1. Introduction
Susan Moller Okin – and especially her critique of libertarianism in Justice, Gender, and the Family – has received only scant attention in the libertarian literature, compared to such other egalitarian liberals as John Rawls, T.M. Scanlon, Amartya Sen and G.A. Cohen (Okin Reference Okin1989, chapter 4). For example, the index to the Cambridge Companion to Nozick’s Anarchy, State, and Utopia includes nine entries for Cohen, four for Scanlon, one for Sen and an entire essay on Nozick’s critique of Rawls. As best as I can tell, Okin’s name never appears anywhere in the volume (Bader and Meadowcroft Reference Bader and Meadowcroft2011). Indeed, her critique has sometimes been dismissed offhand. For instance, since the publication of Justice, Gender, and the Family, the prominent libertarian Jan Narveson has written four single-author books. He discusses Okin’s critique in only one of these, in an endnote to a chapter on children. After the citation to Okin, the entire endnote – and so the entire discussion of Okin’s critique – runs as follows:
The main argument of that chapter, and more generally of the book, cannot, I think, be taken seriously, though the fact that it has been suggests that common sense does not play the role in philosophy that one might hope. (Narveson Reference Narveson2002, 278n3)
I find this neglect of Okin’s argument surprising: while Okin’s original presentation of the argument was at times hyperbolic, the basic structure was quite clear and straightforward. Libertarians generally believe that people are self-ownersFootnote 1 and that (perhaps because they are self-owners) they own the things that they make. But, Okin points out, children are made by their mothers. So mothers own their children, and so children are not self-owners.Footnote 2 The premises here seem to be two core libertarian commitments and a rather well-established fact, and the argument produces a logical contradiction in just a few steps. The argument seems to be sound, and the implications for libertarianism are devastating.
I proceed as follows. In the next section, I present Okin’s argument in her own words, then offer a precise and (apparently) logically valid reconstruction. This reconstruction suggests a taxonomy of possible objections, and the next two sections each consider one kind of possible objection. I discuss several problems for each objection. One problem, which recurs throughout the paper, is an extension of Okin’s argument that I call OKIN’S DILEMMA. In brief, the objections can avoid her original argument, but only by requiring an involuntary transfer of property from one person to another. That is, they retain consistency between self-ownership and their account of the acquisition of property by giving up strong property rights. This would seem to be about as unacceptable for right-libertarianism – and some versions of left-libertarianism – as giving up self-ownership. I conclude with a discussion of some possible implications of giving up strong property rights.
2. Okin’s argument
Let us begin by reviewing Okin’s argument as she presented it. Her feminist-critical aim in giving this argument, as in almost every other chapter of Justice, Gender, and the Family, is to identify androcentric assumptions in the work of major political philosophers: these philosophers have neglected women, and in ways that create serious problems when women are no longer neglected. As she puts it:
What becomes of libertarian arguments when we apply them to all the adult members of society, women as well as men …. Instead of the minimal state that [Robert Nozick] argues for in Anarchy, State and Utopia, what results is a bizarre combination of matriarchy and slavery that all would probably agree is better described as dystopia …. [Hence] libertarianism in any form tacitly assumes, beyond the reach of its principles, a realm of private life in which the reproductive and nurturant needs of human beings are taken care of. It also assumes that work performed in this realm is not work in the same sense, or deserving of the same rewards, as that done outside this sphere. Behind the individualist facade of libertarianism, the family is assumed but ignored. (Okin Reference Okin1989, 75, her emphasis, my brackets)
Before reviewing the essentials of Nozick’s view, Okin claims that
when one considers the fact that women … have the natural capacity to produce people, it becomes apparent that Nozick’s entitlement theory of justice – based as it is on the notion that one owns what one produces – leads to absurd and inconsistent conclusions. (Okin Reference Okin1989, 76, her emphasis)
She notes that the principle that ‘one owns what one produces’ is ‘the crux’ of Nozick’s argument against (re)distributive accounts of justice:
Whoever makes something, having bought or contracted for all other held resources used in the process (transferring some of his holdings for these cooperating factors), is entitled to it. The situation is not one of something’s getting made, and there being an open question of who is to get it. Things come into the world already attached to people having entitlements over them. (Nozick, quoted in Okin Reference Okin1989, 78; parentheses and emphasis as in Okin. Note that Okin is deliberately using Nozick’s non-gender-neutral style for this discussion, 76n.)
In Nozick’s view, the person who creates something owns it from the moment it begins to exist – he acquires it exerting the labour that produced it. Any scheme of redistribution thereby involuntarily (by assumption) transfers some of his property to another person; it does not simply distribute unowned things.
In excerpt, the core of Okin’s argument runs as follows:
The assumption that each person owns himself, however, can work only so long as one neglects two facts. First, persons are not only producers but also the products of human labor and human capacities …. [A]ll human beings are necessarily, at birth …, the products of specifically female capacities and female labor. When this one simple fact of human life is taken seriously … it renders Nozick’s entire theory contradictory to the point of absurdity at its pivotal point – the principle of just acquisition. Instead of a utopian minimal state, Nozick’s individuals are left in a condition of matriarchy, slavery, and dystopia. (Reference Okin1989, 79–80, her emphasis)
Indeed, there is nothing about a woman’s production of an infant that does not easily fulfill the conditions of the principle of acquisition as Nozick specifies them. ‘Whoever makes something, ’ he says, ‘having bought or contracted for all other held resources used in the process …, is entitled to it.’ Pregnancy and birth seem to constitute a paradigm of such processes. Once she is freely given a sperm (as usually happens) or buys one (as is becoming no longer very unusual) – in either case amounting to legitimacy in transfer – a fertile woman can make a baby with no other resources than her own body and its nourishment. (Reference Okin1989, 82–83)
Since [Nozick] so firmly upholds in all other cases the principle that persons are fully entitled to whatever results from their natural talents and capacities, he would seem to have no way of avoiding the conclusion that only women own the children they produce. (Reference Okin1989, 83)
Some of the responses to Okin complain that the logic of her argument is unclear, or that she has overstated her claim to have produced a strict reductio ad absurdum of Nozick’s view (Cohen Reference Cohen1992, 275; Jeske Reference Jeske1996, 138; Perrett Reference Perrett2000, 386). However, attempts (including by some of those who make this complaint) to present her argument in a more logically rigorous way themselves fall short. In part this is because one premise of the argument is, essentially, the entirety of Nozick’s entitlement theory, with all of its various commitments, qualifications and ambiguities. And as the debate over the argument develops, the premises must be expanded and modified. Altogether, I think that Okin’s argument cannot be presented both perspicuously and with full logical rigour. Nonetheless, in the remainder of this section, I will present a simplified reconstruction of the argument that at least seems to be logically valid.
For my reconstruction, I must introduce some terminology to characterize the core commitments of libertarianism that Okin’s argument aims to challenge. Okin’s argument deals with the relationship between the creation of something and ownership of it. But a thing that is owned by someone can be transferred to someone else, and we need a way to logically bracket this possibility. On standard libertarian views, it is necessary for a (just) transfer from one owner X to a new owner Z that the transfer be voluntary on the part of X.Footnote 3 I will capture all of this in the following definition.
X I-OWNS Y: X owns Y as soon as Y begins to exist and X owns Y unless and until X voluntarily transfers Y or either X or Y cease to exist.
This definition can then be used to define two core commitments of libertarianism.
SELF-OWNERSHIP: No person is i-owned by anyone except themselves.
ACQUISITION BY LABOUR: Any person who makes something using resources she or he owns i-owns that thing.
In what follows, I abbreviate ‘acquisition by labour’ as ABL. Finally, we will initially define NOZICK’S ENTITLEMENT THEORY or NET – Okin’s original target – as the conjunction of self-ownership and ABL.
Before moving on, I want to make a few comments about these definitions. Self-ownership is often defined positively rather than negatively – all persons own themselves, or some such (for example Vallentyne Reference Vallentyne2012, Section 1). There are two problems with this definition. First, without further assumptions about ownership or self-ownership, it does not imply exclusive self-ownership. In particular, it does not imply that I am not owned by my mother. And this will be crucial for Okin’s argument. So we need a negative rather than positive definition: I am not owned by anyone other than myself. I do not believe any libertarians would object to this, since they make the needed further assumptions. Second, Nozick and other libertarians explicitly allow that persons voluntarily sell – or indeed give – themselves as slaves to other persons (Nozick Reference Nozick1974, 331; MacIntosh Reference MacIntosh and Murray2007, 166). Such persons would not, I assume, own themselves after the transfer, even though (if) they did previously. This requires that self-ownership be captured in terms of i-ownership rather than ownership simpliciter.
ABL states the relationship between the creation of something and (i-)ownership over it. Libertarians often claim that self-ownership implies ABL (for example, MacIntosh Reference MacIntosh and Murray2007, 163; Narveson Reference Narveson1988, 66ff; note also that some libertarians reject it; see Vallentyne Reference Vallentyne, Vallentyne and Steiner2000, 5). However, I must admit that I have never understood any of the arguments for this implication. When this inference is relevant to my discussion here, I will grant it for the sake of argument.
Finally, we can reconstruct Okin’s argument. For the sake of simplicity, I will present it in terms of a particular scenario involving two individuals – ‘the mother’ and ‘the child, ’ with pronouns ‘she’ and ‘he’ respectively – rather than in terms of all mothers and children or mothers and children in general. The argument runs as follows:
(1) Assume for contradiction: NET
(2) The mother makes the child using resources she owns.
(3) Hence, the mother i-owns the child. (1, 2)
(4) Hence, the mother does not i-own the child. (1)
(5) Hence, NET is false. (3 and 4 contradict)
The only premise that the libertarian might reject while preserving NET without modification is (2). So we can divide objections into three categories: those that challenge (2), those that modify NET from the definition given above – thereby challenging (1) – and those that challenge the validity of the argument. In the next two sections, I discuss the first two objections in turn. Perrett (Reference Perrett2000) and McKitrick (Reference McKitrick and Machan2006, 85–86) give versions of the third objection, Perrett by appealing to sortal identity theory and McKitrick by appeal to mereology. While technically rather interesting, I must omit discussion of their objections here for the sake of space.
3. Mothers do not make children
The first category of objections to Okin’s argument challenge the premise that the mother makes the child using resources she owns. The clause ‘using resources she owns’ is given by hypothesis, so the challenge is to the claim that the mother makes the child. Machan seems to give this objection when he asserts (without further argument or explanation) that ‘Okin confuses the productive and property-creating Lockean activity of “mixing one’s labor with nature”; and the “labor”; of women as they carry to term a newborn’ (Reference Machan2006, 179n1). Narveson also seems to give this objection:
We should also deny that children are ‘made’ by their parents (as distinct from influencing them quite a lot). What parents do is initiate a process that eventuates in human organisms, which grow up. No mother can ‘make’ any such thing …. (Reference Narveson and Murray2007, 234)
And MacIntosh floats it as a possible response, though not his own: ‘Do women “make”; children in Locke’s sense? Maybe women only have babies; they do not make them, any more than they make their livers or urine’ (Reference MacIntosh and Murray2007, 165, his emphasis; contrast Steiner Reference Steiner1994a, 233).
Taken literally – pregnancy does not count as labour – this objection strikes me as spectacularly ad hoc. Libertarians otherwise seem to think that nearly everything humans do can be construed as labour, and indeed as the kind of labour that supports property rights. Even actions in which the amount of labour ‘mixed’ might be trivial and the agent cannot be said to have ‘initiate[d] a process’ – such as picking acorns and wild apples – are taken to support property rights (Locke Reference Locke and Bennett2008, Section 28; Narveson Reference Narveson1988, 81). A literal interpretation of the objection seems to boil down to the archaic and sexist view that a pregnant woman is merely a passive vessel, containing the fetus while it develops spontaneously (Grebowicz Reference Grebowicz2005, 991-2).
I can think of only one charitable interpretation of this objection: Persons come into existence sometime after gestation and birth, perhaps over the course of several years, starting from mid-infancy and culminating in full personhood at 6, 12, 25 or however many years old. The precise timing is not important for my purposes; the basic idea is that, strictly, the mother makes a something, which something is not a person but becomes one only some time later. Thus the mother does not, strictly speaking, make the child; she only makes a something that will later become the child. Call the something that the mother makes – whatever kind of thing it happens to be, and whenever it becomes the child – ‘the protochild’. Then, this objection claims that the mother does not make the child; she only makes the protochild, which later becomes the child.
Assuming this reading of the objection is close to the mark, I have three replies. First, it strikes me as nearly as prima facie implausible as the literal interpretation. That version of the objection seemed to rely on a profound misunderstanding of pregnancy. This version seems to rely on a profound misunderstanding of parenting: that parents just hand their children food and shelter and other resources and wait for them to grow up; that parenting, in other words, is not active, productive labour. (For an especially striking contrast, see Kittay Reference Kittay1999, especially chapters 6 and 7.)
In counter-reply, the libertarian might grant that raising a child from infant to person does indeed require active labour from caring adults. But, he might continue, this labour is spread out among too many people for the mother, by herself, to claim ownership: she is doing some of this work, but so is the child’s father, other family members, teachers, and so on. In a cliché, it takes a village (or many villagers) to raise a child. As MacIntosh puts it, ‘The process of making children into persons is not done solely by mothers, but usually also by fathers, friends, teachers and even by the children themselves, which could complicate things for Okin’ (2007, 166). All this is right, but it makes things worse, not better, for the libertarian, since it seems to imply that all of these individuals, not just the mother, made the child; and thus all of these other individuals can claim either collective ownership or several individual ownership shares over the child. Either is inconsistent with the child’s self-ownership. On this counter-reply, the child seems to be the property of the whole community in which it was raised, not even only one of his parents.
My second reply introduces the second part of Okin’s dilemma. The objection – that the mother only makes the protochild – still implies that she does indeed make the protochild, and thus she i-owns the protochild. Assuming she does not voluntarily give the protochild away, etc., she owns the protochild at the time (or entirely during the time) at which it becomes the child. Now, we seem to have two possibilities for what happens when the protochild becomes the child: either she i-owns the child, or the child i-owns himself. In the first case, we have step (3) of the argument once again. In the second case, it seems that some of her property – namely, the protochild – has been taken from her involuntarily. Her property has either been destroyed (if the protochild ceases to exist when the child comes to exist) or appropriated by the child (e.g., he has appropriated his physical body from her), in either case without her consent.Footnote 4
Now, NET as defined above does not seem to imply that there is anything wrong with taking someone’s property from them involuntarily. Indeed, the definitions above say nothing at all about transfers. Libertarians might claim that self-ownership implies that ownership cannot be transferred against the will of the previous (rightful) owner. Or they might simply include something like Nozick’s principle of justice in transfer – which does clearly imply this – in an expanded version of NET. But, in either case, libertarians – especially right-libertarians – generally wish to claim that it is unjust for one person to take another person’s property without the latter’s consent. This defense of private property is arguably the primary raison d'être (or perhaps raison de croire) for right-libertarianism (Cohen Reference Cohen1981, 3–5, 10–11). In whatever way, I am happy to let libertarians include these strong private property rights in NET. But then they are caught in a dilemma: they must either accept step (3), and thus the reductio ad absurdum of NET, or they must accept an involuntary transfer of property from the mother to the child, which (since it now includes strong private property rights) would also amount to a reductio ad absurdum of NET. This is the argument that I call OKIN’S DILEMMA. We shall see it several times over the course of this paper.
My third response is that the approach taken by this objection, when generalized, undermines all claims to property rights in the products of agriculture and animal husbandry. The farmer and rancher, after all, do not themselves make the corn that grows in their fields and the next generation of livestock; like the parents of the child (on this objection), they initiate and facilitate a process that eventuates in the reproduction of these organisms (compare Curchin Reference Curchin2007). So if, as this objection would have it, is it incorrect to say that that the parents create the child, then in exactly the same way it is incorrect to say that the farmer creates the corn or the rancher creates the calf. And if ABL is the only way to acquire something that was not previously owned, then there is no way for anyone to acquire the new corn and cattle.Footnote 5 Given the prominence of agriculture as examples in the literature on libertarianism and property, I take it that libertarians would not accept this conclusion.
4. The except-for-people clause
The second category of objections – and the category that seems to be the most prominent in the literature – rejects the characterization of the core commitments of libertarianism as aforementioned, namely, as the conjunction of self-ownership and ABL. Specifically, these objections claim that ABL has what I will call an EXCEPT-FOR-PEOPLE CLAUSE. That is, ABL should be replaced with
ABL * : Any person who makes something, except for a person, using resources she or he owns, i-owns that thing.
Let NET * be the conjunction of self-ownership with ABL*. Then, with NET* taking the place of NET in Okin’s argument, premise (2) no longer implies that (3) the mother i-owns the child. McKitrick considers this response to Okin’s argument, though her use of the hypothetical mood makes it unclear whether she endorses it herself:
One could accept a Lockean view of how one comes to own unowned nonsentient property, yet not extend that to owning humans. One could consistently hold that there are certain kinds of beings that cannot legitimately be owned. Rational agents seem a likely candidate. Along these lines, one might hold that if a human being mixed his [sic] labor with silicon and electronic elements and built an artificial intelligence that rose to the level of a rational agent, that claims of ownership would become problematic. The same reasoning would justify denying women ownership of their adult offspring. (Reference McKitrick and Machan2006, 86)
And MacIntosh’s primary response to Okin is certainly in this category:
However [the child] has an argument too: people have Lockean rights; so in our treatment of them we must respect side constraints. No one, then, may come to own persons simply by dint of the act of making them; for to make a person is to make a thing with Lockean rights, one which has S [self-ownership] unless it consents to ownership by another. What grounds a person’s right to negative liberty (and so to S) is not that someone gives or concedes it; rather, as we saw, it is her being owed it simply due to her desiring to exercise it, having a conception of the good life. (Reference MacIntosh and Murray2007, 166)
Earlier in this piece, MacIntosh gives a brief encapsulation of his argument for including such an except-for-people clause:
This is why offspring own themselves: were they owned by their mothers, they would have no liberty; so a greater liberty for all would result from individuals being self-owned upon becoming persons, which is why the rights of people to what they make cannot be so extensive as to permit them to own people they make. (Reference MacIntosh and Murray2007, 159)
I will discuss this argument briefly below. But my major response to this objection does not depend on the reasons for including the clause. Instead, I argue that the objection – whatever reasons are given to support it – leads to the other horn of Okin’s dilemma. (In personal communication, Narveson gave a version of this objection as well.)
In the next subsection, I give three replies to this objection, and then make a brief comment on uptake. The subsequent subsection considers a variation on the objection, which is relevant to the abortion literature. And the final subsection addresses Hillel Steiner’s response to this problem, which has elements of both the mothers-do-not-make-children response (considered in the previous section) and the except-for-people response.
4.1 An ad hoc exception?
First, simple versions of this objection strike me as special pleading, or at least a little ad hoc, from the perspective of ABL: Why should persons be treated any differently from any other manufactured good? Or, why should labour that results in persons be treated any differently from any other kind of labour? Note that, in most of the quotations above, there is very little argument given to support including the except-for-people clause.
The libertarian might claim that, since self-ownership implies ABL – or some version of it, at least – self-ownership claims have priority over property rights claims. But this confuses logical strength with an ethical notion of priority. (Narveson seems to make this mistake; see Reference Narveson and Murray2007, 234.) By way of comparison, any logical contradiction implies ABL, but surely this does not mean that the contradiction enjoys some kind of ethical priority! Further, granting for the sake of argument that self-ownership does imply ABL and the self-ownership is consistent, then there cannot be any logical disagreement between the two, and so there will be no need to determine which one has priority. Put contrapositively, the prima facie tension between self-ownership and ABL established by Okin’s argument gives reason to believe that they are actually logically independent.
Unlike other arguments for the except-for-people clause, Duncan MacIntosh’s argument does not strike me as ad hoc. MacIntosh’s argument – quoted a few paragraphs ago – starts with a commitment to maximal negative liberty. Indeed, in his view, the fundamental right is to maximal negative liberty, a ‘non-overridable right to such liberties as do not interfere with those of others’ (MacIntosh Reference MacIntosh and Murray2007, 163). MacIntosh is still committed to self-ownership – and hence qualifies as a libertarian by Vallentyne’s definition (n. 1 above) – but takes self-ownership to be supported by maximal negative liberty. Since, presumably, some version of ABL is a separate implication of maximal negative liberty, it is logically possible for ABL and self-ownership to conflict. (In other words, MacIntosh might take self-ownership and ABL to be logically independent of each other.) This is the case in Okin’s argument, and MacIntosh’s objection is straightforward: since the mother’s ownership claim over the child would seriously interfere with his liberties, and his right to these liberties is the fundamental one, her ownership claim must fail. She can own anything she makes with resources she owns, except a person.
My second reply to this objection develops Okin’s dilemma again. The mother certainly makes something; again, call this something ‘the protochild’. Depending on the protochild’s relationship to the child, it seems that either the mother owns the child by virtue of owning the protochild – and we have step (3) of the argument after all; or the child appropriates the protochild from the mother against her will; or the child, in coming to be, causes the protochild to cease to exist, again against the mother’s will. So either step (3) is intact or there is an involuntary transfer of property from the mother to the child.Footnote 6
My third reply does not directly rehabilitate Okin’s argument, but instead points to a tension between two attitudes that I would expect libertarians to have. On the one hand, persons, as distinct from mere things, have a special ethical status for libertarians; this is why persons have inviolable rights, are self-owners, and so on. So I would expect libertarians to think that the activity of creating new persons is a special and important kind of activity, and thus deserving some sort of special regard. This special regard need not – indeed, shouldn’t, to avoid Okin’s argument – imply that mothers own their children. But it should imply that the work of pregnancy and childrearing is different from other kinds of human activities, and different in a way that makes this work more important than work that produces mere things. On the other hand, the approach taken in this objection deprives mothers of the entitlements and benefits that creators of even mere things enjoy, and does not confer on them any replacement entitlements and benefits. It thereby seems to treat the activity of creating new persons as less important (that is, of less moral significance) than activities that produce mere things (since producing mere things, but not creating new persons, confers rights). As Curchin puts it:
Left libertarianism only makes sense if you treat the labour of bearing and raising children as a mere hobby, rather than as real and necessary work …. [T]he vitally important work of creating future generations goes unrewarded and unrecognised, and it is simply taken for granted that enough people will find time for parenting in their lives. (Reference Curchin2007, 486n6)
Again, this reply does not directly rehabilitate Okin’s argument from this objection, as it relies on further assumptions that libertarians might consistently deny. For example, libertarians could claim that the child has obligations to compensate the mother for the resources she has spent on making the child. As it were, the child must pay the mother back for appropriating her property without her consent. The compensation could be in-kind rather than monetary, that is, the child could have obligations to care for the mother when she is elderly or sick and dependent on others. The compensation might or might not be enforceable through the coercive apparatus of the state. Bracketing the details, perhaps some proposal along these lines would go towards giving reproduction the special regard that it deserves.
Somewhat similarly, a libertarian might propose that the mother is compensated through a sense of self-satisfaction, social recognition, the love of the child, the good of the parent–child relationship or in some other intangible way. (Thanks to an anonymous reviewer for suggesting this possibility.) But this seems far too contingent. Perhaps, like more than a few women, the mother finds pregnancy and birth unpleasant, unsatisfying and more than a little bit painful (if not literally life-threatening). She may be socially and professionally marginalized as moving on the ‘mommy track’. On the other side, perhaps the child finds the mother’s begrudging and uncaring attitude towards him – after all, she would have kept him as a slave or organ donor, if not for his self-ownership rights – and thus feels no love for her, and the two have the kind of terrible relationship one would expect. The libertarian might suggest that someone has an obligation to ensure that the mother actually enjoys these proposed intangible compensations. But then we would need to clarify who has these obligations, whether someone besides the mother can possibly satisfy some of them (e.g., the mother’s sense of self-satisfaction), and whether these obligations deeply conflict with libertarian anti-paternalism (that is, these obligations would seem to require significant meddling in the lives of the mother and child).
In any case, while some libertarians have considered the obligations that parents have to their children – including, in some cases, positive obligations to provide them with food, shelter, education, and so on (for example, Vallentyne Reference Vallentyne, Archard and Macleod2002) – I have found only one instance of a libertarian discussing the obligations that children have to their parents. And this is a single paragraph, arguing that minor children have an obligation to obey their parents (Machan Reference Machan2006, 217).
In general, to the extent that libertarians have taken up the concerns raised by Okin’s argument at all – reproduction, dependency, and care work – they have generally treated it as an occasion to entrench their commitments without modification, rather than as an opportunity to expand and develop their intellectual tradition (contrast MacIntyre Reference MacIntyre and Brinkman1999). In this respect, McKitrick (discussing care work more than reproduction) is a happy exception:
Okin’s diagnosis still stands, and at this point, it seems to me that libertarians have basically two ways to respond. One is to grant that under the present gender system, people making free choices about work and family results in a situation in which women are ‘differently advantaged’ and that it would require substantial interference with personal liberty to alter this situation. But, the libertarian argues, such interferences are unacceptable. So, unfortunately for women, that is just the way it is. And unfortunately, many libertarians seem to have this response. They hail the traditional gendered division of labor as marvelously efficient or ‘natural’ or acquiesce to it as inevitable or beyond question, and they seem oblivious or insensitive to the vulnerabilities it creates. (Reference McKitrick and Machan2006, 95)
A better response, McKitrick thinks, would be to offer a libertarian critique of the gender status quo, and develop proposals for addressing the problems of that status quo without infringing on individual liberty.
[L]ibertarians have to offer other strategies for dealing with the cycle of inequality. The problem, for rhetorical purposes anyway, is that there is not, nor can there be, a grand libertarian plan for coordinating and changing behavior, of employers and individuals, both in their private and professional lives. (McKitrick Reference McKitrick and Machan2006, 98)
I shall return to this thought at the end of the paper.
4.2 Tacit consent
A specific version of the objection that I am considering in this section might be thought to avoid all of the replies given in the last few pages. This version modifies ABL by conjoining with it a claim that, when the mother makes the child, she automatically or tacitly consents to give up her ownership claim immediately before the child begins to exist. Thus, concludes the libertarian, the transfer of ownership is actually voluntary, and both horns of Okin’s dilemma have been avoided. This version of ABL might be defended, in part, by the thought that parenthood is voluntary (compare Machan Reference Machan2006, 219); then, since the mother has consented to have the child, it might be claimed to follow that she has consented to give up her claim to own the child. As with the modification of NET suggested earlier, automatic consent prevents (2) from implying (3), and the transfer of ownership is not involuntary.
This line of thought is similar to a line of thought that appears in the abortion debate: since (goes the argument) among the foreseeable actions of sex is pregnancy, a woman who consents to sex also automatically or tacitly consents to carry the resulting pregnancy to term. (For an extensive critical discussion of this line of thought, see Boonin Reference Boonin2002, Section 4.3. Citations to examples are given by Boonin at 148n6.) There are two standard problems with this line of thought, and the argument of the last paragraph also seems to have these problems. First, in many cases, sex is not voluntary. In the USA, on the order of 20 million women have been raped during their lifetime, with on the order of 1 million raped in 2006. While presumably only some of these attacks resulted in pregnancy, and only some of these pregnancies were carried to term, this still means that some pregnancies and some children were conceived without any consent at all (Kilpatrick et al. Reference Kilpatrick, Resnick, Ruggiero, Conoscenti and McCauley2007). (The lack of access to abortion in many parts of the USA also means that some pregnancies are involuntary and so some children are involuntary – even in cases where the sex is voluntary.)
Second, even in cases where the sex is voluntary, this does not imply automatic consent to any of the consequences. This is especially easy to see if a sexually transmitted disease is among the consequences: the fact that I consent to the sex does not imply that I consent to contract, say, HIV. (Thanks to Samantha Brennan for suggesting this example.) This is not to say that I do not have ethical obligations with respect to these consequences, such as obligations to notify potential sexual partners that I have HIV; rather, the obligations that I have cannot depend on my consent. Specifically, it is entirely consistent for a woman to consent to sex without consenting to carry the resulting pregnancy to term, and likewise entirely consistent for her to consent to having the child without consenting to give up her ownership claim.
The libertarian – like the abortion opponent – might claim that the obligations with respect to the consequences of the action, even without consenting to them, are sufficient for their objection: by virtue of consenting to sex, the abortion opponent says, the pregnant woman has an obligation to carry the pregnancy to term, even if she does not consent to do so; and similarly, the libertarian claims that, by virtue of consenting to the pregnancy, the mother has an obligation to give up her ownership claim over the child, even if she does not consent to do so.
At this point, there is a relevant disanalogy between many libertarians and abortion opponents. Libertarians generally recognize a distinction between enforceable and non-enforceable (that is, through the use of coercion) obligations, and generally claim that the only enforceable obligations are voluntary ones (Narveson Reference Narveson2003). (While I myself reject the second conjunct – it seems to founder clearly on the obligation to notify one’s partner if one has HIV – I am willing to grant it for the sake of the current discussion.) Then any obligations that the mother has to give up ownership over the child would seem to be non-enforceable. No one can force her to give up her ownership claim over the child – not even the child himself, or the state that would otherwise protect his self-ownership rights. So, even with these obligations, if she does not voluntarily give up ownership, she still i-owns the child. In this case, (3) still holds. And the child’s self-ownership claims – like other non-enforceable obligations – are rather toothless.
The libertarian can get around the last paragraph, of course, but only by making some non-voluntary obligations enforceable. That is, only by requiring an involuntary transfer from the mother to the child, albeit one required by the mother’s obligations to the child. Once again, the libertarian is caught in Okin’s dilemma, between accepting (3) and accepting involuntary transfers of property.
4.3 Steiner’s version
In An Essay on Rights, left-libertarian Hillel Steiner offers a response to ‘the paradox of universal self-ownership’ (Steiner Reference Steiner1994a, 237–248; I thank an anonymous reviewer for bringing Steiner’s response to my attention). While Steiner attributes this paradox to Robert Filmer, he notes in passing that it is similar to Okin’s argument (Reference Steiner1994a, 245n21). Steiner’s response combines the two responses that I have considered already: because (in a sense to be explained below) the child is not made by the mother, there is room for an except-for-people clause.
The first move is to argue that the ‘germ-line genetic information’ used to make the child is a natural resource, that is, it is not made by the mother or any other person. As Steiner puts it, ‘[The child] is not fully the fruit of his parents’ labour. For his production required them to mix their labour with natural resources in the form of germ-line genetic information transmitted from his grandparents' (Steiner Reference Steiner1994a, 248, my emphasis). In terms of ABL, the mother makes the child using resources she owns, but not using only these resources. Hence, the mother’s ownership claim over the child may be ‘encumbered’, that is, may be overridden. Continuing the previous quotation:
Hence denying his parents full liberal ownership of him doesn’t amount to a denial of [ABL]. That is, encumbering their ownership is permissible. And what the children’s rights discussion does is to mobilize that permissibility to sustain [self-ownership], by implying that this encumbrance consists in the liability of their ownership to expiry on the occasion of [the child] attaining his majority. (Steiner Reference Steiner1994a, 248)
Some later articles, which present the same argument, and an exchange with Curchin, help to clarify this argument (Steiner Reference Steiner, Dyson and Harris1994b, Reference Steiner, Burley and Harris2002, Reference Steiner2008; Curchin Reference Curchin2007). On Steiner’s left-libertarianism, individuals have a right to an equal share of the world’s natural resources. Individuals may claim ownership of these natural resources, but only if they compensate everyone else (Reference Steiner1994a, 235–236, 267–272; Reference Steiner, Dyson and Harris1994b, 71). As Curchin glosses Steiner’s view, ‘Unilateral appropriation of unappropriated resources is permissible so long as compensation is paid to those who are deprived of the value of the resource to which they have a claim’ (Reference Curchin2007, 486). Furthermore, if an individual does not own all of the resources she or he uses to make something, then she or he cannot claim i-ownership of it:
We each own the fruits of our labour inasmuch as [read: if, and only if] all the factors entering into their production are either things already owned by us or initially unowned things amounting to no more than an equal portion of them. (Steiner Reference Steiner1994a, 236)
In particular, the mother does not own the germ-line information used to make the child, and so his self-ownership claim overrides her claim from ABL.
The inference to the previous sentence is flawed. The universal claim is that, for all natural resources, either they are unowned or the original owner must compensate everyone else. This implies that either the germ-line information is unowned or the person who first claimed ownership over it must compensate everyone else. This is one of Curchin’s two objections to Steiner:
Steiner gives no reason why parents of children cannot simply compensate all others for the appropriation of the necessary germ-line genetic information, and then freely dispose of them however they wish for as long as they wish. This leaves us with no solution to the paradox of universal self-ownership. (Curchin Reference Curchin2007, 492)
In his response, Steiner seems to recognize the force of this objection. However, his response is simply to reassert the child’s self-ownership claim:
A theory that vests every moral agent with the right of self-ownership is thus a theory that does offer a reason why the encumbrance on the ownership of minors by their parents should not be dischargeable [through compensation]. Denying that persons who are minors themselves have rights, it clearly offers a reason why their rightless state should be one of a duration limited to the period of their minority and, hence, why the form of the encumbrance on their parents' ownership of them should reflect this. (Steiner Reference Steiner2008, 348)
Earlier, Steiner’s argument was for a claim that mothers do not make (every part of) their children. At this point, it seems clear that he is also relying on an except-for-people clause in ABL. In this way, his response to Okin’s argument combines both of the strategies that I have already discussed.
There are several problems with Steiner’s argument. First and second are Curchin’s two objections. Curchin’s first objection is that germ-line genetic information is a non-excludable good (in the economic sense) and that Steiner’s restrictions on the appropriation of natural resources only applies to excludable goods (Reference Curchin2007, 488–491; Steiner Reference Steiner2008, 352–353). Since this point is particular to Steiner’s view, I will not discuss it here.
Curchin’s second objection, and Steiner’s response, has been discussed earlier: per Steiner’s requirements for natural resources in general, the mother might appropriate the germ-line genetic information simply by compensating everyone for it. Steiner’s response seems to boil down to an except-for-people clause, and so seems to be vulnerable to all of the problems discussed in section 4.1.
A third problem is related to this second one. Steiner’s argument depends heavily on the claim that germ-line information is never owned by anyone. To the picture of the mother and the child, add all the matrilineal ancestors: the grandmother, the great-grandmother and so on. According to Steiner’s analysis, the germ-line information used by the mother to make the child comes from these ancestors, and can be traced back to some non-person ancestors (who, since they were not persons, were natural resources) (Reference Steiner1994a, 247). Through all of these generations, the germ-line information remains unowned. This must be the case, since otherwise (we could suppose that) the mother owns it, and so can claim i-ownership over the child. Furthermore, the child cannot own it either, for the same reason: if he could, then it could not be used to protect the self-ownership of his descendants. But then this means that the child’s self-ownership claim is based on a natural resource that existed long before he did and that he does not, and indeed cannot, own.
This last sentence does not strictly undermine Steiner’s argument. But I take to be highly implausible. And it seems to clash with the standard intuitions and claims that libertarians offer concerning self-ownership. First, self-ownership is usually tied to moral personhood and autonomy. But – unless we make the mistake of conflating personhood and humanity (Warren Reference Warren1973; see also the next section) – the connection between germ-line genetic information and personhood seems to be too thin.
Furthermore, Steiner’s analysis implies that we, as self-owners, do not own all of the parts of our bodies (Reference Steiner1994a, cf. 248–9n26). It is true that no one else can claim ownership over the unowned parts, but this still falls short of the intuitive meaning of ‘self-ownership’. This also seems to imply that we do not actually have ‘full liberal ownership’ over our bodies: since my body contains things that I do not and cannot own, I do not have the right, e.g., to destroy these things, and so have no right to destroy my body, or at least those parts of it that contain my germ-line information. Besides suicide, this seems to rule out certain kinds of self-sterilization and birth control, and may significantly complicate the libertarian analysis of abortion.
Fourth – and perhaps undermining several of the proceeding problems – the appeal to germ-line genetic information seems to be superfluous. Recall the exchange between Curchin and Steiner: Steiner argues that, since the information is unowned, the mother does not make the child using only resources she owns. Curchin points out that the mother could come to own the information. And Steiner’s response is that the child’s self-ownership claim takes precedence. But then it makes no difference whether the germ-line information is owned or not. In other words, Steiner responds to Curchin’s critique of his mothers-do-not-make-children view by dropping it and asserting an except-for-people clause. He could simply have asserted that from the beginning, without the unnecessary and problematic appeal to germ-line genetic information.
Last, but certainly not least, is Okin’s dilemma. Bracketing the preceding problems, and given that the mother does not own all of the resources used to make the child, Steiner grants that she does own some of them. By hypothesis, the mother does not voluntarily transfer these to the child. But they must be transferred to him if he is to be a self-owner. Hence, his solution requires an involuntary transfer of property.
This is easiest to see if we assume that the unowned genetic information is confined to the child’s gonads (gamete-producing organs). Given this assumption, we can conceptually distinguish (a) the part(s) of him that the mother made using only resources that she owns from (b) his gonads. Recognizing the child’s self-ownership claim requires an involuntary transfer of all of his non-gonad parts to him from the mother.
More plausibly, the unowned information has been utilized to make every part of his body in such a way that there are no parts of his body made using only resources that the mother owns. In this case, the trick of the last paragraph will not work. Call the set of all of the resources owned by the mother (that will be used to make the child) ‘the protochild.’ That is, the child is made by combining the protochild with the germ-line genetic information (in the appropriate way). By hypothesis, the mother owns the protochild. When the child comes to be, there seem to be two possibilities: either the protochild becomes the child (in an Aristotelean sense, it is ‘informed’ by the genetic information) or the protochild ceases to exist. If the protochild becomes the child, then, when the child becomes a self-owner, the mother’s property – the protochild – is transferred from her involuntarily. If the protochild ceases to exist, then her property has been destroyed, again against her will. In either case, she is involuntarily deprived of her property. (Compare this instance of Okin’s dilemma to the instance in the text at note 4)
4.4 Summary
Since this has been a somewhat long and articulated section, permit me to recap briefly. The libertarian objected to Okin’s argument on the grounds that ABL should include what I called an except-for-people clause, which implies that persons cannot be owned by the people who made them. My major response to this objection is to concede that this does indeed avoid Okin’s original argument; however, in one way or another depending on the details, it requires an involuntary transfer of property from the mother to the child. Thus, this objection is vulnerable to the broader argument that I am calling Okin’s dilemma.
5. Conclusion
My defense of Okin’s argument in this paper has not been completely unqualified, for the libertarian can resist the inference to the crucial claim (3) that the mother i-owns the child. However, doing so requires the libertarian take the other horn of what I called Okin’s dilemma: the child has the right to appropriate at least some of the mother’s property without, or even against, her consent. This would seem to threaten right-libertarianism with a tumble down the slippery slope to left-libertarianism, and raises serious problems for certain versions of left-libertarianism as well. In the remainder of this paper, I point out some features of the terrain of this slope.
First, put generally, the basic concession the libertarian must make (in order to avoid Okin’s original argument) is that the requirements of self-ownership can override property claims, in at least some cases and even when these are supported by other libertarian principles. Then, an obvious strategy for the egalitarian is to claim that self-ownership has further requirements, which override further property claims. To take a strong example, the egalitarian could try to argue that self-ownership implies that I have a right to appropriate property nominally owned by others without their consent, if it is surplus to their survival but necessary to maintain my own (compare Sterba Reference Sterba1998, 44–56). Some left-libertarians might be happy to move in this direction. But clearly, right-libertarians would want to resist.
A right-libertarian could respond by claiming that self-ownership does not imply this much; in other words, by claiming that self-ownership is more ‘formal’ and less ‘substantive’ than this argument requires. At this point, the debate may move on to familiar ground, with egalitarians advocating substantive conceptions of freedom and self-ownership and right-libertarians advocating formal conceptions of these concepts. Note, however, that a more formal account of self-ownership would perhaps not support more substantive property rights.
Another egalitarian strategy would be to extend the argument given by the libertarian in §4, namely, the introduction of the except-for-people clause. That is, perhaps ABL should include other except-for-X clauses. For example, the egalitarian might try to make a case for the following:
ABL**: Any person who makes something, except for a person or something required by another person for the latter’s continued survival, using resources she or he owns i-owns that thing.
This move is not as strong as the example floated two paragraphs ago, but it is, as it were, pointing in the same direction. Even stronger clauses would be more favourable to egalitarianism, such as except-for-things-others-require-to-flourish and except-for-what-goes-beyond-strict-equality clauses.
I take it that many libertarians would find these moves implausible. But it seems to me that Okin’s dilemma puts the burden of proof on the libertarian rather than the egalitarian: given that property rights can be limited in such-and-such ways (whichever horn of Okin’s dilemma the libertarian prefers), then why not in these other, analogous ways?
Next, as a logical point, Okin’s dilemma shows that NET – either the original version or the starred version – does not imply that justice forbids all involuntary transfers. Indeed, it shows that NET implies that at least some involuntary transfers are permitted by justice, and further that some involuntary transfers are required by justice. As mentioned earlier, a major attraction of right-libertarianism – at least for people who are attracted to it – is the seemingly strong protections it gives to property claims. Okin’s dilemma logically undercuts this attraction, and thus creates an urgent need for right-libertarians to examine the variety of kinds of involuntary transfers and to determine which ones justice (on their view) requires, permits and forbids. It also indicates that simply appealing to self-ownership would not seem to be sufficient to make these determinations. (This point might be construed as a generalization of what is sometimes called ‘indeterminacy’; see Fried Reference Fried2004; Vallentyne, Steiner, and Otsuka Reference Vallentyne, Steiner and Otsuka2005; Fried Reference Fried2005.)
Finally, we return to a point raised by McKitrick above, the problem of ‘offer[ing] other strategies for dealing with the cycle of inequality’. In order to avoid Okin’s dilemma, the libertarian requires the involuntary sacrifice of some people’s property claims. But which people? In the gender status quo, this burden falls on women much more than men. This should raise concerns among libertarians about both women’s (external) property and their self-ownership. Women must make significant material sacrifices to care and raise their children into autonomous self-owners. And doing so often requires that they sacrifice their other lifelong plans and projects – they must take time away from their careers, which limits both their possibilities for advancement and their earning power. Okin’s dilemma indicates that there are good (libertarian) reasons for requiring this sacrifice – it is necessary for the future generation to be autonomous self-owners. But the formal egalitarianism of libertarianism – equal negative liberty for all – suggests that the burden of making involuntary sacrifices should not be required of only mothers, or perhaps that other people should be required to sacrifice some of their property in such a way that the burden is shifted – one is tempted to say ‘so that the burden is distributed fairly’.
One egalitarian proposal would be for the state to pay full-time caregivers a living wage, using tax money collected from citizens at large (all of whom, of course, were raised by caregivers when they were children) (compare Alstott Reference Alstott2004; Bou-Habib and Olsaretti Reference Bou-Habib and Olsaretti2013). An even stronger, socialist proposal would be for much of the work of caregivers – cleaning, food preparation and early childhood education – to be professionalized and available freely to anyone who wishes to take advantage of these services, paid for by the state (Davis Reference Davis1981, chapter 13). Clearly any right-libertarian would balk at the first proposal and be downright horrified by the second. But then the challenge is laid at their feet, to develop less-statist schemes that do not require mothers, and only mothers, to make involuntary sacrifices.
Acknowledgements
Thanks to Jim Sterba, Hallie Liberto, Jennifer McKitrick, Jan Narveson, Duncan MacIntosh, Samantha Brennan and anonymous reviewers for comments on earlier versions.
