The Family: Friend or Foe?
The family is one of the issues that divide liberals from conservatives. In general, conservatives tend to see private associations — the family, the church, the corporation — as bulwarks of freedom against the state. Few conservatives question the need for a powerful state apparatus, but they insist that it operate in the service of private associations rather than supplant them. Liberals, by contrast, are more likely to see these private associations themselves — family, church, corporation — as threats to autonomy, and to view state intervention as a guarantor of freedom against the oppressive tendencies of private associations. Few liberals seek to abolish such associations, but they do want to subordinate them to the state — just as conservatives want to subordinate the state to the private associations.
This dispute, like so many between the right and the left, is one that libertarians have to sit out. Libertarians agree with conservatives that the state is the chief threat to freedom, and that private associations must be protected from governmental interference. But libertarians are also sensitive to the potential for oppression in private associations, especially when these associations are the beneficiaries of government favoritism. The conservative approach of putting the state in the service of family, church, and corporation simply hands the reins of power to these institutions, which are no more to be trusted with such power than are governmental bureaucracies.
Conservatives see the family as the fundamental unit of society. But for libertarians the fundamental unit is the individual. Hence libertarians have traditionally been ambivalent about the family (as about its kin, the church and the corporation). The family, as a locus of influence and loyalty separate from the state, is certainly something that opponents of centralized power are eager to defend. But on the other hand, libertarians are keenly aware that the family has not always been a sphere of individual freedom, particularly for women and children. How, then, should libertarians think about the family?
The Origin of the Family
In biological terms, the family originates in the need to nurture offspring. The lowest animals often have no families, because they do not need them; they come into the world with a full adult repertoire of survival behavior genetically programmed into them. In many insect and fish species, the parent is either dead or long since absent by the time the young organism hatches. But the learning-to-instinct ratio is higher in more intelligent, more flexible species, and such species therefore need a longer period of childhood. In such species, one or both parents stay with the young until this vulnerable learning period has passed. This is the most primitive form of the family.
This first family is often ephemeral. In many animal species, the family unit dissolves as soon as the young are full-grown; from then on, offspring and former mates are treated in more or less the same way as any other member of one’s species.
But the evolutionary process is resourceful. A trait that initially emerges to meet one need, may then be pressed into service to meet another. There are evolutionary advantages to maintaining a cooperative relationship among family members beyond the point needed to ensure the continuation of the species. And with the highest animals, not only biological evolution but cultural evolution can come into play (e.g., a cat who is raised to regard mice as playmates rather than prey may in turn raise a whole generation of peacenik cats).
Among humans, the family still serves the original function of childrearing, but it has acquired a robust range of new functions as well, serving both the economic and the emotional needs of its members. The family has grown beyond its original biological basis, thus dramatically increasing the number of possible family structures.
A parallel can be made to language. Presumably, language first evolved in order to convey information vital for survival, such as “There’s a sabretooth tiger behind that outcropping” or “Don’t eat those, they’re the mushrooms that made me sick before.” And language still serves that function. But today language also serves a broad range of spiritual needs whose relation to physical survival is tenuous at best. To condemn (as many conservatives do) family relationships that are not for the purpose of childrearing is like condemning Shakespeare’s Hamlet for not telling us where the sabretooth tiger is.
In his book The Psychology of Romantic Love, libertarian psychologist Nathaniel Branden traces the institution of marriage from primitive times to the present. In ancient times, he points out, it was expected that marriage would be based on economic and social considerations, not on love; the phenomenon of romantic love was regarded as an antisocial obsession, an unfortunate madness that people sometimes fell into. In the Middle Ages, marriage for love remained socially impracticable for most, but the literature of the time (in opposition to official Church doctrine) began to celebrate romantic love as one of the highest human experiences, and to portray marriage not based on love as an oppressive institution. But the mediæval romancers were not social revolutionaries; rather than conceiving of a fundamental change in the nature of marriage, they generally portrayed romantic love as glorious but adulterous and tragically doomed. It was the rise of industrial capitalism, Branden argues, that first gave women enough economic independence to postpone marriage, and this greater equality, he says, along with the capitalistic ethic of individualism, is what led to the expectation in present-day society that marriage will ordinarily center on romantic attachment above everything else. To the extent that this change is a good thing, as I think it is, human beings have managed to make out of the sexual pair-bond something superior to what nature originally provided.
Unfortunately, the human intellectual and social skills that allow us to improve on nature, also allow us to do worse than nature. Historically, human families have often been oppressive and exploitative institutions, in a way that animal families do not seem to be. The purest example of this is the Roman family, in which the male head of household (the paterfamilias) was legally entitled to put his wife and children (even grown children) to death. This aspect of family relationships is called patriarchy (“father-rule”), signifying the subordination of wives to husbands and of children to parents. Those who defend patriarchy as “natural” often point to the animal kingdom as a model; but traditionally, parental authority and sexual inequality have been far more pronounced in human societies than in most animal societies. Recent political developments — springing in part from the libertarian urge to subordinate patriarchal authority to individual rights, and in part from the welfare-liberal urge to subordinate patriarchal authority to that of the state — have weakened the institution of patriarchy, but not eliminated it entirely. In her valuable book Justice, Gender, and the Family, Susan Okin points out some of the ways in which contemporary society still systematically reinforces patriarchal family structures.  How might families in a truly free society develop beyond this patriarchal paradigm?
Family Structures as Voluntary
As mentioned above, human reliance on learning over instinct allows us to progress beyond the limitations of our genetic programming, thus increasing the number of family structures available to us. Kinship relations and procreative unions, while they will remain one important basis for family structures, are no longer the only such basis. Yet most human societies have laws mandating only certain sorts of family structure, and forbidding others. Conservatives argue that such laws are necessary if society is not to collapse; they see heterosexual monogamy as a prerequisite for a healthy culture, and thus as an institution deserving legal protection. Yet conservatives also see themselves as defenders of the Western cultural tradition originating with the ancient Jews and Greeks, two groups whose commitment to heterosexuality (in the case of the Greeks) and monogamy (in the case of the Jews) is hardly notable; were their cultures defective?
A libertarian legal system would not grant special protection to certain types of family, but would allow any arrangement that was consensual and peaceful. Monogamy or polygamy; heterosexual or homosexual marriage;  extended families or nuclear families or single-parent families;  group marriages (sexual or nonsexual) — any of these relationships would be permitted. It is a mistake to suppose that there is just one kind of family structure that is right for everybody; and even if there were, it would be a mistake to think we could be justifiably confident that we had found it if we did not allow the discovery process of competition among alternative family structures to operate freely.
Another way in which libertarian society would differ is in the greater variety of marriage contracts that legal institutions would be willing to recognize and enforce. (I say “legal institutions” rather than “the state,” to leave open the possibility of an anarchist society.) There would be some limits here, however; I have argued in previous articles that indentured-servitude contracts are not legitimate on libertarian principles, and the same reasoning would apply to contracts forbidding divorce. Many statists (originally on the right, but they are now being joined by voices on the left) argue that marriage laws should make divorce more difficult, primarily in order to “protect children.” While this might have worked in the days when social mores were different, the result of such legislation if it were implemented today would be, not unhappy couples staying together, but unhappy couples separating without divorce, and moving in with new partners without remarrying. How this would make the children any better off is unclear. (Conservatives say we should try to encourage stable marriages by “restoring the stigma of illegitimacy.” The notion that this would benefit the children involved is still more bizarre.) In any case, the parents as sovereign individuals have the right of free association and disassociation, and to force them to remain in a relationship with someone they no longer love is tyrannical. (I also think the idea that parents should stay in a phony marriage for the sake of the children is immoral, a kind of sacrilege against marriage itself — though of course the parents have the right to make such a decision if they choose.) But, leaving aside no-exit contracts, libertarian legal institutions would respect a greater variety of marriage contracts. Couples who find themselves in a dispute not covered by their contract, or who do not have a contract, may be treated by the courts as if they had signed whatever the “default” contract is in the society — though they can always opt out of any of the provisions of the default contract by making an explicit contract to the contrary.
How would children and women fare, under a libertarian model of family? To this question I now turn.
The Rights of Children
The libertarian ideal is one of independence. Yet we all come into the world as dependent beings, beings who must obey people who in turn must provide us with care. Such a situation seems contrary to libertarian values, yet it is one of the basic facts of our existence; how can libertarianism accommodate the fact of childhood? The parental right to make decisions for one’s child is an exception to the libertarian principle that no one should make decisions for another; the parental duty to provide care for one’s child is an exception to the libertarian principle that no one should be required to provide assistance to another. What justifies these exceptions?
One possible reply is that these exceptions are beneficial. Consider the toddler who starts to wander into traffic, until the parent swoops down and pulls the child back to safety. Hasn’t the parent coerced the child, preventing it from doing what it wanted to do? It seems so. But if the parent hadn’t intervened, the child might have been injured or killed; so it is in the interest of the toddler to be coerced.
No doubt it is; but can this be what justifies parental authority? After all, libertarians generally reject the paternalistic notion of coercing people in order to benefit them, and argue instead that people have the right to make their own mistakes. Why doesn’t this apply to children? If we allow adults to engage in risky behavior like bungee jumping or mountain climbing or engaging in unprotected sex, why not allow toddlers to engage in risky behavior like walking into traffic or drinking Clorox?
Some libertarians have concluded that the anti-paternalist argument does indeed apply to children, and maintain that it is wrong to restrain children in any way as long as the children aren’t hurting anybody else; such libertarians maintain that children should have full rights to sign contracts or have sex with adults. Reacting against this, other libertarians have gone to the opposite extreme, holding that children are their parents’ property and that parents may do with them as they please. Most libertarians take an intermediate position, regarding parents neither as the equals nor the owners of their children, but rather as their guardians, entitled to make decisions for them and obligated to provide for their welfare. This is surely the commonsensical position; but does it constitute a departure from strict libertarianism?
I don’t think so. In my view, what justifies paternalistic treatment of children is not simply that such treatment benefits children (it might benefit foolish adults as well), but rather that children lack the capacity to make rational decisions about their lives (whereas foolish adults may have that capacity even if they don’t use it much). Consider the analogy of a person in a coma; we make medical decisions for such persons without their consent, because we assume they would consent if they were able to do so. If a person in a coma has left instructions not to use certain kinds of treatment, then most libertarians will agree that we should refrain from using them. So this is not a case of paternalistically overriding someone’s will, but rather of acting as an agent for someone currently unable to exercise his will. We can also extend the analysis to cases where the capacity for rational decision-making is not completely blocked (as in the case of an unconscious person), but simply diminished, as with persons who are drugged or delirious or mentally impaired. I suggest that children may be considered as instances of diminished capacity; guardians act as agents for children, treating the children as they judge the children would consent to be treated if their faculties were fully developed. The standard that justifies paternalism is not benefit but counterfactual consent; the two are different because a person with fully developed faculties can still fail to use them and so make dumb decisions.
This helps to explain why the rights and responsibilities of guardianship go together in the way that they do. Specifically, guardianship is a bundle of one right (the right to make decisions about what happens to the child) and one responsibility (the duty to care for the child’s welfare). These come as a unit because it is only when the decisions we make are those that the impaired person would consent to if unimpaired (as far as we can determine) that we are justified in acting as an agent and substituting our judgment for his.
The fact that the guardian-ward relationship depends on diminished capacity has an important implication for children’s rights. Diminished capacity is a matter of degree; a 13-year-old’s capacity for rational decision-making is not as impaired as a 4-year-old’s, which in turn is not as impaired as a newborn’s. So it is unrealistic to have an absolute cut-off age, below which a child is completely under his guardian’s authority (and unable to engage in any binding financial transactions, from buying a house to buying a pack of gum) and above which he is suddenly a fully responsible agent. The older a child is, the stronger the presumption becomes that a child’s expressed will is an accurate reflection of the will he would have if unimpaired. So, for example, a teenager’s desire to have an ear pierced has to be given more weight than a toddler’s desire to have an ear pierced; and a rational capacity that is not up to giving informed consent in the case of purchasing a house may be quite up to the task of purchasing gum. These sorts of grey areas could probably be handled better by evolving court precedents than by statutory fiat.
I have said that the standard for how a child should be treated is not the child’s benefit, but rather that to which the child would consent if its rational faculties were not impaired — a standard that will presumably track fairly closely with the child’s welfare, but will not match it entirely, especially as the child grows older. (For example, we may think little Nemo would be better off as a stockbroker than as a sidewalk artist, but if all the evidence suggests that Nemo is overwhelmingly likely to choose sidewalk art as his career when he is an adult, then we are not justified in forcing him to go to stockbrokers’ camp, if there is such a thing.) But of course, what the child is likely to consent to retroactively, as an adult, is to a large extent (though not completely) determined by decisions made by the parents in early childhood. In other words, if you were raised a Muslim you will probably look back later and say, “I’m glad I was raised a Muslim”; but if you were not raised a Muslim, then you’ll probably be glad you weren’t. In cases where the child’s likely future preferences are being shaped by present treatment, how do we then turn around and use those likely future preferences as a standard by which to evaluate that present treatment?
This is a difficult case. On the one hand, libertarians generally want to say that the parent is in a better position than anyone else to decide, e.g., which religion a child should be raised in, and this is a matter in which outsiders representing the child’s interests should not interfere, even if we think being raised in one religion is objectively better for the child than being raised in another. On the other hand, when it comes to abusive procedures like female genital mutilation (popularly known by the euphemism “female circumcision,” falsely conveying the impression of being comparable in seriousness to male circumcision), we generally think parents do not have the right to do this, even though women who have had this procedure done when young will usually endorse it in retrospect when they are grown, because they have been inculcated with the relevant cultural attitudes and values. (Cases like Christian Scientists and Jehovah’s Witnesses denying their children medical care seem to come somewhere in between.) Neither the benefit standard nor the counterfactual-consent standard gives precisely the answers we want in such cases, which suggests that I may need to do more tinkering with my theory and somehow incorporate aspects of the benefit standard into the counterfactual-consent standard, without doing so in such a way as to justify a like paternalism toward adults. I haven’t fully figured out how to do this, but perhaps something along the following lines would work: when we consider the child’s likely future preferences, those preferences include both a generic preference for being benefited, and a (possibly mistaken) specific preference for particular treatment regarded as beneficial. Since these preferences are non-actual, we cannot treat one as more expressive of the child’s will than another (whereas once the child is grown and acts on the latter preference, that does give it priority over the former one). So the guardian is obligated to balance the generic desire to be benefited (which requires the guardian to provide what is actually beneficial) with the specific desire for whatever the child is likely to regard, in the future, as having been beneficial. So the more harmful a particular treatment actually is, the more weight the case for abstaining from that treatment has against the contrary weight that the child will end up endorsing it when grown.
How are guardianships acquired? Presumably in the same way as other property rights: by homesteading or transfer. The simplest way to homestead a guardianship would be finding an abandoned infant and undertaking to provide care for it. Another way to homestead guardianship of a child is to give birth to the child; the mother starts out as the child’s guardian, a position to which no one else (not even the father) can have a claim unless the mother grants it. (I do not think an expectant mother could grant guardianship rights in advance, by contract, for the same reason one cannot sell one’s blood before it has been removed from one’s body; one cannot alienate a possession that is still incorporated into oneself.)  One can also obtain a guardianship by gift or sale from someone else who relinquishes it (i.e., adoption).
The fact that what is owned is guardianship over a child, rather than simply the child itself, places restrictions on how one can get rid of a guardianship. As long as one has the guardianship, one is required to use it in ways consistent with the child’s welfare, and so (since renouncing guardianship is itself an exercise of guardianship) one cannot renounce guardianship by throwing the baby in a trash bin or selling it to someone you know plans to cook and eat it. By analogy, if you rescue a comatose patient from a hospital fire, you cannot renounce your guardianship duties by dumping your patient in a river, but must convey the patient to another hospital.
The Status of Women
Libertarians have an uneasy relationship with feminism. Many endorse Christina Sommers’ distinction between “liberal feminism” and “gender feminism.” Liberal feminists, Sommers says, are concerned with legal equality, i.e., with ensuring that men and women have the same rights before the law, while gender feminists go beyond this and assert that sexual inequality pervades every aspect of society, and that a mere equality before the law is insufficient to redress this problem. Sommers’ distinction, and her preference for liberal feminism over gender feminism, is shared by many in the libertarian community.
Libertarian feminist Wendy McElroy offers a more subtle analysis  in the introduction to her book Freedom, Feminism, and the State. There she distinguishes not two but three kinds of feminism. First there is “mainstream feminism,” which simply seeks to include women equally with men in whatever the existing legal status quo is. If there are male Senators, there should also be female Senators; if males can be drafted into the army or compulsory labor camps, so should females; and so on. This position is contrasted with what McElroy calls “radical feminism,” which sees sexual equality as a symptom of a deeper inequality that pervades society as a whole and is inherent in the status quo (so that mere inclusion in the status quo won’t do). There are, says McElroy, two kinds of radical feminism: “socialist feminism,” which sees socioeconomic inequality as the culprit, and individualist (i.e., libertarian) feminism, which regards the problem as stemming from political inequality (where by “political inequality” McElroy means any coercive subordination of one person to another person’s will — statism being the paradigm case of political inequality).
McElroy’s distinction is better than Sommers’, because Sommers would lump mainstream feminists and individualist feminists together into the single camp of liberal feminism, ignoring the important differences between them. But even McElroy’s distinction, it seems to me, does not go far enough. McElroy seems to believe that it is un-libertarian to care about socioeconomic differences between men and women, except to the extent that those differences are the result of coercive state action. Now it is true that libertarian feminists should avoid seeking governmental solutions to such inequalities, but that is not to say they should not regard such inequalities as undesirable, and in need of some sort of (non-governmental) solution. Surely the so-called “gender feminists” are right to point out that undesirable sexual inequalities are extremely pervasive in our society.
As Susan Okin points out in the book I mentioned above, most political theories (and this is certainly true of libertarianism) tend to assume as their subject-matter a mature agent who has been raised by someone else’s labor, usually female labor. The employment conditions in our society (working hours, structure of leaves and benefits, etc.) also seem to be designed with the assumption that the worker has a wife at home, even when the worker is female. Women still do the majority of unpaid household labor, even when they are working, and tend to put their husbands’ careers ahead of their own; as a result, if the marriage breaks up it is the man, not the woman, who is best prepared to prosper on the job market. Okin argues that this fact gives the husband disproportionate power in the relationship, since he has less to lose by exiting. (Okin also points out ways in which existing marriage laws exacerbate this situation; her chapter “Vulnerability by Marriage” is one that libertarian judges and legislators might well read with profit.) In addition, Okin emphasizes that the family is the first school of morality, that is, it is the first context in which people learn about appropriate interpersonal behavior, and if the family is characterized by one-sided exploitative relationships, it will not produce the sort of citizens who can be relied upon to maintain a just society.
I think Okin’s concerns are important ones. Okin’s own solutions, of course, are coercive and statist in nature; but we need not dismiss her account of the problems simply because we doubt both the morality and the utility of her solutions.
A libertarian society would not automatically solve all the problems Okin mentions; cultural biases can survive even without governmental support. However, the absence of such support does weaken the effectiveness of those biases, thus making it easier to combat them through voluntary means, if only we undertake to do so. In particular, the explosion of prosperity that a libertarian society would see, would go a long way toward providing women with an economic safety net more effective than any government welfare program. (One possibility is that women could form mutual-support networks of a kind that today’s governmental regulations would render impossible.) And I have discussed in previous articles why competition would tend to undermine the impact of sexist bias in the marketplace.
I want to close by saying a bit about the issue of spousal abuse, one of the ugliest remnants of patriarchy in the modern family. How should a libertarian legal system handle this problem? Today, our predominantly male (and often macho-oriented) police force is well-known for not being particularly helpful at addressing this question. Tracy Chapman’s song “Behind the Wall” (from the album Tracy Chapman) expresses a familiar complaint:
Last night I heard the screaming
loud voices behind the wall
another sleepless night for me
it won’t do no good to call
the police always come late
if they come at all
and when they arrive
they say they can’t interfere
with domestic affairs
between a man and his wife
and as they walk out the door
the tears well up in her eyes
last night I heard the screaming
then a silence that chilled my soul
I prayed that I was dreaming
when I saw the ambulance in the road
and the policeman said
I’m here to keep the peace
will the crowd disperse
I think we all could use some sleep
Could the fact that current police forces enjoy a coercive monopoly on the provision of security within their respective territories have anything to do with this situation? Imagine a scenario in which different kinds of police agencies, specializing in different kinds of problem, could compete on the open market. A feminist police agency (perhaps a mutual-support network, perhaps a fee-for-service business, perhaps a nonprofit organization depending on charitable contributions, perhaps some combination of the above) would most likely be far more sensitive and responsive to issues of spousal abuse than are present-day police agencies. A wife batterer might have to contend with three feminists armed with Uzis showing up on his doorstep to investigate. (In this connection, remember that gun control (which would not exist in a free nation) is one of the most effective tools of patriarchy, since it favors those with greater physical strength; widespread gun ownership and training undermine female vulnerability to male violence by compensating for average strength differences between men and women.)
A related issue is that of self-defense against spousal abuse. In a number of recent cases, a woman has killed or maimed her abusive husband because she feared a continuation of abuse, even though he was not abusing her at the precise moment she attacked him. Our legal system tends to treat these women as criminals, on the grounds that violent self-defense is justified only when the threat is immediate (except when it’s government that is doing the defending, at which point the criteria for justifiable pre-emptive violence seem to become extremely lax). The argument is that an abused woman should flee the home rather than staying and assaulting her abuser. But why should she have to leave her own home, simply because it is also the abuser’s home? Even our degraded legal system generally recognizes that one has no duty to retreat from an attacker when one is in one’s own home. If you are the victim of a persistent pattern of severe rights-violations, a pattern you have every reason to expect will continue, and if external authorities offer no reliable protection, it seems to me that you are justified in undertaking your own defense, and that a libertarian court should recognize this. A competitive legal system would allow women’s perspectives a greater voice in deciding the treatment of such cases than is possible under our monopolistic system.
Conservatives are right: the family is an institution of paramount value and importance, both in its own right and as a bulwark against the encroachments of the state. Liberals are also right: the family has often served as a sphere of oppression and exploitation, thanks to the tradition of patriarchy, in which women are unjustly subordinated to men, and children are unjustly subordinated to parents. The proper libertarian response to both concerns is to see how, consistent with our anti-interventionist principles, we can foster a family structure free of patriarchal influence.
In the case of parents and children, this means recognizing that in deciding how to treat their children, parents must attempt to track not just the child’s welfare but also what the child is likely (once mature) to prefer; since a child’s expressed preferences become a more and more accurate guide to its mature preferences as time passes, this means that parents have less and less justification, as their child grows older, for imposing on it their own conceptions of benefit when these clash with the child’s. This model of the parent-child relationship is thus anti-patriarchal, in that it gives children a greater right to a say in their own treatment than the benefit standard does, while at the same time recognizing enough distance between expressed and mature preferences to avoid the extreme consequences of “kid lib.”
In the case of husbands and wives, going beyond patriarchy means seeking to foster both a work environment and a home environment that do not systematically disadvantage women in relation to men. In the economic sphere, this involves removing regulatory barriers to competition, thus giving employees generally, including women, more clout in the job market, thus putting them in a better position to negotiate for higher pay, parental leave, and the like (which the employers, also benefiting from the economic boom that freedom would bring, would be in a better position to provide). In the legal sphere, it involves abolishing laws that discriminate against women, and more importantly, opening up the services of adjudication and enforcement to competition so that the concerns of women could be more adequately represented. And in the cultural sphere, it involves inculcating an attitude of reciprocity and mutual respect.
Some libertarians may say that we don’t need this last aspect: if there is any serious problem, the market will take care of it, so we don’t need to do any cultivating. I think this attitude is a mistake, and tends to encourage discriminatory attitudes (if the market hasn’t taken care of it, then it must not be a serious problem; e.g., if women aren’t making as much money as men on the market, it must be their own fault). Libertarians are often reluctant to recognize entrenched power structures when they don’t come attached to governmental offices; but we should always remember that power and tyranny are older than the state. Indeed, Herbert Spencer intriguingly suggests (in his Principles of Sociology) that the subordination of women by men is the initial form of oppression from which all later ones grew, including the state. We should also remember, when we say “the market will take care of it,” that we are the market, that its successful operation depends on the alertness of Kirznerian entrepreneurs, and that we who have noticed a problem are in the best position to fill that entrepreneurial role. Stressing the Hayekian strand within Austrian socioeconomic thought at the expense of the Kirznerian strand can lead to excessive passivity in the face of the omniscient, omnipotent forces of history.
Translations for this article:
 Susan Moller Okin, Justice, Gender, and the Family (New York: Basic Books, 1989). This book has gotten something of a bad press among libertarians, first because of its bizarre attack on libertarianism, and second because of Okin’s own rather socialistic policy proposals. It is true that Okin tends to misunderstand and misrepresent the positions of her opponents, and her chapter on libertarianism is particularly egregious in this regard; it is also true that her policy proposals would be a statist nightmare if enacted. Nevertheless, I think there is a great deal of value in her book that libertarians need to consider carefully.
 The argument is sometimes made that even if homosexual relationships should be permitted, they should not be called “marriage,” because marriage has historically been a relationship between men and women. But by that logic, contemporary heterosexual relationships should not count as marriages either. After all, marriage has historically involved the wife’s legal absorption into and subordination to her husband, so one could argue that no relationship between equals should be considered a marriage. (In fact, this is exactly what many 19th-century “free love” advocates did; the free-love movement’s antagonism toward marriage was not (in most cases) an endorsement of promiscuity, but rather a hostility to what they saw as an inherently one-sided and exploitative relationship.) But I think this would be a mistake; the nature of marriage is not inherently determined by the particular form it takes in a given society. Marriage and the family are historical phenomena, and cannot be defined in separation from the way they develop over time.
 Single-parent families are currently under attack from conservatives, who cite statistics showing that children from two-parent homes tend to do better than those from one-parent homes. One question that is seldom asked is how much of this difference derives from an inherent advantage of two parents over one, and how much instead from the economic hardship and reduced parent-child time that a (politically manufactured) low-wage economy imposes on single-parent families?
 Actually, the two cases are somewhat different. As I understand it, Jehovah’s Witnesses simply refuse certain kinds of medical treatments on religious grounds, without offering alternative treatment, arguing that the child is better off dead than alive but damned. Christian Scientists, by contrast, treat their children by means of spiritual healing, a method that has an impressive success rate but many unexplained failures, just as mainstream medicine has an impressive success rate but many unexplained failures; so disputes over Christian Science treatment for children have more to do with the medical profession’s claiming a government-sanctioned monopoly in the field of health care than with issues of child neglect and so forth.
 This raises the complicated issue of surrogacy contracts. One side wants to enforce them, the other side to forbid them. As I see it, the correct position is that specific performance should not be enforceable (because an expectant mother cannot alienate guardianship rights while the child is still in her body), but money damages should be enforceable.
 At least, she once did. In more recent writings, however, she unfortunately seems to have adopted Sommers’ terminology.
 Okin cites statistics showing that after divorce, the average man’s economic position improves while the average woman’s declines. Since she wrote her book, the particular study on which she relied has been discredited; but this shows only that the post-divorce difference is less extreme than Okin supposed, not that it is insignificant.