The Duty to Rescue and the Registry for Caregivers: A Guest Post

We have recently featured several guest posts (here, here, and here) by the authors of a new book about criminal justice and the family called Privilege or Punish: Criminal Justice and the Challenge of Family Ties. The authors are Ethan Leib, who is a scholar-in-residence at Columbia Law School, and an associate professor of law at the University of California-Hastings College of the Law; Dan Markel, the D’Alemberte Professor of Law at the Florida State University in Tallahassee; and Jennifer Collins, a professor of law at Wake Forest University in Winston-Salem. Leib and Markel usually blog at Markel has offered to send interested parties a free PDF of their new book upon request. This is their final post, and we thank them for their stirring contributions.

The Duty to Rescue and the Registry for Caregivers
A Guest Post
By Jennifer Collins, Ethan J. Leib, and Dan Markel

In two previous posts, we examined laws exempting family members from prosecution for harboring fugitives and laws either granting or permitting sentencing discounts on account of one’s family status, ties, or responsibilities. These are two of the benefits defendants receive on account of their family status in the criminal justice system.

Today, we explore one of the burdens defendants face in the criminal justice system as a result of their family status. Specifically, we’ll look at the phenomenon of omissions liability, a legal doctrine which places criminal responsibility on certain persons because they didn’t do anything; they’re punished, in other words, because they had a duty to perform a relatively costless rescue, and they breached that duty. We will focus our discussion on the spousal obligation in particular.

The Law and Its Rationale

Generally speaking, most American citizens are under no obligation to rescue each other from peril. Two well-known exceptions to the rule in most jurisdictions (in the U.S.) exist: parents must make (relatively costless) efforts to save children, and spouses must make the same efforts to save each other.

Hmmm. Only parents and spouses. Why not grandparents, cousins, siblings? We think the answer has something to do with the fact that parental and spousal relations are the two familial relationships that persons enter into with some degree of real voluntariness. I can’t choose to have a grandfather or a sister. But I can avoid marriage — notwithstanding the subtle pressures or inducements from Jewish mothers or the government’s social policies. Similarly, obligations to one’s child might be legitimate in part because it’s largely a choice to have a child — at least in an age and polity where contraception, abortion, and the chance to terminate one’s parental rights exist.

Thus, despite the general common law rule in favor of maximizing personal autonomy, the government’s imposition of an obligation to rescue spouses and children is not so illiberal; it reflects the fact that the underlying conduct of having a spouse or child is such that it can be avoided without much difficulty.

What are the rationales offered for penalizing a failure to rescue between spouses? Usually, they are 1) saving human lives in danger and 2) affirming the significance of marital obligations.

The problem with the first interest is that the means used here — spousal obligations to rescue each other, policed through the criminal law — is woefully underinclusive, so much so that it’s hard to take seriously the idea that this is what’s motivating the use of this family-ties burden.

The second objective, by contrast, makes relatively more sense. Although the obligation to undertake easy rescues is not specifically articulated in many wedding vows, it reasonably falls under the language that is often used in those vows.

Our Take

The problem with relying principally on wedding vows in these contexts is that the state ends up drawing weird and inexplicable lines. For example, why do the partners of a newly married heterosexual couple have duties to rescue each other but not those in the long-term homosexual couples living in states that won’t recognize their unions? If the key feature of the spousal relationship (for purposes of assigning omissions liability) is the voluntary assumption of caregiving responsibilities, then we don’t see why the breach of other voluntarily assumed caregiving obligations are not similarly subject to criminal law sanction.

As we argue in greater detail in our book, any potential prosecution of a person for failing to protect his or her spouse from harm also has the potential to have a discriminatory impact, in a different and critical sense: it treats differently those who cannot or choose not to enter a spousal relationship sanctioned by the state. For example, these laws currently do not clearly give the family members of homosexual couples the comfort of knowing that omissions liability is parceled out in a non-discriminatory fashion.

One way to see this discrimination is through analogy: if omissions liability were distributed on the basis of race, such that whites had a duty to rescue their spouses but blacks did not unless they separately contracted for that duty, what message would that send? Our sense is that it exhibits a lack of respect of the value of the spouses of black people. The same is true by restricting omissions liability along lines that are tethered to the few family status relationships recognized by the state. Why should a heterosexual man have an obligation to protect his spouse from harm while a gay man in a similarly meaningful and voluntary partnership does not? In both instances, imposing liability serves the same valuable functions: increasing safety and promoting an ethos of caregiving relations triggered by voluntary choices. Thus, limiting omissions liability to those in a state-sanctioned relationship seems plainly underinclusive; it leaves out those who cannot get married because of a plainly troubling moral choice made by the state.

For the most part, we do not have much problem with marriage being an overinclusive obligation because divorce is an option by which the obligation can be terminated. But because marriage is an underinclusive basis for imposing omissions liability, we think several options should be explored.

Some Options For Designing Policy

One solution would be to decouple omissions liability from marriage altogether, and instead ask parties to any relationship to register with a state registry. This would treat all persons the same and without favor. But a no-duty-to-rescue rule in marriage could act like a penalty default rule. On the one hand, it would probably encourage more people outside of marriage to think about whom they wish to rescue. On the other hand, it might also add needless costs associated with persons who by virtue of marriage would already be willing to undertake a duty to rescue.

A better solution, based on reducing the social costs of the scheme, would be to require duties to rescue in marriages and to create a registry for all others who want to participate in a “compact of care” such that they have a duty to perform easy rescues. Marriages would simply have the implicit term of duty to rescue built into them and others outside marriage (including those in polyamourous relationships) could opt into it. This would also allow persons to insist on seeing evidence of opt-in by another person before they decide to jointly acquire property, cohabit, or perform caregiving tasks for one another.

Some might raise concerns that we are too focused on spouses and parents as paradigmatic relationships here. The concern here would be that we are insisting that the sexual family or marriage be the normative ideal for adult interactions with each other. We respectfully disagree. Indeed, the point of our registry system is to obviate this concern entirely. People who are not married but “act” as though they are do not have to register, but they may choose to do so; or just one may decide to do so for the other since the registry is a place of declaring one’s assumption of obligation — it is not predicated on norms of reciprocity, nor does it require contractual formalities. To be sure, our slight preference for assigning duties to rescue in the context of marriage and custodial parenting is responsive to what we think of as the specific features of caregiving written into the “scripts” of marriage, but no one should be forced into those roles.

What’s more, people should be free to and encouraged to assume these obligations outside the scripts of marriage. The registry we endorse permits siblings or cousins or roommates or friends to enter into compacts of care, but the idea is not to require it through the criminal law outside voluntary choices or the specific circumstances of the parent-child or spousal relationships. Indeed, we would resist any state’s attempt to impose a legally enforceable relationship of caregiving or a duty to rescue on those persons outside the parent-child or spousal context because we simply cannot say these relationships have been entered into voluntarily.

In the context of platonic roommates, imposing a duty of care through the criminal law would be a drastic restructuring of the traditional boundaries of that relationship. On the other hand, we certainly believe that individuals should be able to choose a legally enforceable relationship of caregiving through the use of a registry. This allows individuals to signal their commitment both to each other and to those around them.

It is, of course, possible that very few individuals will choose to register; why would they voluntarily assume the risk of a legal liability that they currently do not face? But if that is the outcome, we are no worse off than we are now, as these individuals do not currently face liability. If, on the other hand, some individuals do choose to undertake an obligation to rescue, the benefits that decision conveys in terms of promoting safety and promoting an ethos of care and compassion certainly seem worth the effort.

We can also imagine the state incentivizing such registrations through small tax breaks; alternatively, perhaps norm entrepreneurs (private employers or faith groups) will start “opt-in days” to foster solidarity among members of their communities. Of course, because peoples’ relationships ebb and flow, the registry would have to permit people to withdraw from these compacts of caregiving if notice is given to the affected parties. Crazy? Feel free to let us know (gently) in the comments. Thanks again to our wonderful hosts for letting us share our freaky, if not quite freakonomical, ideas.


This post baffled me. Can you give examples of what a "simple rescue" would be? Why would anyone opt into one of these relationships? If I had someone I cared about, I would choose to help them in whatever way I reasonably could, whether or not I was legally obligated to do so. But why would I risk criminal prosecution and legally oblige myself to do so, on the off chance that one day I might not be able to or want to? Maybe I just missed the point, but this one was confusing. Perhaps an explanation of "simple rescue" is in order.


Can you provide any examples of prosecutions for failure to perform a "relatively costless rescue?" This sounds like much ado about nothing.


Say, calling 911 in the case of an emergency.
Their point with the registry is that assuming no one wants to opt in once it's enacted, the situation with the registry would be the same without. The assumption is that some sort of incentive would be put in place similar to the incentives in place for marriage and having children. The proposal would allow people to choose to opt if they wanted to, but I'd be interested to see how much a registry would cost vs. the benefit this would actually provide.


As #1 BSK asked, what is a "simple rescue"?

"Specifically, we'll look at the phenomenon of omissions liability, a legal doctrine which places criminal responsibility on certain persons because they didn't do anything; they're punished, in other words, because they had a duty to perform a relatively costless rescue, and they breached that duty. We will focus our discussion on the spousal obligation in particular."

I understand criminal liability in the chlid example of a mother who has knowledge but fails to "rescue" her daughter from being molested by a her husband who is a step-father to the minor child.

But what is the spouse version of this?

Is this when a spouse is required to bankrupt his/her self when the other spouse has a chronic or terminal illness? These are the cases when the spouses "divorce" in order to protect the assets of one and push the cost of care of the other onto the public system.

Is this blog post subject an end run around the homosexual marriage issue? (Not that I care- let everyone get married - even let plural marriages between adults be legal- whatever adults want to do.)



Your notion of registry reminds me of the PACS (Pacte de solidarité civile) which exists under french civil law. The PACS that does just that: allow two persons, including homosexual couples or even platonic roommates, to jointly acquire property, recieve tax breaks, etc.

Great post and fascinating topic!
Thank you!

Jason Walters

If you are trying to eliminate the discrimination it seems a much easier solution is to eliminate the mandatory marital obligations and allow any pair or group of people who want to enter into such a contract to do so freely.


Gee, I've never heard of a law that says I need to rescue my spouse. Can you give text from one state law that illustrates what you're talking about. I'm looking for something like: "There shall be no legal duty to rescue another person from physical harm unless you are married to that person."

Seth Glasgow

This issue does indeed confuse me. It seems almost aimless that one would opt-in to a situation such as this. As was echoed by earlier posters, people will make a reasonable and concerted effort to rescue any one they care that deeply about regardless of the criminal statutes that may apply. Secondly, what would be the state's incentive to offer a similar incentive for people to rescue each other? Why would the state give me a tax break (or whatever the incentive would be) because I placed myself on a list of people who would save each other? Unlike having children, which obviously requires the parents to provide a great deal of care, there is no garauntee that I would ever have to fulfill my promise of "reasonable rescue". The incentive seems to reward the possibility of something happening as oppposed to the actual action as is the case with caregiving to children which cannot be avoided. Furthermore, it does seem that this premise is simply to get at the gay rights issue (which I am totally in favor of). It serves to point out that inconsistencies in famlial definition have legal ramifications in regards to rights and duties. But, at the end of the day, the fundamental issue is still about denying basic rights to a segment of the population which has done nothing to deserve it (unlike, say, felons). This denial of freedoms will always be wrong (in an ethical sense), regardless of how it is framed legally.



This seems to put the cart before the horse. Shouldn't we start with a discussion of the basic construction of the duty to rescue, without the specific criteria of parenthood or marriage that are used to limit it in certain jurisdictions? If we understand the rationale for a legal duty to rescue, then we can start talking about what classes of relationships it should apply to. As is, though, this just doesn't make sense.


There is among us moderns a somewhat cynical tendency to overly rationalize laws and customs. We say that Jewish law prohibited the eating of pork as a way of avoiding trichinosis, or that the REAL reason Hindus treat cows as sacred is that they are more valuable for milk and fuel than they are for meat.

Searching for the "state's interest" in various laws is similar. Lawmakers may well pass laws not so much because of what they think the state's interest is as just because something seems right to them from an idealistic standpoint. A perfect example is Good Samaritan laws. What interest does the state have in giving protection to incompetent roadside medical assistance? None. But it makes people feel better if someone who means well doesn't get punished for not knowing what they're doing.


I fail to see the point of this article. Regardless of whom the law affects, people generally don't need external incentives to save loved ones, and as a result, any discriminatory effect this law may have in theory is in practice so small as to be negligible. I sincerely doubt that there are a significant number of cases where this law becomes an issue.

Robot Mistake

The craziest part of this idea is that we are dealing with the liability for the failure to act.

You are suggesting a person that was capable of providing assistance to someone in need, who has no duty under the law, would go and sign a registry creating a duty to act. Then when the time came to act, would decide not to.

There is only a rational reason to create the duty if there is some insentive. Such as immunity from negligence when action is taken.

The arguement seems very weak due to the fact that any party can make the choice to act, independant of whether thier is a legal duty.

Frances Howard-Snyder

Do these laws require one to save one's spouse or child RATHER than a stranger? For example, suppose an agent can easily save either of two people but not both. If one is her husband, could she be prosecuted for neglecting him to save the other person? If the husband is an unpleasant, even abusive man whom she hasn't had the opportunity to divorce yet, it might make sense for her to save the innocent child instead.

Peter Merriam

For those who have said they don't understand the duties of a spouse or parent, one writer used the example of calling 911, which would naturally be included. But think of the cases (admittedly rare) where a parent, or person with similar responsibility (e.g. foster parent) starves or malnourishes their kid. Only the parent is responsible, not the neighbor, not the teacher. Likewise, if your spouse falls ill, and needs someone to give them a shot once a day, the responsibility would naturally fall to the healthy spouse. Failure to provide this simple care on the part of a stranger (or cousin, for that matter) would never be seen to be criminal. However it might be criminal if the healthy spouse is aware of what is needed, capable of performing the action, and blows it off anyway.

Just as a matter of perspective, some other countries (I know of at least one, Turkey, where I lived), have good Samaritan laws that state that any passerby, regardless of medical training, is required to render aid to an injured person. Something our litigious society has a hard time with, alas.

I think the author's point is that we, as a society, give tax breaks, and other breaks (insurance rates) to folks who are married, or are parents is because we assume that these caregiver relationships add value to society, and deserve at least a little recognition, and perhaps remuneration. If other caregiver relationships result in similar societal benefits (fewer folks dying or going into toxic shock or whatever from lack of simple care), why not give them similar breaks?




This is a very interesting post, but I have a few concerns:

(1) Isn't the suggested registry somewhat superfluous? Couldn't I now sign a "rescue contract" with a stranger or a family member, giving us a legal obligation to rescue each other? I suppose that breaking the contract may not be a CRIMINAL offense (I think -- I'm not that well versed in legal matters), but so what? There would be a legal remedy for the injured party, and isn't that what matters here?

(2) I'm a bit puzzled as to why it is so important that the relationship be VOLUNTARY. Consider a mother's obligation to rescue her child. On the view these authors propose, whether she has this obligation would depend on whether she CHOSE to have this child. Of course, in most cases mothers choose to have their children, but not always. Do we really want to say that a child who was the product of rape is not entitled to be rescued by his mother? Suppose that I, but not my brother, is the product of rape. Should we say that he is entitled to be rescued by our mother but that I am not? That seems rather arbitrary.

Now, one might reply that, in the rape case, the mother at least CHOSE to go through with the pregnancy rather than have an abortion. But suppose that, on one of my nightly walks, I come across a stranger drowning in a pond whom I could easily rescue. Couldn't we then say that, since my being in that location is voluntary (I CHOSE to take a walk), that I owe it to him to rescue him? My worry, in short, is that if one insists that there is voluntariness even in the rape case, then what principled reason is there to say that there isn't also voluntariness in the "rescue a drowning stranger" case? We could even build into this case, if you like, that people often drown in this pond, so that you have some reason to expect that, should you walk by the pond, someone will cry out for help.

My own view, for what it's worth, would be to have a generic obligation to rescue, regardless of the voluntariness of the requisite relationship.

Great post, though! Very thought provoking.



While an interesting topic, I have to say I found the writing fairly dense. Some concrete examples to help anchor the opinions/theories stated would help my understanding, as I found this as esoteric as the tax code.

Charles Long

It doesn't matter what laws they pass. Some folks will do what's right, others won't. Doctors have this dumped on them. Most really try for the best outcome. Irresponsible spouses are just the way they are. Some folks can't handle it. Somebody, usually emergency room personnel, are left with the task. Neglect is abuse, even if not intentional.
What is the suggestion that solves this? I missed it.


"It's largely a choice to have a child - at least in an age and polity where contraception, abortion, and the chance to terminate one's parental rights exist." ...for the woman.

Thomas Sewell

For the reasons pointed out already, the authors' solution is silly. It's a government solution looking for a problem.

How about the alternative solution of simply dissolving government's involvement in the matter altogether?

Julian Goetz

The analogy of black and white citizens made the imbalance of the current situation quite clear and interesting. I don't believe that discrimination like that could be measured; therefore, to seek a solution is necessary.