Duties to Rescue and the Anticooperative Effects of Law

Prof. Eugene Volokh, UCLA Law School *


(88 Georgetown Law Journal 105 (1999))

              The Good Samaritan rescues the crime victim, or at least promptly reports the crime 1 to the police; the Bad Samaritan stands idly by.  Discussions about laws that require bystanders to help a crime victim or report the crime "to the extent that [they] can do so without danger" 2 generally divide people into these two groups.  Surely it would be good, the argument goes, to pressure Bad Samaritans into acting Good.

              But real people aren't so neatly divisible into these two internally homogeneous categories.  Rather, Samaritans come in at least five different stripes:

              The Good Samaritan helps the victim by calling the police.  He might even help by physically interceding while the crime is happening, but this is almost never required even under laws that require "assistance" and not just "report[ing]"; these laws explicitly do not mandate any intercession that would pose "danger or peril to [s]elf or others." 3  Duty-to-rescue and duty-to-report laws are thus interchangeable for my analysis.

              The Hopelessly Bad Samaritan refuses to help -- perhaps because of loyalty to the criminal, unreasonable fear of retaliation, or callousness coupled with a perceived improbability of being identified and prosecuted under the duty-to-rescue/report law -- and can't be budged from this by conscience, the threat of punishment, or the law's normative force. 4

              The Legally Swayable Samaritan would be Bad in the absence of the duty-to-rescue/report law, but would be swayed by such a law's normative or coercive effect.

              The Delayed Samaritan initially fails -- because of loyalty, panic, hurry, fear, or uncertainty -- to help or to report, but later changes his mind and wants to come forward with information about the incident, perhaps prompted by a day's contemplation or by hearing that the police are looking for evidence about the crime.

              The Passive Samaritan never calls the authorities, but when the police come to his door looking for witnesses, is willing -- because of remorse, a felt duty to answer questions, fear of lying to the police, or just the natural tendency to respond to questions asked by those in authority -- to tell them what he saw.

              This typology tells us two things.  First, and most obvious, duty-to-rescue/report laws by definition won't do much about the Hopelessly Bad Samaritan.  The laws will affect those who are Bad enough that they don't do the right thing on their own but are nonetheless so sensitive to the law's normative or coercive effect that they're Legally Swayable to being Good (or at least acting as the Good do).  Some people will fall into this category, but I doubt many will:  Those who don't respond to the social norm of helping those in distress -- at least by calling 911 --probably aren't that likely to be swayed by the normative effect of a new duty-to-rescue/report law; and the law's coercive force will usually be rather low because the witnesses know they're generally unlikely to be conclusively identified if they just stay quiet.  This suggests the laws will do relatively little good, something the laws' supporters don't deny; the laws, they often agree, will influence only a few people, but they argue that even this small benefit is justification enough. 5

              The trouble, though, is that as to the two other varieties of Samaritan -- the Delayed Samaritan and the Passive Samaritan, two groups usually forgotten in the public outrage at the genuinely Bad -- the laws may actually cause harm. 6  Imagine a man who sees a neighbor seriously abusing the neighbor's child.  The Good Samaritan would intervene or at least call the authorities, but some otherwise decent people fail to do this.

              Some might be afraid, reasonably or not, that the abusive parent will physically hurt them too if they intervene or even if they call the police.  Others may wrongly feel a sense of loyalty to their neighbor, or wrongly conclude it's none of their business.  Others may misperceive the magnitude of the abuse and hesitate to call the police over what they think might be permissible discipline; people are often reluctant to get the police involved in such matters unless they're sure that a serious crime is taking place.  Still others may be paralyzed with indecision when confronted with an unusual situation.

              These are generally unworthy responses, but fortunately after reflection or after observing other events, the witnesses may well change their minds.  A gnawing conscience may move them to call the police a day later.  Seeing a second (or tenth) incident of abuse may finally steel them to action.  Seeing the child the next day with a black eye may convince them that the abuse was in fact serious, and that they have a moral obligation to act to stop it.  And even if they don't volunteer, a visit from the police or from child welfare officials investigating the matter may prompt them to tell the truth about what they saw, thus helping save the child from further abuse.

              People who react this way may still be morally culpable for their initial failure to aid or to report, but the more practically important fact is that the legal system -- and the victim and possible future victims -- can nonetheless use their help.  Delayed Samaritanism and Passive Samaritanism aren't as good as Good Samaritanism, but they are much better than nothing, and shouldn't be discouraged.

              Under a duty-to-rescue/report law, though, both the Delayed and the Passive Samaritans have committed a crime by not helping or responding "as soon as reasonably practicable." 7  By the time the remorse sets in, the Delayed and Passive Samaritans are legally guilty, 8 and either volunteering or honestly answering police questions will incriminate him.

              The Delayed and Passive Samaritans therefore have a choice: continue keeping quiet, and likely never be conclusively found out to have been a witness, 9 or speak up and risk a criminal conviction (and perhaps a civil lawsuit 10) for the initial failure to act.  And when silence or "I didn't see anything, officer" is legally cheap (though morally expensive) and volunteering or even answering police questions risks even a modest legal cost, many will stay mum. 11  The same can happen in many other situations, for instance when a person sees a robbery or killing outside his window and at first doesn't report the crime (out of fear, indecision, or a habit of mistrusting the police), or when a frat boy happens to see a friend raping an unconscious girl, and initially keeps quiet out of misguided loyalty.

              Thus, in trying to achieve a "procooperative effect" -- cooperation with the social goal of helping crime victims -- duty-to-rescue/report laws inadvertently cause an "anticooperative effect":  They deter Delayed Samaritans from coming forward, and deter Passive Samaritans from helping the police who knock on their doors looking for witnesses.

              Of course, many people won't know about the duty to rescue or to report, 12 and some who know about it won't think much about it in making their decisions; the risk of punishment for having violated this duty thus wouldn't deter them from coming forward belatedly.  But the same ignorance or lack of concern would equally weaken the law's supposed positive effect -- people who don't know or don't care about the law won't be led by it to promptly help or report.

              Some others may assume prosecutors will decline to prosecute them for the initial failure to help so long as they come forward with help later.  But not all Delayed or Passive Samaritans will have such faith in prosecutorial mercy, given that they know their actions were less than noble and that their own skins are at stake. 13  Even a small chance of being prosecuted 14 may be enough to scare off many witnesses who, by hypothesis, aren't public-spirited Good Samaritans in any event.  (Conversely, if one defends the duty-to-rescue/report laws by stressing that people will expect them to remain unenforced, then one must also acknowledge that the laws will also have virtually zero coercive effect, and probably relatively little normative effect.)

              In some cases, prosecutors could avoid the anticooperative effect by offering an important witness immunity from prosecution for failure to aid or report, but this will work only for witnesses whom the prosecutor knows to be important.  Many witnesses may never come to the prosecutor's attention precisely because they'll keep quiet to avoid any risk of prosecution.

              Moreover, a witness who does testify under grant of immunity would tend to be somewhat less credible to the jury than a witness who, in the absence of a duty-to-rescue/report law, testifies without need for immunity.  ("Isn't it true, Mr. Witness, that the prosecution has agreed to drop criminal charges against you in exchange for your testimony against the defendant?")  Likewise, categorically exempting anyone who tardily comes forward with information will encourage the cooperation of the Delayed Samaritans but will provide no safe harbor for the Passive Samaritans, who'll still be deterred from answering questions when the police come around.

              More radical changes to duty-to-rescue/report laws may indeed avoid the anticooperative effect, but only by dramatically reducing the scope of liability.  My colleague Peter Arenella, for instance, suggests that drafters should include a "second chance" defense, which would exempt from liability anyone who either volunteers information or responds to police questions; this would in effect change the law from a duty-to-rescue/report law to something that it will in practice more resemble a "duty to cooperate" law. 15

              I have rather little objection to such a law, largely because it isn't terribly different from the situation today, where prosecutors may (at least in theory) use subpoenas and the threat of contempt prosecution to require witnesses to tell what they saw.  For the same reason, the law should arouse less libertarian opposition than orthodox duty-to-rescue/report laws would -- the burden it imposes on private liberty is not much greater than that imposed by the general duty to testify when called on to do so.  But in any event, such a law differs considerably from the duty-to-rescue/report laws that have been enacted and proposed so far.

              Sometimes, the existence of a duty-to-rescue/report law may give the prosecutor extra leverage over the witness:  If the prosecutor can prove that a reluctant witness saw the crime but failed to promptly help or report, the prosecutor can use this as pressure -- "tell us what you saw, or we'll prosecute you for failing to rescue or report."

              But I suspect there will be few such cases relative to the others I describe, because without the witness's cooperation, failure-to-rescue/report cases will usually be hard to make stick.  Too often there will be uncertainty about whether the witness was really at the scene, whether he really knew there was a crime in progress, and possibly whether he was afraid of retaliation for rescue or for reporting. 16  This is especially so because other witnesses who could shed light on these questions may themselves be guilty of failing to promptly report and thus be reluctant to come forward.  Sometimes, the victim or someone else can identify the witness and testify that the witness must have known what was going on; but often the witness will only be put in legal jeopardy if he comes forward and admits he was a witness, which is why the Delayed and Passive Samaritans are particularly likely to be scared off by such laws.

              Moreover, a prosecutor's offer of immunity wouldn't provide immunity from possible civil liability, if the duty-to-rescue/report regime allows such liability. 17  Witnesses who are more concerned by the risk of civil liability than of the likely rather modest criminal liability may thus remain silent even in the face of the prosecutorial threat, reckoning that the risks of talking and thus implicitly admitting conduct that could lose them their savings and homes far exceed the risks of silence and possible petty misdemeanor prosecution.

              Finally, some Delayed and Passive Samaritans could try to act on their remorse while minimizing their legal exposure, for instance by calling in an anonymous tip.  But such tips are much less useful to the police than are witnesses who actually come forward, and some risk-averse witnesses might be reluctant even to call in an anonymous tip, fearing that it won't stay anonymous.

              This anticooperative effect should give us real reason to worry.  True, delayed reports, which can usually at most help solve the crime and prevent future crimes, are less valuable than rescues and prompt reports, which can sometimes interrupt a crime or save an injured victim's life.  Still, delayed reports are important; the criminal is likely to act again, either against the same victim or against another.  And crimesolving is a difficult business, one that often requires cooperation from bystanders who have little to gain from helping.  It would be made even harder if delayed cooperation became legally costly.

              The laws' positive effect, I argue above, is likely to be quite modest -- few people, I think, expect the contrary.  But the anticooperative effect could be quite substantial, probably more than enough to outweigh whatever practical benefit the laws may provide.  At the very least, supporters of duty-to-rescue laws must confront this anticooperative effect in making their arguments, something they have not done so far.

              One could of course try to justify the duty-to-rescue/report laws primarily on retributive or denunciatory grounds, not pragmatic ones:  Under this theory, those who failed to rescue or promptly report simply deserve punishment, even if such punishment may ultimately make other criminals harder to catch.

              I'm unpersuaded by this approach, though; even if one accepts that failure to rescue or report is so immoral that it deserves criminal punishment -- a point that's at least controversial 18 -- it hardly seems so grievous that it demands punishment even when the punishment interferes with solving other crimes.  And the lack of prosecutions under the existing duty-to-rescue/report laws suggests that, at least in prosecutors' minds, such failures to rescue or report aren't loudly crying for retribution.  The practical costs created by the anticooperative effect seem to me quite adequate to outweigh any retributive benefits of the law.

              Looking at the matter more broadly, the problem is that duty-to-rescue/report laws focus on a single time -- the time the bystander witnesses a crime -- and criminalize certain conduct in order to change bystander's behavior at that particular time.  But citizens have a continuous relationship to the legal system, a relationship from which the system can benefit at many times.  Making the legal system into the citizen's adversary rather than his protector and servant jeopardizes this relationship, and can deprive the legal system of the citizen's assistance. 19

* * *

              This analysis doesn't show that all duties to rescue or report are unjustified.  In fact, it helps explain why some such narrow duties may be sensible while other, broader ones, are counterproductive (though I don't claim that this is the reasoning historically used to determine when such duties should be imposed).

              Consider doctors' duty to report to the police any gunshot or stab wounds they treat. 20  Doctors who fail to comply with this duty are more likely to be found out than the typical witness would be:  There will often be records that describe the treatment, and other hospital personnel who will testify about it; moreover, the doctor would usually know that failing to report puts his fate in the hands of the patient, who will often be a rather unreliable character.  If the patient does turn out to be a criminal, and the police capture him, he may well implicate the doctor, especially when questioned by the police about the wound.  The doctor, realizing his potential exposure, will be particularly likely to promptly comply with the law. 21

              Furthermore, the normative impact of the law should be considerably greater than the normative impact of more general duty-to-rescue/report laws.  Most witnesses who fail to rescue or report act out of callousness, fear, or deep-seated loyalty to family, friends, or confederates; it's unlikely that knowing about a legal duty to rescue or report will have much of a normative effect on their behavior.  Doctors who fail to report gunshot wounds, on the other hand, are more likely to be acting that way not because they are callous but because they feel a professional norm of loyalty to their patients.  So long as this norm exists, conscience alone might not push doctors to report the wound, but a duty-to-report law is especially likely to effectively undermine this norm because the norm isn't as deeply seated as are more personal norms of loyalty to family or friends.

              Sometimes the legal duty to report could indeed have an anticooperative effect, for instance when a doctor performs the surgery alone, keeps no records, and thinks that if the matter eventually comes to light he can plausibly (though falsely) claim that he hadn't realized the wound was made by a gunshot.  In such a case, a doctor who, out of loyalty or unreasonable fear or hesitation, initially fails to report but later changes his mind could indeed be deterred from reporting by the risk of punishment for his initial silence.  But I suspect the cases where all these factors are present would be rare, and the procooperative effect of the law would thus exceed its anticooperative effect.

              My analysis also doesn't apply to the less common 22 duties to help victims of accidents or natural disasters.  In these situations, late reporting will usually do no good, because there's no criminal to be caught.  It's either prompt rescue (or at least a prompt call for help) or nothing, so the anticooperative effect of a duty to rescue becomes largely irrelevant.  This isn't to say that legal duties to help accident victims are a good idea; they may still be improper on libertarian grounds or for other practical reasons.  But my argument doesn't say much about them.

              The analysis does, however, apply in spades to duties enforced through tort law. 23  Many people rightly fear the risk of civil liability, which could mean the loss of their savings or even their homes, even more than the risk of prosecution under a rarely-enforced misdemeanor statute.  Even if some Delayed or Passive Samaritans might expect a prosecutor to exercise his discretion not to go after them if they report late, they can have no such assurance about the victim, who may be angry, greedy, or both.  The risk of civil liability will thus make them less likely to belatedly come forward or to answer police questions.

* * *

              The anticooperative effect I describe is particularly present in duty-to-rescue/report laws, but it may arise elsewhere, too.  Illegal aliens, for instance, are often thought to be especially vulnerable to criminal abuse, because the risk of being deported may deter them from reporting such crimes; likewise for prostitutes.  Their reporting the crimes may help society, both by preventing future victimization of illegal aliens and prostitutes, and by helping catch criminals who may also prey on otherwise law-abiding citizens.  But such cooperation often won't be forthcoming, because of the anticooperative effects of laws that make it dangerous for illegal aliens and prostitutes to talk to the police.

              These situations of course differ in some ways from bystanders' duty to rescue or report.  The anticooperative effect is increased by the fact that the likely penalty for prostitutes or illegal aliens who in reporting the crime incidentally confess their own criminality is much higher than the likely penalty for bystanders who didn't rescue or report in time.  On the other hand, the anticooperative effect is somewhat diminished by the fact that crime victims -- unlike the mere bystanders -- have a strong incentive to report the crime to the police, since otherwise the criminals could victimize them again. 24  More importantly, if one thinks prostitution or illegal immigration really must be stamped out, the anticooperative effects in these cases might be outweighed by the perceived need for assiduous enforcement of prostitution and immigration laws.

              Still, those who are ambivalent about these laws in the first place might conclude that the anticooperative effects sometimes weigh against maximally stringent enforcement.  The police might thus, for instance, announce a policy of not going after prostitutes or illegal aliens who come to them as crime victims.  Though this might seem counterintuitive at first, given our general assumption that more law enforcement is better than less, the concerns about anticooperative effects of certain kinds of law enforcement can explain this. 25  Likewise, this concern about anticooperative effects of law explains why people announce rewards for the return of lost property with "No Questions Asked." 26

              Similarly, a woman who illegally carries a gun in her purse and uses it to scare off a robber or rapist probably isn't going to report the attack on her, even if her testimony can help catch the criminal and thus make the streets safer for others; after all, if the criminal is caught, his story will incriminate her, too.  Her only incentive to go to the police is public-spiritedness, and even a slight risk of prosecution on a weapons charge is more than enough to counteract that.

              Again, if we think that law-abiding citizens can't be trusted to carry guns, we might conclude that the carry ban's positive effect exceeds the anticooperative effect.  But if we think that it's a close call whether such carrying actually poses a danger -- in fact, 31 states now let pretty much all law-abiding adults carry concealed weapons, 27 holders of concealed weapon permits very rarely abuse their guns, 28 and a recent study suggests that allowing concealed carry may decrease violent crime 29 -- then the anticooperative effect of concealed carry laws may break the tie in favor of concealed carry liberalization.

              More generally, all this illustrates the need to consider what happens when people refuse to comply with a proposed new law, rather than just evaluating the law on the assumption that they will comply.  Returning to the duty to rescue/report, we see a proposed law that many think is morally proper.  It at first seems to create a good incentive.  If people reacted to laws simply by obeying them, such a law would probably increase social welfare (setting aside libertarian objections).  But some people won't obey the law, and this noncompliance won't simply mean the law won't affect their behavior:  The law's anticooperative effect -- through which an initial violation of the law makes one less likely to act in certain socially useful ways later -- will at least counteract the law's positive effects, and may end up outweighing them.

              When a law turns people into outlaws, even only modestly punishable ones, it naturally makes them less likely to cooperate with the legal system that is threatening to prosecute them, especially when their cooperation can alert the legal system to their misdeeds.  This is an inevitable cost of using the threat of force to coerce people's conduct; while it may often have to be paid, it ought not be ignored, and for some laws, such as duty-to-rescue/report laws, it may be dispositive.  


    *    Acting Professor of Law, UCLA Law School (volokh@law.ucla.edu).  Many thanks to Peter Arenella, David Dolinko, James Ho, Dan Klerman, Elaine Mandel, Gene Meyer, Herbert Morris, David Sklansky, and James Wiley for their help.

    1.    I focus here -- as do most duty-to-rescue/report statutes -- only on crime victims, not victims of accidents or natural disasters.  See infra note 22.

    2.    See, e.g., Mass. Gen. Laws Ann. ch. 268, § 40 ("Whoever knows that another person is a victim of aggravated rape, rape, murder, manslaughter or armed robbery and is at the scene of said crime shall, to the extent that said person can do so without danger or peril to himself or others, report said crime to an appropriate law enforcement official as soon as reasonably practicable"); Haw. Rev. Stat. § 663-1.6 (applying to all crimes in which the victim suffers "serious physical harm"); Ohio Rev. Code § 2921.22 ("No person, knowing that a felony has been or is being committed, shall knowingly fail to report such information to law enforcement authorities"); 12 Vt. Stat. Ann. § 519 (applying to all situations where "another is exposed to grave physical harm" and requiring "reasonable assistance" rather than just reporting to the authorities); State v. Miccichi, No. 86AP08066, 1987 WL 14481, *2 (interpreting the Ohio statute as requiring a report within a "reasonable time").  See also Fla. Stat. Ann. ch. 794.027; Mass. Gen. Laws Ann. ch. 269, § 18; Minn. Stat. Ann. § 604A.01; R.I. Stat. §§ 11-1-5.1, 11-56-1; Rev. Code Wash. Ann. § 9.69.100(1)(c); Wisc. Stat. Ann. § 939.20.  See also 105 H.R. 4531, 105 S. 2452, 105th Cong., 2nd Sess. (Sept. 14, 1998) (proposing that federal funds be withdrawn from any state that doesn't impose a duty to report on people who witness sexual abuse of a child); 1998 N.J. A.B. 2517 (Oct. 5, 1998) (proposing more general duty to report); Earl Ofari Hutchinson, Protecting Children Always a Good Cause, L.A. Times, Sept. 6, 1998, at M5 (describing proposed California and Nevada laws that would impose duty to report on people who witness assault against a child).

    3.    See, e.g., statutes cited supra note 2; State v. Joyce, 139 Vt. 638, 641, 433 A.2d 271, 273 (1981) (dictum) (duty-to-rescue law "does not create a duty to intervene in a fight"); Daniel B. Yeager, A Radical Community of Aid: A Rejoinder to Opponents of Affirmative Duties to Help Strangers, 71 Wash. U. L.Q. 1, 24 (1993) ("None of the easy-rescue states would require individual bystanders in [the context of a rape in progress] to do anything more than make a phone call.").  I also assume, as have others, that duty-to-rescue laws implicitly impose a duty to report, since one way of providing assistance to a crime victim (especially when direct assistance is dangerous) is to promptly call the police.  In any event, the points I make in this article apply equally to both duty-to-rescue and duty-to-report laws.

    4.    By "normative force" I mean the law's capacity to mold people's moral norms and to lead them to act in certain ways because they've been influenced to think that those ways are right.

    5.    See, e.g., Yeager, supra note 2, at 57 (acknowledging that "widespread adoption of rescue and reporting laws may not significantly change the behavior of bystanders when another person is in distress," but arguing that it may sometimes "tip the balance toward the desired action" and concluding that this marginal benefit is enough to justify the law).

    6.    I set aside in this article the moral debates about whether a duty to rescue or report would impose an impermissible burden on individual liberty; these matters have been adequately aired elsewhere.  Compare, e.g., Steven J. Heyman, Foundations of the Duty to Rescue, 47 Vand. L. Rev. 673 (1994) (supporting such duties), and Yeager, supra note 2 (same), and A.D. Woozley, A Duty to Rescue: Some Thoughts on Criminal Liability, 69 Va. L. Rev. 1273 (1983) (same) with Richard A. Epstein, A Theory of Strict Liability, 2 J. Legal Stud. 151, 197-204 (1973) (taking the opposite view), and Robert Nozick, Anarchy, State, and Utopia ix (1974) (likewise arguing that the state "may not use its coercive apparatus for the purpose of getting some citizens to aid others"), and George Fletcher, Law and Morality: A Kantian Perspective, 87 Colum. L. Rev. 533, 543-46 (1987) (describing Kant's view as being opposed to a legally enforceable duty to rescue).

              I do think that failing to rescue or to report ought not be considered harming another, because it generally leaves the victim in the same position as he would have been had the bystander not existed; I therefore believe people should be presumptively free from a duty to rescue or report.  On the other hand, a duty to tell the police about crimes is not far removed from the traditionally accepted duty to testify in court when subpoenaed.  The fact that we consider the duty to testify to be a permissible -- even a generally uncontroversial -- infringement of personal liberty suggests that duties to report might likewise be permissible infringements.  I therefore think that, viewed from a purely moral perspective as opposed to the pragmatic one I take in this article, a duty to report (though not a duty to rescue) poses a genuinely close question about ordered liberty.

    7.    See statutes cited supra note 2.  I assume here that the state duty-to-report law covers serious child abuse; some laws do and some don't.  Compare, e.g., Haw. Rev. Stat. § 663-1.6 (covering all "crime[s]" that cause "serious physical harm"), and Rev. Code Wash. Ann. § 9.69.100(1)(c) (specifically covering "[a]n assault of a child that appears reasonably likely to cause substantial bodily harm") with, e.g., Mass. Gen. Laws Ann. ch. 268, § 40 (covering only "aggravated rape, rape, murder, manslaughter or armed robbery"), ch. 269, § 18 (covering hazing).

    8.    If the Samaritan was honestly mistaken about the character of what he saw -- for instance, if he honestly hadn't recognized that the child was "exposed to grave physical harm," see, e.g., 12 Vt. Stat. Ann. § 519 -- then he might not be guilty of failure to report, but he might still be concerned that the police might assume the contrary; his going to the police and admitting he was a witness would thus lead to the risk of legal jeopardy.  Likewise, the Samaritan might not be guilty if he at first honestly (but incorrectly) thought that reporting the crime would expose him to "danger or peril," see, e.g., id., but here too he might be concerned that the police wouldn't believe that he was indeed scared.  This would be especially likely if the law puts the burden on the bystander to prove that he felt a risk of harm, see, e.g., Minn. Stat. Ann. § 609.662 subd. 4 (defining risk of harm as an affirmative defense), or that the perception of risk must have been reasonable, see, e.g., id.

    9.    Cases such as the notorious New Bedford rape (dramatized in the movie The Accused), in which dozens of people stood by clearly seeing that a rape was going on but doing nothing, might fit this mold; but my suspicion is that these are likely a small fraction of all failure-to-rescue/report cases.

    10.    See Sabia v. State, 164 Vt. 293, 304, 669 A.2d 1187, 1194-95 (1995) (indirectly suggesting that state duty-to-rescue law may impose tort liability for violation of the duty); Heyman, supra note 6, at 707 (urging such a tort regime); Jay Silver, The Duty to Rescue: A Reexamination and Proposal, 26 Wm. & Mary L. Rev. 423 (1985) (same); Robert Justin Lipkin, Comment, Beyond Good Samaritans and Moral Monsters: An Individualistic Justification of the General Legal Duty to Rescue, 31 UCLA L. Rev. 252 (1983) (same); compare Richard L. Hasen, The Efficient Duty to Rescue, 15 Int'l Rev. L. & Econ. 141 (1995) (arguing that a general tort law duty to rescue may be economically efficient) with William M. Landes & Richard A. Posner, The Economic Structure of Tort Law (1987) (taking the opposite view).

    11.    It would be great to have some empirical data on how large this effect would be, but I don't think any such data is available; the few American duty-to-rescue laws are heavily underused and largely little-known by the public, but even if they were more common, it would be hard to gather data on behavior (unknown witnesses' failure to go to the police) that by definition doesn't come to light.  I'm also unaware of any data about the effects of the European laws, and am in any event skeptical that any such data would tell us much about what effect such laws would have in American culture.

    12.    I suspect that eventually many people will learn about this duty, probably through media coverage of prosecutions for failure to rescue or report in high-profile cases.

    13.    Though the Massachusetts, Minnesota, and Vermont statutes impose only a small fine, the other laws authorize jail time.  Cf. State v. Miccichi, No. 86AP08066, 1987 WL 14481, *2 (sentencing a pharmacist to 30 days in jail for failing to report a theft of drugs from his pharmacy; the earlier theft came to the attention of the police only when the pharmacist promptly reported a later theft).

    14.    Indeed, prosecutions under these laws have seemingly been very rare.  See, e.g., Yeager, supra note 2, at 34 ("Scant trial and appellate court precedents indicate that [recent duty-to-rescue/report laws] have been symbolic at most."); id. at 8 n.37 ("In a written response to a questionnaire that I sent to 387 prosecutors in the eight states that impose duties to render easy aid or duties to report serious crimes, none of the 139 prosecutors who responded could recall filing a complaint under the relevant statute.") (citation omitted).

    15.    Peter Arenella, Taking Character Seriously (work in progress).

    16.    See, e.g., statutes cited supra note 2.

    17.    See supra note 10.

    18.    See supra note 6.

    19.    I don't want to overstate the importance of the anticooperative effect; in many situations, it's not a powerful consideration.  We are perfectly right, for instance, to persecute those who murder or rape or steal, even though this alienates them from the justice system:  First, the harm of not punishing these crimes would easily outweigh the harm of this alienation, and, second, the people who commit these crimes are probably morally corrupt enough that they can't be expected to much help the legal system anyway.  I argue only that in certain situations, especially where a proposed new law would do relatively little good, the law's anticooperative effect may be enough to tilt the scales against it.

    20.    Cf., e.g., Cal. Penal Code §§ 11160-11161.

    21.    Some doctors may be reluctant to tattle on their criminal patients because they're afraid that the patients will retaliate; but this anticooperative effect flows from the criminal, not from the duty-to-report law, and thus tells us little about whether the law makes sense.

    22.    Only four of the nine state duty-to-rescue/report regimes -- Minnesota, Rhode Island, Vermont, and Wisconsin -- apply to perilous situations generally as well as to crimes.  See statutes cited supra note 2.  Cf. Heyman, supra note 6, at 678-79 (suggesting that the normative case for a legal duty to help crime victims is stronger than the normative case for a legal duty to help accident victims).

    23.    See sources cited supra note 10.

    24.    Even victims who aren't concerned about victimization by the same criminal may want to report the crime to the police so as to deter attacks by other criminals.  Getting a reputation as someone who won't report attacks on oneself to the police is a good way of getting attacked more often.

    25.    See supra text preceding note 16 for an explanation of why such an approach won't work in duty-to-rescue/report cases.

    26.    A related concern explains why many states give immunity from negligence lawsuits to bystanders who help crime or accident victims, see, e.g., Minn. Stat. Ann. § 604A.01 subd. 2:  Just as the fear of incidentally implicating oneself in a crime can discourage people from doing the right thing by going to the police, or from doing the right thing by returning property to its owner, so the fear of inadvertently incurring civil liability can discourage people from doing the right thing by trying to give emergency aid.  Cf. also Model Penal Code and Commentaries pt. I, §§ 5.01(8), 5.03(6) (1985) (describing the renunciation defense in attempt and conspiracy cases, and justifying it as an incentive for people who have begun an attempt or conspiracy to do the right thing by stopping their course of conduct before actual harm is done); but see Paul R. Hoeber, The Abandonment Defense to Criminal Attempt and Other Problems of Temporal Individuation, 74 Calif. L. Rev. 377, 397-98 (1986) (rejecting this argument on empirical grounds).

    27.    Thirty states require the police to give concealed carry licenses to pretty much all law-abiding adults who apply; one state, Vermont, allows concealed carry without a license.  See John R. Lott, Jr., More Guns, Less Crime (1998).

    28.    In Florida, for instance (where the Department of Licensing has kept thorough records since concealed carry was liberalized in 1987), there are now a bit under 230,000 active concealed carry licenses.  Over the 11 years since 1987, on average 10 licenses per year have been revoked due to a gun crime by the licenseholder.  See http://licgweb.dos.state.fl.us/stats/cw_monthly.html (visited Mar. 4, 1999).  Even considering that not all gun-abusing licenseholders will be caught, this still reflects an extremely low rate of abuse.

    29.    See Lott, supra note 27.