then suffers serious harm or death in a very short time period. The extended time frame and the repetitive nature of the injuries makes proving causation less of a problem in child abuse cases than in the general crime reporting context. n137

In sum, it may well be more justifiable to impose a civil duty to report child abuse than a civil duty to rescue or a duty to report crime. n138 The ultimate test of whether to create a civil duty of the sort created in J.S. v. R.T.H. may be whether the goal of the criminal law needs bolstering and it appears to need that bolstering in the child abuse area. n139

These examples of mandated conduct n140 show that whenever  [*1025]  legislatures address civil liability in these contexts, they have taken steps to protect (at least to some extent, if not completely) those who try in good faith to comply. This would suggest that if a court created a common-law duty to rescue it should not adopt the due care standard as the measure of compliance.

Although the pressing situation of rescuing a stranger in imminent danger of death or grave physical harm has understandably captured the public's attention, it has given rise to criminal obligations in very few states and to civil obligations in none. Although a civil duty to report child abuse may be warranted, it would not support a civil duty to rescue or to report crime generally.

V. Conclusion
The nature of incentives, of criminal duties to rescue, and other mandatory criminal and civil duties to rescue or report, all shed light on the main question: should there be a civil duty to rescue? This essay has argued that the presence of incentives might reasonably be seen as the right way to induce the desired conduct, and that more incentives might be considered. A criminal duty to rescue should not generally lead the court to impose a civil duty, even though the existence of the criminal statute does remove several impediments to the creation of civil liability. n141 Tort liability is not an appendage to be grafted onto criminal liability in order to add "muscle." The situations explored in this essay demonstrate  [*1026]  the fallacy of assuming that tort law is necessarily "less harsh" on defendants than criminal law.

Most of the mandatory criminal duties imposed on citizens have no statutory or common-law civil counterparts. Those that do have related statutory or common-law civil duties involve situations in which the obligation being imposed on the citizen is much more easily defined and reviewable after the fact (e.g., a report or a phone call) than in rescues.

In conclusion, tort obligations to rescue (though perhaps not those to report child abuse) have not been shown to be necessary. Society may conclude that incentives have not sufficed and cannot be strengthened, and decide that some duty is appropriate. If so, it should consider a criminal duty - the duty that focuses solely on changing the behavior of those unwilling to perform easy rescues, without regard to what brought forth the need for the rescue.

n1. Whether the proposed duty to rescue has been framed in terms of criminal law, see, e.g., Joshua Dressler, Some Brief Thoughts (Mostly Negative) About "Bad Samaritan" Laws, 40 Santa Clara L. Rev. 971 (2000), or civil law, it has always been framed in terms of "easy" rescue. For various attempts to define "easy" rescue, see Daniel B. Yeager, A Radical Community of Aid: A Rejoinder to Opponents of Affirmative Duties to Help Strangers, 71 Wash. U. L.Q. 1, 25 (1993) (questioning what easy rescue means and finding the statutes ambiguous and that no courts have addressed the issue); Jennifer L. Groninger, Comment, No Duty to Rescue: Can Americans Really Leave a Victim Lying in the Street? What Is Left of the American Rule, and Will it Survive Unabated?, 26 Pepp. L. Rev. 353, 368 (1999) ("An easy rescue is one where a victim is in danger and a potential rescuer is in a position to alleviate the harm without any significant cost to himself."); Sungeeta Jain, Note, How Many People Does it Take to Save a Drowning Baby?: A Good Samaritan Statute in Washington State, 74 Wash. L. Rev. 1181, 1189 (1999) (defining "easy rescues" as those that "involve no danger to the rescuer and do not interfere with duties the rescuer owes to others").

n2. Although this symposium focuses on situations in which immediate action is required, we will range more broadly, considering cases and statutes that address situations more remote in time and space from the typical "rescue" - though they may involve more lives. For example, one person may mislead others to the detriment of still others. See, e.g., Randi W. v. Muroc Joint Unified Sch. Dist., 929 P.2d 582 (Cal. 1997) (examining instance when employment references that were silent about prior predatory conduct led school district to hire man who assaulted student). See infra notes 36-37 and accompanying text.

n3. See Desmond S. Greer, A Transatlantic Perspective on the Compensation of Crime Victims in the United States, 85 J. Crim. L. & Criminology 333, 334 & n.8 (1994) ("By 1992,...every state had enacted a compensation program."). In 1984, Congress enacted the Victims of Crimes Acts, establishing the Crime Victims Fund, which provided additional funding for state crime victim compensation programs. See Victims of Crimes Acts, Pub. L. No. 98-473, 98 Stat. 2170 (codified as amended at 42 U.S.C. 10601-10604 (1988 & Supp. 1993)).

n4. See Greer, supra note 3, at 349.

n5. See id. at 368-70; see also, e.g., N.Y. Exec. Law 620-635 (McKinney 1999).

n6. See Greer, supra note 3, at 351 & nn.90-91, 369 n.183; see also, e.g., N.Y. Exec. Law 631(5)(c).

n7. See Greer, supra note 3, at 351 & n.91, 369 n.183; see also, e.g., N.Y. Exec. Law 631(5)(b) & (c) (exempting Good Samaritans from provisions that reduce the amount of awards based on contributory conduct and allowing Good Samaritans to recovery for property damage).

n8. See, e.g., Cal. Gov't Code 13970 (West 1999). In addition, some states formally encourage heroic behavior on the part of prisoners by rewarding such conduct. Idaho allows the director of the department of corrections to reduce the sentence of a prisoner who performs "an extraordinary act of heroism at the risk of his own life or for outstanding service to the state of Idaho which results in the saving of lives, prevention of destruction or major property loss during a riot, or the prevention of an escape from a correctional facility." Idaho Code 20-101D (1986). For example, two prisoners at a work camp in Idaho recently risked their lives in an unsuccessful attempt to save a driver who had crashed his car into a canal. Under the Idaho law, they may petition the director for reductions in their sentences. See 2 Prisoners May Get Time Cut for Heroism, N.Y. Times, Jan. 6, 2000, at A18. To similar effect, Iowa allows for the restoration of previously revoked good conduct time in recognition of heroism. See Iowa Code Ann. 903A.3 (West 1998).

n9. Wagner v. International Ry. Co., 133 N.E. 437 (N.Y. 1921).

n10. Id. at 437-38.

n11. See Sears v. Morrison, 90 Cal. Rptr. 2d 528 (Ct. App. 1999) (holding that a defendant who negligently became trapped under a swamp cooler was liable to the person who was injured while trying to rescue him); Carney v. Buyea, 65 N.Y.S.2d 902, 908-09 (App. Div. 1946) ("We think the defendant, by parking her car as she did, exposed herself to undue risk .... Her act in that respect was wrongful to the plaintiff since it brought about an undue risk of injury to him causing him to undertake her rescue to his injury and damage."); see also F. Bohlen, Liability in Tort of Infants and Insane Persons, 23 Mich. L. Rev. 9, 30 n.33 (1925) ("[A] person who carelessly exposes himself to danger or who attempts to take his life in a place where others may be expected to be, does commit a wrongful act towards them in that it exposes them to a recognizable risk of injury.").

n12. See Soldano v. O'Daniels, 190 Cal. Rptr. 310 (Ct. App. 1983) (holding that a bartender had a duty to permit a third party to call the police to aid the plaintiff); Griffith v. Southland Corp., 617 A.2d 598 (Md. Ct. Spec. App. 1992) (holding that a convenience store employee had a duty to call for police to aid a police officer who was being assaulted outside the store). The Restatement (Second) of Torts states the law this way:

One who knows or has reason to know that a third person is giving or is ready to give to another aid necessary to prevent physical harm to him, and negligently prevents or disables the third person from giving such aid, is subject to liability for physical harm caused to the other by the absence of the aid which he has prevented the third person from giving.
Restatement (Second) of Torts 327 (1979).

n13. See Riggs v. Colis, 695 P.2d 413 (Idaho 1985) (recognizing an action for aiding and abetting assault and battery where defendant prevented, at knife point, a bystander from intervening in an attack by a thrid party on the plaintiff); Barnes v. Dungan, 690 N.Y.S.2d 338 (App. Div. 1999) (recognizing a wrongful death action alleging intentional interference with lifesaving medical assistance where supervisor prevented worker from administering CPR to the victim of a heart attack); see also Restatement (Second) of Torts 326 ("One who intentionally prevents a third person from giving to another aid necessary to prevent physical harm to him, is subject to liability for physical harm caused to the other by the absence of the aid which he has prevented the third person from giving.").

n14. Contributory negligence is still a complete bar in five jurisdictions (Alabama, Maryland, North Carolina, Virginia, and the District of Columbia).

n15. See Eckert v. Long Island R.R. Co., 43 N.Y. 502 (1871). There is a minority strain of opinions holding that a rescuer who knew the danger of the situation, perhaps in entering a burning inhabited house, "assumed the risk" of the resulting injury, freeing the defendant who negligently set the house on fire from any liability to the rescuer. See id. at 505 (Allen, J., dissenting).

n16. Most courts addressing this issue have focused on the fact that comparative negligence removes the harsh consequences of contributory negligence and have ruled that a plaintiff who is negligent in performing a rescue should recover only a pro rata share of the damages sustained attributable to the defendant. See, e.g., Zimny v. Cooper-Jarrett, Inc., 513 A.2d 1235 (Conn. App. Ct. 1986); Ryder Truck Rental, Inc. v. Korte, 357 So. 2d 228 (Fla. Dist. Ct. App. 1978); Sweetman v. State Highway Dep't, 357 N.W.2d 783 (Mich. Ct. App. 1984); Pachesky v. Getz, 510 A.2d 776 (Pa. Super. Ct. 1986); Oullette v. Carde, 612 A.2d 687 (R.I. 1992).

n17. See Oullette, 612 A.2d at 690. The Supreme Court of Rhode Island stated:

We are of the opinion, however, that the comparative-negligence doctrine does not fully protect the rescue doctrine's underlying policy of promoting rescue. No common-law duties changed as a result of the enactment of Rhode Island's comparative-negligence statute, and there is nothing other than an individual's moral conscience to induce a person under no legal duty to undertake a rescue attempt. The law places a premium on human life, and one who voluntarily attempts to save a life of another should not be barred from complete recovery. Only if a person is rash or reckless in the rescue attempt should recovery be limited; accordingly we hold that the rescue doctrine survives the adoption of the comparative-negligence statute and that principles of comparative negligence apply only if a defendant establishes that the rescuer's actions were rash or reckless.

n18. See, e.g., Rodney Foo, Teenager Saved from Rampage, Police Say, San Jose Mercury News, Mar. 22, 2000, at 1B-2B (reporting on an incident in which a man rescued a teenage girl from an attacker by wrestling him to the ground and disarming him).

n19. See Dressler, supra note 1, at 972 n.10 & 983 n.53. By definition, it is difficult to identify situations where a rescue did not occur, since there is no reason for those on the scene to publicize their behavior. Occasionally, as with the Genovese case and the event that spurred this symposium, others do learn about such conduct. There is no way to know how many of these go unreported. On the other hand, some who address matters of decent behavior do espouse the view that we are a nation of rescuers. See Ann Landers, S.F. Examiner, Feb. 17, 2000, at B9 (responding to a letter about strangers who saved the writer's husband who had suffered heat stroke and a heart attack while jogging: "I never cease to be amazed at the number of good Samaritans out there"). Media coverage of non-rescues has also led to legislation in Massachusetts, Minnesota, Rhode Island, and Nevada. See Jay Silver, The Duty to Rescue: A Reexamination and Proposal, 26 Wm. & Mary L. Rev. 423, 423 & n.3 (1985); see also Dressler, supra note 1, at 973-74 n.15 (discussing proposed legislation after the David Cash incident).

n20. See, e.g., Sharyn Obsatz, Teen and Her Brother Aid Flood Victims with $ 1,000, The Press-Enterprise (Riverside, Cal.), Mar. 22, 2000, at B2.

n21. See infra notes 104-07 and accompanying text.

n22. See Frank B. Mapel, III & Charles J. Weigel, II, Good Samaritan Laws - Who Needs Them?: The Current State of Good Samaritan Protection in the United States, 21 So. Tex. L.J. 327, 327 (1980).

n23. See id. at 331-38 (summarizing who is covered under the various statutes).

n24. See id. at 342-46 (summarizing the standard of care required to receive immunity under the various statutes).

n25. For a tracing of the history of the Vermont legislation, see Marc A. Franklin, Vermont Requires Rescue: A Comment, 25 Stan. L. Rev. 51 (1972).

n26. See, e.g., Mass. Gen. Laws Ann. ch. 71, 55A (West 1996) (protecting school personnel who aid students); id. ch. 94, 305D (restaurant employees who aid choking patrons); id. ch. 111C, 14 (emergency medical technicians, police officers, and firefighters who aid anyone); id. ch. 112, 12B (medical professionals who volunteer emergency treatment); id. ch. 112, 12C (physicians and nurses who administer immunizations).

n27. See, e.g., Mass. Gen. Laws Ann. ch. 71, 55A applies to school personnel who help students, and Mass. Gen. Laws Ann. ch. 94, 305D does not apply to restaurant employees who aid noncustomers.

n28. See, e.g., Mass. Gen. Laws Ann. ch. 112, 12V (applying to persons trained in CPR who administer CPR); Cal. Civ. Code 1714.21 (West 1999) (applying to persons trained in CPR or use of defibrillator who administer CPR or use a defibrillator in an emergency).

n29. Marian Burros, Eating Well, N.Y. Times, Dec. 11, 1996, at C8.

n30. 42 U.S.C.A. 1791(c)(1) (West 1999).

n31. See id. 1791(c)(2). California has sought to encourage food donation in a different way. The donor of food that "is fit for human consumption at the time it was donated to a nonprofit charitable organization" is protected from liability "except for injury resulting from negligence or a willful act." Cal. Civ. Code 1714.25(a) (West 1998). This protection applies "regardless of compliance with any laws, regulations or ordinances regulating the packaging or labeling of food...." Id. The nonprofit organization, if it "in good faith, receives and distributes food without charge that is fit for human consumption at the time it was distributed is not liable for an injury or death due to the food unless the injury or death is a direct result of the negligence, recklessness, or intentional misconduct of the organization." Id. 1714.25(b).

n32. See generally Robert D. Sack, Sack on Defamation (3d ed. 1999).

n33. See id. 9.2.3.

n34. See id. 9.2.5.

n35. See id.

n36. Robert S. Adler & Ellen R. Peirce, Encouraging Employers to Abandon Their "No Comment" Policies Regarding Job References: A Reform Proposal, 53 Wash. & Lee L. Rev. 1381, 1383 (1996).

n37. See id. Note that the employer's desire to avoid libel claims by former employees is reinforced by the desire to avoid suits brought by those hurt as the result of an inaccurate reference. This may include those who directly rely on the defendant's reference, as well as those hurt by the reliance of others.

In Randi W. v. Muroc Joint Unified School District, 929 P.2d 582 (Cal. 1997), for example, three defendant school districts placed in a central placement file their reports on the same former employee. See id. at 585-86. The reports consistently spoke highly of the employee and none of the three mentioned that during his stay in that district there had been charges or complaints of sexual misconduct and impropriety with female students. See id. A fourth school district hired the man in question, who then sexually assaulted the plaintiff-student in that district. See id. In the resulting suit, a unanimous court concluded that, although the referring districts had no duty whatever to say anything about the employee, once they undertook to do so they had a duty to exercise reasonable care not to mislead. See id. Although defendants argued that no such duty arose here because they had said nothing about sexual behavior in their reports, the court thought that the generally affirmative comments about the employee might reasonably lead a reader to conclude that the writer knew nothing that might discredit the employee. See id.

These concerns, which have not been much affected by the constitutionalization, may be allayed only through government incentives or compulsion combined with protection. As usual, the existence and nature of an incentive will depend on the government's assessment of the importance of the desired conduct and what it will take to obtain that conduct.

The court in Randi W. asserted in dictum that the defendants were free to say nothing. See id. at 589. On the other hand, the court in Tarasoff v. Regents of University of California, 551 P.2d 334 (Cal. 1976), hinted that it might have been willing to impose a duty based on foreseeability alone if it had not found an alternate basis. See id. at 342-43. Such an approach would assuredly result in imposition of a duty to rescue.

n38. See W. Page Keeton et al., Prosser and Keeton on The Law of Torts 107 (5th ed. 1984) [hereinafter Prosser & Keeton]. In determining whether to impose a duty, courts balance a number of considerations. See, e.g., J.S. v. R.T.H., 714 A.2d 924 (N.J. 1998). In J.S., the New Jersey Supreme Court stated:

In determining whether a duty is to be imposed, courts must engage in a rather complex analysis that weighs and balances several, related factors, including the nature of the underlying risk of harm, that is, its foreseeability and severity, the opportunity and ability to exercise care to prevent the harm, the comparative interests of, and the relationships between or among, the parties, and, ultimately, based on considerations of public policy and fairness, the societal interest in the proposed solution.
Id. at 928. Similarly, the California Supreme Court, in Rowland v. Christian, 443 P.2d 561 (Cal. 1968), noted that determination of duty

involves the balancing of a number of considerations; the major ones are the foreseeability of harm to the plaintiff, the degree of certainty that the plaintiff suffered injury, the closeness of the connection between the defendant's conduct and the injury suffered, the moral blame attached to the defendant's conduct, the policy of preventing future harm, the extent of the burden to the defendant and consequences to the community of imposing a duty to exercise care with resulting liability for breach, and the availability, cost, and prevalence of insurance for the risk involved.
Id. at 564.

n39. This is a question for criminal law as well. Cf. United States v. Hollingsworth, 27 F.3d 1196, 1203 (7th Cir. 1994) (en banc) ("The proper use of the criminal law in a society such as ours is to prevent harmful conduct for the protection of the law abiding, rather than to purify thoughts and perfect character."); see also infra note 90.

n40. Indeed, we believe that the courts should continue their search for special relationships that justify imposing duties to rescue or aid.

n41. Note that some of the arguments in this section may argue against both civil and criminal duties.

n42. See infra note 56 and accompanying text.

n43. See Dressler, supra note 1, at 983 n.54; Melody J. Stewart, How Making the Failure to Assist Illegal Fails to Assist: An Observation of Expanding Criminal Omission Liability, 25 Am. J. Crim. L. 385, 431-32 (1998) (reviewing empirical studies documenting the social inhibition phenomenon).

n44. Vt. Stat. Ann. tit. 12, 519(a) (1973). See also Minn. Stat. 640A.01(1) (1996) ("A person at the scene of an emergency who knows that another person is exposed to or has suffered grave physical harm shall, to the extent that the person can do so without danger or peril to self or others, give reasonable assistance to the exposed person."); R.I. Gen. Laws 11-56-1 (1994) ("Any person at the scene of an emergency who knows that another person is exposed to, or has suffered, grave physical harm shall, to the extent that he or she can do so without danger or peril to himself or herself or to others, give reasonable assistance ....").

n45. See supra note 1. Some proponents of a civil duty suggest additional limitations. See, e.g., Thomas C. Galligan, Jr., Aiding and Altruism: A Mythopsycholegal Analysis, 27 U. Mich. J.L. Reform 439, 519 (1994) (suggesting a scheme for evaluating a duty to rescue in which ease of rescue is only one of six factors). Is there any safe rescue these days other than pulling a baby from a puddle? Even if the rescue appears to pose no obvious dangers, it may still be dangerous. An apparently injured person lying on the side of the road could be a trick to lure the Good Samaritan into a criminal trap. Even if the person is truly injured, the situation may involve other hidden dangers. See, e.g., Walker killed on 101 had HIV, CHP says, S.F. Examiner, Mar. 6, 2000, at A3 ("The Highway Patrol is looking for Good Samaritans who may have been exposed to the AIDS virus when they performed mouth-to-mouth resuscitation and CPR on a man struck and killed by a car Saturday on U.S. 101 in Windsor."). Of course, the world is dangerous enough even when one is not involved in rescuing others. For example, Ennis Cosby was robbed and murdered when he stopped to change a flat tire on his own car. See B. Drummond Ayres Jr., Bill Cosby's Son Is Slain Along Freeway, N.Y. Times, Jan. 17, 1997, at A14.

n46. See Richard A. Epstein, A Theory of Strict Liability, 2 J. Legal Stud. 151, 197-204 (1973); Dressler, supra note 1, at 987-88 n.73. This area is explored at length by Peter Singer in Peter Singer, Practical Ethics (2d ed. 1999). Singer explores the situation of a man who "has invested most of his savings in a very rare and valuable old Bugatti, which he has not been able to insure." Id. The car will "allow him to live comfortably after retirement." Id. The man is then confronted with a situation in which a boy will be killed on a railroad track unless the man throws a switch that will veer the train onto a siding where it will destroy the Bugatti. See generally Peter Berkowitz, Other People's Mothers, The New Republic, Jan. 10, 2000, at 27 (reviewing two Singer books and discussing this example). Consider also the argument of opponents of Proposition 19 on the March, 2000 California ballot. The claim was that if transit police were treated as other police, transit police would be empowered to order citizens to join a posse to chase a fugitive. This potential danger was enough to lead libertarians to oppose the proposition, whose main purpose was to increase the penalty for killing transit police. See California Secretary of State, California Voter Information Guide: March 7, 2000 Primary Election, at 39 (visited May 2, 2000) <<uscore>bp.htm> (discussing the statement of Gail K. Lightfoot, past chair of the Libertarian Party of California, and others in opposition to Proposition 19).

n47. See Robert M. Ackerman, Tort Law and Communitarianism: Where Rights Meet Responsibilities, 30 Wake Forest L. Rev. 649, 660-61 (1995); see also Yeager, supra note 1, at 38-54.

n48. See Epstein, supra note 46, at 200. Professor Anthony D'Amato recognizes that making an act criminal destroys the altruism argument:

Once a law is passed, S will not be acting out of altruism but out of his own self-interest in avoiding the criminal sanctions of the law. Thus, the law has arguably destroyed the basis for altruistic behavior by requiring it. However, the fact that a majority of the members of a state might find it in their self-interest to pass such legislation does not necessarily deprive any smaller class of people of the possibility of moral behavior. If the behavior is morally required, then it should not be significant that it be labeled "altruistic."
Anthony D'Amato, The "Bad Samaritan" Paradigm, 70 Nw. U. L. Rev. 798, 805 & n.30 (1976). Note that this also applies to a state in which civil law obligates rescue.

n49. Some are concerned that any obligation to rescue would deter law-abiders from going to beaches and other places where rescue situations are more likely to occur than to other (safer) places. See William M. Landes & Richard A. Posner, Salvors, Finders, Good Samaritans, and Other Rescuers: An Economic Study of Law and Altruism, 7 J. Legal Stud. 83, 119-24 (1978); see also Eugene Volokh, Duties to Rescue and the Anticooperative Effects of Law, 88 Geo. L.J. 105, 108 n.12 (1999).

n50. See supra note 19 and accompanying text.

n51. See, e.g., Abraham M. Rosenthal, Thirty-Eight Witnesses (1964) (describing how Kitty Genovese was repeatedly attacked and then murdered while her neighbors did nothing to help); Dressler, supra note 1, at 971-72 (discussing how Sherrice Iverson was raped and murdered in a Nevada casino bathroom by Jeremy Strohmeyer while his friend, David Cash, waited outside); Stewart, supra note 43, at 389-90 (relating how two men severely beat Joey Levick and left him to die in a drainage ditch and then told several people who did nothing to help Levick).

n52. Consider the potential liability in the Genovese situation if a witness had called the police and had then shouted out the window that he had made the call - to give courage to the victim and to try to get the attacker to depart. What if the attacker quickly killed the victim and ran away? Might the victim's estate argue that a jury should decide whether the caller was negligent to announce what he had done - that the victim would only have been wounded if he had not shouted to the murderer? Might a jury find that it was negligent to alert an attacker to the imminent arrival of the police?

n53. See James Barr Ames, Law and Morals, 22 Harv. L. Rev. 97 (1908); see also, e.g., Jain, supra note 1, at 1200 (relating examples of Good Samaritans being victimized by criminals feigning injury).

n54. The ambiguity of the accident scene is a factor that a prosecutor should consider before charging a criminal failure to rescue - whether or not the statute explicitly addressed that issue. In civil litigation, this issue may be raised by pretrial motion but deferred until a trial at which the jury will be free to conclude that there was no risk. In addition, as noted later, see infra note 82, the civil case will lack most - if not all - of the other criminal law protections, such as the judicial tendency to narrowly construe criminal statutes, the burden of proof, and the common requirement for a unanimous verdict.

n55. See Minn. Stat. 604A.01 & 609.02 (1996) (imposing a maximum fine of $ 200); R.I. Gen. Laws 11-56-1 (1994) (imposing a maximum of six months in jail, $ 500 fine, or both); Vt. Stat. Ann. tit. 12, 519(c) (1973) (imposing a maximum fine of $ 100).

n56. See, e.g., BMW of North Amer., Inc. v. Gore, 517 U.S. 559 (1996) (contrasting punitive damage award of $ 2,000,000 with the state's $ 2,000 maximum fine for the same conduct); New York Times Co. v. Sullivan, 376 U.S. 254 (1964) (contrasting civil award of $ 500,000 with maximum criminal sanction of fine of up to $ 500 and prison sentence of six months). We strongly suspect that citizens in Vermont who failed to rescue would prefer, if given a choice, to be prosecuted under the criminal law and pay a fine of $ 100 than to be exposed to tort liability for tens or hundreds of thousands of dollars. We are not proposing that law violators should be given their choice; we are simply observing that whatever generic attitudes we hold toward the relationship between the two systems almost surely do not apply in the rescue situation.

n57. The usual claim that even a judgment-proof person has incentives to act safely because of physical dangers, as in driving or crossing the street, is unlikely to apply in rescue cases.

n58. We refer readers to Professor Dressler's article for the facts of the actual case. See Dressler, supra note 1, at 972-73.

n59. Courts have allowed recovery for pre-death emotional distress. See, e.g., Shu-Tao Lin v. McDonnell Douglas Corp., 742 F.2d 45 (2d Cir. 1984) (allowing recovery for plane passenger's pre-impact fright); Solomon v. Warren, 540 F.2d 777 (5th Cir. 1976) (allowing recovery for passenger in plane that was running low on fuel over the open sea and was never seen again), cert. dismissed, 434 U.S. 801 (1977); Nelson v. Dolan, 434 N.W.2d 25 (Neb. 1989) (permitting recovery from pre-impact fright in a car-motorcycle accident).

n60. See, e.g., Gibson v. Gibson, 479 P.2d 648 (Cal. 1971). To avoid complications of apportioning liability in some death cases, if a prospective beneficiary is negligent we assume that only the non-negligent parent sues.

n61. See, e.g., Restatement (Third) of Torts: Apportionment of Liability (Proposed Final Draft (Revised) 1999).

n62. That is, if we assume that Cash was not in any danger from an angry Strohmeyer. See Dressler, supra note 1, at 971 n.1.

n63. See Marc A. Franklin & Robert L. Rabin, Tort Law and Its Alternatives 325 (6th ed. 1996).

n64. See Cal. Civ. Code 1430-32 (West 2000).

n65. Further, there are subtle issues of causation with such a suit. Are we prepared to conclude that all four actors "caused" the harm? This problem is not unique to tort law. But the civil focus is on the damages allegedly caused by the inaction. Again, although this is not the central argument against tort liability, it is important to note that causation here is more difficult to find than when defendant's car runs into plaintiff's. See Volokh, supra note 49, at 106 n.6.

n66. See D'Amato, supra note 48, at 802.

n67. Compare Whitehead v. Linkous, 404 So. 2d 377 (Fla. Dist. Ct. App. 1981), with Whitehead v. Toyota Motor Corp., 897 S.W.2d 684 (Tenn. 1995).

n68. See, e.g., Great Northern Ry. Co. v. Sunburst Oil & Ref. Co., 287 U.S. 358 (1932); Kelly v. Gwinnell, 476 A.2d 1219 (N.J. 1984); see also Walter V. Schaefer, The Control of "Sunbursts": Techniques of Prospective Overruling, 42 N.Y.U. L. Rev. 631, 631-41 (1967); Michael Treanor & Gene B. Sperling, Prospective Overruling and the Revival of "Unconstitutional" Statutes, 93 Colum. L. Rev. 1902 (1993).

n69. Further, such judicial action would not be a case of "overruling" because there is no prior decision to overrule. Yet, it is a situation in which the universal understanding was one way and the court is now preparing to move the other way. One possible solution is a prospective civil statute. See infra note 141 and accompanying text.

n70. Compare the unwillingness of most courts to reject the doctrine of contributory negligence, even where there had been no judicial decision explicitly adopting it. See Symposium, Comments on Maki v. Frelk - Comparative v. Contributory Negligence: Should the Court of Legislature Decide?, 21 Vand. L. Rev. 889 (1968) (including comments by Professors James, Kalven, Keeton, Leflar, Malone, and Wade on the Illinois Supreme Court's reversal of the Illinois Court of Appeals' judicial adoption of comparative negligence); see also, e.g., Williams v. Delta Intern. Mach. Corp., 619 So. 2d 1330 (Ala. 1993). But see Davenport v. Cotton Hope Plantation Horizontal Property Regime, 508 S.E.2d 565 (S.C. 1998). The comparative negligence situation is not as strong as our situation because the defendant in that case by hypothesis had violated a duty of care and the judicial overruling, though a surprise to the defendant, could hardly be thought to be an unfair surprise.

n71. See, e.g., Tarasoff v. Regents of the Univ. of Cal., 551 P.2d 334 (Cal.1976).

n72. A similar concern exists in criminal law. See Dressler, supra note 1.

n73. To the extent that criminal conduct created the danger, recall that the claim for compensation is even more attenuated, though not necessarily extinguished. See Hutcherson v. City of Phoenix, 961 P.2d 449 (Ariz. 1998); see also supra note 69 and accompanying text. Indeed, the criminal law seems so much the better candidate for imposition of a duty to rescue that one might think the only reason proponents of a civil duty have turned to tort law is that criminal law usually requires prospective legislative action and tort law can be retroactively created by judges.

n74. Although earlier we recognized that some may find incentives inherently inadequate, see supra text following note 37, we do not share that view. If one finds current incentives inadequate we believe the first recourse should be to consider other incentives. One candidate for added incentives might be changing the common law rule so that those who attempt rescues that are not required are held only to a good faith standard rather than the due care standard now in place. See Restatement (Second) of Torts 324 (1979). Saul Levmore argues that courts hold the person who initially attempts to aid but withdraws before completing the rescue liable not because he has made matters worse, but because he can be identified. See Saul Levmore, Waiting for Rescue: An Essay on the Evolution and Incentive of the Law of Affirmative Action, 72 Va. L. Rev. 879, 937-38 (1986). There are, of course, arguments against changing the common law rule, such as concern about overeager attention seekers who see the need to rescue though reasonable observers would not, or who cause harm because of their unusually clumsy rescue efforts. Our point is that discussion of new incentives should precede discussion of coercion.

Of course, for an incentive to have any effect on behavior, bystanders must be aware of the legal rules. There may be, however, some evidence that people are generally aware of the legal rules regarding rescue. See Hans Zeisel, An International Experiment on the Effects of a Good Samaritan Law, in The Good Samaritan and the Law 209, 210 (James M. Ratcliffe ed., 1966) (reporting the results of study comparing surveys of university students conducted in Germany, which has a Good Samaritan law, and the United States and Austria, which do not).

n75. See Minn. Stat. 604A.01 (1996); R.I. Gen. Laws 11-56-1 (1994); Vt. Stat. Ann. tit. 12, 519 (1973).

n76. See, e.g., R.I. Gen. Laws 11-56-1.

n77. See supra note 44.

n78. Proponents of a civil duty to rescue might rely on section 874A of the Restatement (Second) of Torts, which provides:

When a legislative provision protects a class of persons by proscribing or requiring certain conduct but does not provide a civil remedy for the violation, the court may, if it determines that the remedy is appropriate in furtherance of the purpose of the legislation and needed to assure the effectiveness of the provision, accord to an injured member of the class a right of action, using a suitable existing tort action or a new cause of action analogous to an existing tort action.
Restatement (Second) of Torts 874A.

As discussed earlier, however, a civil remedy may not further the goals of criminal duty to rescue legislation. See supra notes 55-57 and accompanying text. Further, a civil duty is not "needed to assure the effectiveness" of the criminal statute. See supra notes 50-54 and accompanying text. Lastly, those in peril may not be a recognizable class of persons. Section 874A is much more applicable to the child abuse reporting situation discussed later. See infra note 125.

n79. The impact of criminal legislation on tort duties has long been debated. Compare Ezra Ripley Thayer, Public Wrong and Private Actions, 27 Harv. L. Rev. 317 (1914) (should have no impact), with Clarence Morris & C. Robert Morris, Jr., Morris on Torts 146-51 (2d ed. 1980) (may be relevant).

n80. Courts may sometimes conclude that the overall legislative scheme is so intricate and balanced that it should be treated as a complete package, implicitly barring judicial imposition of additional duties. However, there is no such complex scheme here. But see infra note 82.

n81. This assumes that the criminal legislation was in effect before the failure to rescue that gave rise to the tort claim.

n82. In fact, in one sense it can be argued that the legislation works against the creation of a duty. Even though there is no elaborate legislative scheme that signals a complete package, some aspect of the legislation may induce judicial caution. The fact that the criminal statute sets the maximum penalty at a fine of $ 100 may be such a signal. See Vt. Stat. Ann. tit. 12, 519 (1973). A tort action for wrongful death or for injury aggravated by delayed rescue might lead to liability for tens or hundreds of thousands of dollars, or more. Since this would be a civil action it would lack basic protections that were implicit in the criminal law enactment: the requirements of proof beyond a reasonable doubt and a unanimous jury. A court should think twice about whether the legislature's decision to impose a minimal fine should influence it to create a tort duty that might impose massive civil liability. Perhaps the legislative goal was to change attitudes toward mutual respect and assistance in the community, but to impose only a symbolic sanction - at least until lingering public doubts pass. A rational court might conclude that the criminal sanction argues against adding a civil duty because the criminal sanction already reflects an appropriate "total" governmental sanction.

Furthermore, Professor D'Amato worries that after criminal law has set the appropriate sanction the civil liability may totally upset that determination: the combination of criminal and tort sanctions where both can be effective might be heavier than the society's judgment of what is warranted. See D'Amato, supra note 48, at 810-11. The preemption argument discussed earlier, see supra note 80, is that the legislature wanted no more liability. The argument here is that the court itself concludes that the total package of criminal and civil sanctions would be excessive.

n83. See Prosser & Keeton, supra note 38, 36.

n84. For example, the crimes of murder and rape implicate the tort of battery, arson implicates trespass to land, and theft implicates trespass to chattels.

n85. See Haw. Rev. Stat. 663-1.6 (1993).

n86. See D'Amato, supra note 48, at 810-12 (advocating "the adoption of a Vermont-type statute coupled with a statutory provision explicitly prohibiting any private action based upon violation of the statutory standard").

n87. See infra note 90 and accompanying text.

n88. The wisdom of criminal liability in these situations is beyond the scope of this essay and not relevant to what follows.

n89. See D'Amato, supra note 48, at 808.

n90. Note that only someone who has opted to impose criminal liability can use as an argument against civil liability the point that the defendant will be punished under criminal law without giving any award to the plaintiff. We find the fear of encouraging risky behavior much weakened by the fact that Professor D'Amato would recognize a criminal duty to rescue no matter how the victim got into the perilous situation, and has noted that criminal law may be more effective there than civil law. As Dressler notes, however, the reach of Anglo-American criminal law is limited: "The penal law does not seek to punish every morally bad act that we commit." Dressler, supra note 1, at 987. This limitation also applies to tort law if the goal is not compensation, but rather changing behavior.

n91. D'Amato, supra note 48, at 808; see also supra note 49.

n92. See D'Amato, supra note 48, at 810. Professor Dressler is concerned about giving prosecutors this discretion. See Dressler, supra note 1, at 983. Nonetheless, that discretion is far more likely to be protective than its counterpart in the civil area.

n93. See Spickler v. Greenberg, 644 A.2d 469, 470 (Me. 1994) (noting that Maine is the only state that recognizes a civil action for perjury).

n94. See generally Franklin v. Terr, 201 F.3d 1098 (9th Cir. 2000) (extending this protection to a 1983 claim that a psychiatrist who testified at a trial on induced memory conspired with other trial witnesses to assure that their testimony was consistent); Neidert v. Rieger, 200 F.3d 522 (7th Cir. 1999).

n95. See Sack, supra note 32, at 8-1.

n96. See Prosser & Keeton, supra note 38, 74.

n97. See 1 Ronald E. Mallen & Jeffrey M. Smith, Legal Malpractice 7.12-13 (4th ed. 1996).

n98. Mitchell v. Chapman, 10 S.W.3d 810 (Tex. App. 2000).

n99. See Barker v. Kallash, 468 N.E.2d 39 (N.Y. 1984). But see Ashmore v. Cleanweld Prods., Inc., 672 P.2d 1230 (Or. Ct. App. 1983).

n100. See Victor E. Schwartz, Comparative Negligence 5-4 (3d ed. 1994); see also, e.g., Lewis v. Miller, 543 A.2d 590 (Pa. Super. Ct. 1988).

n101. But this will not explain cases in which the defendant is only negligent and thus not deterred at all in a case involving a reckless plaintiff. States utilizing modified comparative negligence bar these suits, but there is no criminal deterrent. Denial is based on a rejection of the fairness of comparative negligence where the plaintiff is more negligent or reckless than the defendant.

A well-known hypothetical illustrates this point in an even more clearly criminal context. A person who has stolen asks a friend to chop off the offending hand. The friend complies - and is sued. Professor Morris argued that the plaintiff would be barred because of the consent (not the theft). See Morris, supra note 79, at 25. But he assumed that the defendant might still be criminally liable for mayhem despite the consent. See id.

n102. See Wayne R. LaFave & Austin W. Scott, Jr., Substantive Criminal Law 1.3 (1986).

n103. See supra notes 90-91 and accompanying text.

n104. See Vt. Stat. Ann. tit. 12, 519(b) (1973).

n105. See Volokh, supra note 49, at 105-09 (describing a typology of Samaritans, including the "Legally Swayable Samaritan" who "would be bad in the absence of the duty-to-rescue/report law, but would be swayed by such a law's coercive or normative effect").

n106. See Prosser & Keeton, supra note 38, 32.

n107. See Minn. Stat. 640A.01(1) (1996); R.I. Gen. Laws 11-56-1 (1994); Vt. Stat. Ann. tit. 12, 519(b). But see D'Amato, supra note 48, at 811 n.47 (arguing that if the criminal law should require rescue, "one sees no reason why a reasonable man standard cannot be applied to the rescue attempt. Why should the law encourage bumbling or negligent rescues once the duty has devolved upon everyone to render assistance to an imperiled victim?").

n108. See Franklin, supra note 25, at 53-54.

n109. On the other hand, the existence of section 519(b) does answer at least one additional question: when a volunteer undertakes a rescue and is thereafter sued, what standard should be used? In that case, it would be hard to deny any rescuer in Vermont the benefit of the "gross negligence" standard. If the duty to rescue did not apply (because the person had an important duty with which the rescue would have interfered or because there was risk involved), but the rescuer decided nonetheless to try the rescue, it would be implausible to make such a person worse off than a person who was in fact obligated by section 519(a) to attempt the rescue. Although the legislature extended this protection to anyone "acting in compliance with section (a)," Vt. Stat. Ann. tit. 12, 519(b), it would be unreasonable not to protect to (at least) the same extent those who sought to rescue without being obliged to do so by the statute.

n110. Curiously, in at least one situation the existence of a criminal statute - with or without civil provisions - might make it more difficult for a court to impose a civil duty if the criminal law would not have required rescue. If the criminal statute tells citizens that no rescue is required in a particular situation - because of a risk or because of an overriding important duty - a judicial decision retroactively imposing a civil duty to rescue in these circumstances might be thought to unfairly surprise citizens. In this sense, the enactment of limited criminal liability may properly inhibit the imposition of a civil duty. That is, if the Las Vegas events had occurred in Vermont, but the statute would not have covered David Cash (because, for example, he showed that he risked harm from Jeremy Strohmeyer if he intervened), it might be harder for a Vermont court inclined to create a civil duty to do so than if the state had no criminal statute at all.

n111. See supra note 55 and accompanying text.

n112. See, e.g., Fla. Stat. ch. 794.027 (1993) (duty to report sexual battery); Haw. Rev. Stat. 663-1.6 (1993) (duty to obtain aid for victims of crime); Mass. Gen. Laws Ann. ch. 268, 40 (West 1990) (duty to report crimes to law enforcement officials); R.I. Gen. Laws 11-1-5.1 (1994) (duty to report crimes to law enforcement officials); R.I. Gen. Laws 11-37-3.1 (duty to report sexual assault); Wash. Rev. Code 9.69.100 (1998) (duty to report offense against child or any violent offense); Wis. Stat. 940.34 (1995-96) (duty to aid victim or report crime). Note that victims of an act of God or persons who put themselves in peril would not receive the benefit of these statutes. That is, a witness would not be required to call for help after seeing a baby drowning in six inches of water. If the witness saw an adult put the baby in the water would a call to 911 suffice?

n113. The statutes appear to be geared more toward thwarting the crime and aiding the victim rather than catching the perpetrator. The statutes mandate reporting where the crime involves a victim who is suffering from or is in danger of physical harm. If the sole purpose of these statutes were law enforcement, then there would be no such limitation. Some statutes are explicit in their purpose, see, e.g., Fla. Stat. ch. 794.027(2) (requiring witnesses to "seek assistance for the victim or victims"); Haw. Rev. Stat. 663-1.6(a) ("Any person at the scene of a crime who knows that a victim of the crime is suffering from serious physical harm shall obtain or attempt to obtain aid from law enforcement or medical personnel."), while in other statutes the purpose is implicit, see, e.g., Mass. Gen. Laws Ann. ch. 268, 40 (requiring reporting to law enforcement officials without explicitly stating that the goal is to aid the victim). In either case, however, the goal of aiding the victim is apparent from the language of the statute. Most of the statutes apply only to those who witness the commission of the crime. See, e.g., Fla. Stat. ch. 794.027; Haw. Rev. Stat. 663-1.6; Mass. Gen. Laws Ann. ch. 268, 40; R.I. Gen. Laws 11-1-5.1; Wash. Rev. Code 9.69.100(1). The statutes also generally contain a time frame component: witnesses are required to report crimes that are in progress and to summon aid for victims who are suffering. See, e.g., Haw. Rev. Stat. 663-1.6(a) (noting that the statute applies when the victim "is suffering"); R.I. Gen. Laws 11-37-3.1 (describing that the duty applies when the crime "is taking place"); Wis. Stat. 940.34(2)(a) (addressing situations when a "crime is being committed and a victim is exposed to bodily harm"). Furthermore, the statutes generally require that the witness report the crime or otherwise summon aid immediately. See, e.g., Fla. Stat. ch. 794.027(2) ("immediately reporting"); Mass. Gen. Laws Ann. ch. 268, 40 ("as soon as reasonably practicable"); R.I. Gen. Laws 11-1-5.1 ("as soon as reasonably practicable"); R.I. Gen. Laws 11-37-3.1 ("immediately notify"); Wash. Rev. Code 9.69.100(1)(c) ("as soon as reasonably possible").

The Nevada statute passed in the aftermath of Sherrice Iverson's death, seems to be less victim-focused than the others. See Nev. Rev. Stat. Ann. 202.882 (Michie 1999). The statute requires the witness to report a crime that she knows or believes that another person "has committed," not limiting its application to an ongoing offense. See id. Furthermore, although the witness is required to report the crime "as soon as reasonably practicable," the statute sets an outer deadline of no more than 24 hours after the assault. See id. This provision further suggests that the attack has been completed and that reporting it as required under the statute may help law enforcement personnel apprehend the offender, but does nothing to help the victim. Of course, the fact that the statute is limited to crimes involving minors may imply some victim focus.

n114. For example, the statutes do not require that witnesses yell (perhaps from a safe distance) to the perpetrators to stop, or to shout that they have called the police, thereby alerting the perpetrators that they are not alone and may be apprehended if they continue. See supra note 112.

n115. Most statutes specify reporting the crime to law enforcement officials, but at least one (Hawaii) also specifies medical personnel. See Haw. Rev. Stat. 663-1.6(a). Wisconsin requires that the witnesses either provide aid themselves or call law enforcement officials for help. See Wis. Stat. 940.34(2)(a). This, of course, does not require direct intervention, but rather allows it as an alternative to calling the police for help. In many cases, however, the action required by general duty-to-aid statutes (like Vermont's) and duty to summon aid for victims of crime statutes may in fact be the same. If it would be too dangerous to attempt to intervene directly to aid a victim of crime, "reasonable assistance" may require only calling for help. This suggests an interesting point about the other statutes: if a witness chose to intervene directly instead of calling for help, has he or she violated the statute? There may be good reasons to discourage direct intervention by witnesses who may themselves be injured.

n116. Recall that this is Professor D'Amato's proposition: a criminal statute that explicitly rejects civil liability. See supra note 86 and accompanying text.

n117. Hawaii and Wisconsin, however, are exceptions. See Haw. Rev. Stat. 663-1.6(b) & (c) (providing immunity from civil liability for both those who comply with the statute and those who do not); Wis. Stat. 940.34(3) (providing immunity for those who comply with the statute).

n118. Given this lack of need, section 874A of the Restatement (Second) of Torts provides little support for a civil duty to report crime. See supra note 78.

n119. See supra note 113.

n120. Many questions remain unanswered in this situation. Was the victim still alive by time the defendant should have called? How many other emergencies were arising at around the same time? How would the 911 operators prioritize the call? What were the traffic conditions from the police station to the site? Would the victim still be alive and in a survivable state when the aid should have arrived? On the uncertainties of calling 911, see Hutcherson v. City of Phoenix, 961 P.2d 449 (Ariz. 1998); Stewart, supra note 43, at 420; Jain, supra note 1, at 1195 n.131.

n121. See Caroline T. Trost, Note, Chilling Child Abuse Reporting: Rethinking the CAPTA Amendments, 51 Vand. L. Rev. 183, 194-95 n.63 (1998) (collecting statutes).

n122. See, e.g., Tenn. Code Ann. 37-1-403(a) (1996) (imposing a duty to report on "any person having knowledge of or called upon to render aid to any child who is suffering from [any injury] which is of such a nature as to reasonably indicate that it has been caused by brutality, abuse or neglect").

n123. See, e.g., Mo. Rev. Stat. 210.115 (1986). The Missouri statute imposes a duty to report on the following persons:

Any physician, medical examiner, coroner, dentist, chiropractor, optometrist, podiatrist, resident, intern, nurse, hospital or clinic personnel that are engaged in the examination, care, treatment, or research of persons, and any other health practitioner, psychologist, mental health professional, social worker, day care center worker or other child care worker, juvenile officer, probation or parole officer, teacher, principal, or other school official, or other person with responsibility for the care of children [who] has reasonable cause to suspect that a child has been or may be subjected to abuse or neglect or observes a child being subjected to conditions or circumstances which would reasonably result in abuse or neglect....

n124. See Trost, supra note 121, at 196 n.73 ("Every state except California, Alabama, Ohio, and possibly New Jersey, which provide "absolute immunity,' has chosen to implement "good faith' reporting immunity provisions.").

n125. At least seven state statutes mandate civil liability for failure to report child abuse. See Ark. Code Ann. 12-12-504(b) (Michie Supp. 1993); Colo. Rev. Stat. 19-3-304(4)(b) (Supp. 1994); Iowa Code Ann. 232.75(2) (West 1994); Mich. Comp. Laws Ann. 722.633(1) (West Supp. 1994); Mont. Code Ann. 41-3-207(1) (1993); N.Y. Soc. Serv. Law 420 (McKinney 1992); R.I. Gen. Laws 40-11-6.1 (1990).

n126. See Ham v. Hospital of Morristown, 917 F. Supp. 531, 534-37 (E.D. Tenn. 1995) (holding that, although Tennessee does not impose a common law duty to report or prevent suspected child abuse, Tennessee's reporting statute creates a private right of action based on the physician's statutory duty to report); cf. Sabia v. State, 669 A.2d 1187, 1193 (Vt. 1995) (alternative holding imposing a duty in a case involving a claim of inaction by a state agency after it received a report of child abuse, using the Vermont duty-to-aid statute by analogy).

n127. J.S. v. R.T.H., 714 A.2d 924 (N.J. 1998).

n128. See id. at 935; see also First Commercial Trust Co. v. Rank, 915 S.W.2d 262 (Ark. 1996) (holding a physician's duty arises out of the Medial Malpractice Act, which "encompasses a cause of action for failure to diagnose child abuse"); Eric J. v. Betty M., 90 Cal. Rptr. 2d 549, 551 n.2 (Ct. App. 1999) (court, holding that family members had no duty to warn others of danger of convicted child molester, noted that molester's mother, who had once left her son and a young boy alone in her house, had been held liable below and did not appeal).

n129. See J.S., 714 A.2d at 933 (citing Restatement (Second) of Torts 874A (1979)). Note that the question before the J.S. court was whether to create a tort duty, using the statute or otherwise. This is different from the situation in which a court uses the statutory criminal duty to find breach once a duty has been established. See, e.g., Bentley v. Carroll, 734 A.2d 697 (Md. 1999) (holding that where duty existed due to a doctor-patient relationship, failing to report child abuse in violation of the child abuse reporting statute constitutes evidence of negligence).

n130. See J.S., 714 A.2d at 930-31, 933.

n131. Id. at 933.

n132. Id.

n133. See, e.g., Isley v. Capuchin Province, 880 F. Supp. 1138 (E.D. Mich. 1995); Letlow v. Evans, 857 F. Supp. 676 (W.D. Mo. 1994); Thelma D. v. Board of Educ., 669 F. Supp. 947 (E.D. Mo. 1987); Doe "A" v. Special Sch. Dist. of St. Louis County, 637 F. Supp. 1138 (E.D. Mo. 1987); Eric J., 90 Cal. Rptr. 2d at 549; Freehauf v. School Bd. of Seminole County, 623 So. 2d 761 (Fla. Dist. Ct. App. 1993); Fischer v. Metcalf, 543 So. 2d 785 (Fla. Dist. Ct. App. 1989); Vance v. T.R.C., 494 S.E.2d 714 (Ga. 1998); J.A.W. v. Roberts, 627 N.E.2d 802 (Ind. 1990); Borne v. Northwest Allen County Sch. Corp., 532 N.E.2d 1196 (Ind. Ct. App. 1989); see also Perry v. S.N., 973 S.W.2d 301 (Tex. 1998) (holding that the reporting statute did not create a civil cause of action but expressly noted that it was not addressing the question of a common law duty).

n134. See J.S., 714 A.2d at 933 (noting that "ninety-five percent to ninety-eight percent of child sexual abuse "is hidden behind closed doors'").

n135. See id.; see also Trost, supra note 121, at 214-215.

n136. See supra notes 121-28 and accompanying text. It seems implicit that a court creating such a duty without an existing statute of any sort should offer a comparably lenient standard. In a voluntary system, a bystander who is unsure about his ability to acquit himself reasonably in a rescue attempt has the option of simply doing nothing, thereby not exposing himself to liability. The immunities that the statutes provide attempt to encourage the bystander to get involved by reducing fears that he may be exposing himself to liability. In a mandatory system, however, that bystander cannot simply sit this one out; he must get involved. To the extent that many of the child abuse statutes apply only to individuals who have a special relationship with the child abuse victim, those individuals may already have a civil duty regardless of the statute.

n137. Proving causation, however, still requires speculation about how the report would have been processed and what intervention, if any, it would have spurred.

n138. There are, however, still some concerns that militate against finding such a civil duty. As discussed above, see supra notes 55-57 and accompanying text, imposing a civil duty to report child abuse exposes a person to potentially ruinous civil liability for failing to report. The penalties imposed by the criminal statutes are mild in comparison. Also, courts must always consider whether civil liability would impair what the legislature thought was the appropriate balance of duty and liability.

n139. Contrary to the duty-to-rescue or crime reporting contexts, section 874A of the Restatement (Second) of Torts seems particularly on point here. There is no doubt that requiring reports of child abuse is meant to protect "a class of persons" and that the age and dependence of that class and the secrecy of the criminal act suggest that tort law may be needed to further "the purposes of the legislation" and "to assure the effectiveness of the [criminal] provision." Compare supra note 78, with supra note 118.

n140. There are other examples of such mandates, but some are far afield from our question. Nonetheless, they do suggest that such coercion warrants some level of protection from civil liability. In California, the legislature has considered background checks on prospective peace officers to be so important that under specified conditions a former employer of an applicant "shall disclose employment information" when asked by an appropriate inquirer. Cal. Gov't Code 1031.1(a) (West 1995). An employer's refusal to comply "shall constitute grounds for a civil action for injunctive relief requiring disclosure on the part of an employer." Id. 1031.1(d). In return, the legislature granted the responding employer extensive protection against civil liability. See id. 1031.1(b) (granting immunity against civil liability "in the absence of fraud or malice"). A court has increased that protection to absolute immunity. See Bardin v. Lockheed Aeronautical Systems Co., 82 Cal. Rptr. 2d 726 (Ct. App. 1999). Here, coercing action that requires the revelation of information that the speaker may not wish to reveal - but compliance should not be onerous if the files and memories exist. For our purposes, the relevant point is that again the tradeoff for coercion is protection from some kinds of legal liability.

n141. As suggested earlier, see supra note 68-72 and accompanying text, if the state has no prior history of criminal or civil obligation to rescue, we believe that, at the least, given the long-established common-law no-duty rule, a court inclined to create such a duty should reject the traditional retrospective approach to tort cases. There is simply inadequate notice to law-abiding citizens that they risk the loss of all their assets if they refrain from an effort to rescue that a court later concludes should have produced a rescue attempt. The use of a prospective ruling can ameliorate this concern. This is not truly a case of "overruling" since by hypothesis the court has never ruled on the issue before. This situation differs from the case of someone who behaves badly and then argues that it is unfair to change the tort rule retroactively. In the perjury example, see supra notes 93-96 and accompanying text, if a court were to decide that perjurers should no longer be protected against civil liability, a defendant's reliance argument seems quite weak.

Prepared: April 16, 2003 - 5:02:29 PM
Edited and Updated, April 17, 2003

Back to
Noah Rachlin