Custody law reform: letting the child decide

Twilight Zone - "It's a Good Life"

Twilight Zone – “It’s a Good Life” episode (1961)

The suggestion is sometimes made that custody decisions should not be based on either parental rights or a judge’s opinions about what is best for a child. Instead, children should be empowered to make their own choices as to custody. Since the child is the one who will be most impacted by the decision, this approach does have a certain amount of appeal.

The problem with it is that children are children, not adults. Their capacities for judgment and reason are not yet fully developed. This is why, as a society, we do not permit children to execute wills, enter into marriages, or be irrevocably bound by the contracts they sign. It is why we delegate to a parent or guardian the authority to make decisions on a child’s behalf. Until children attain a sufficient level of maturity and judgment, parents choose their school, religion, dietary regimen, and so on.

Just as a child generally is not mature enough to use good judgment when making those kinds of decisions, so a child is not likely to be mature enough to use good judgment when deciding on a custodian. For example, a child may be likely to select the one who is constantly giving her candy and keeping her entertained, and shun the one who makes her study and eat healthy food, even though placement in the latter parent’s custody is more likely to be in her best interests.

Even if a child has reached a sufficient level of maturity to make a wise decision, there would still be a big problem with requiring a child to select one or the other parent in the event of a divorce or breakup of her parents’ relationship. Requiring a child to take sides in a custody battle is “empowering” to a child only in the way that giving a prisoner a choice between a firing squad or a gas chamber is “empowering” to the prisoner.

Putting a child in the position of having to hurt the feelings of one or the other of his parents is very likely to impair the child’s relationship with that parent. Worse, it will almost certainly damage a child psychologically. Children of divorce are already at risk of internalizing responsibility for their parents’ divorce. They should not be subjected to the additional risk of internalizing responsibility for a parent’s feelings of rejection. A person’s internal attribution of control over a negative event leads to depression and other mental health problems.1

Allowing children to make their own custody decisions also creates an incentive for adults seeking custody to compete for the child’s affections. Competition is sometimes a good thing, but it is not necessarily a healthy thing in this context. A parent competing with the other parent for the child’s affection will have an incentive to prevent or discourage the child from developing a healthy relationship with the other parent. This might be accomplished by limiting the amount of time the child is permitted to spend with the other parent, and/or by disparaging or criticizing the other parent in the child’s presence. A competing parent will also have an incentive to try to influence the child by doting on him, being overly generous, and withholding discipline. None of these things is healthy for children.

Courts already encounter these kinds of problems because the child’s preference is a factor that they consider when deciding custody. The problems can be expected to intensify considerably if the child’s preference were made the determinative or presumptive basis for every custody decision.

When formulating general policy concerning the custody of children, it might make sense to consider children’s preferences concerning the kind of custodial arrangement to which they are subjected. For example, it might be reasonable to consider statistical evidence concerning whether children prefer sole vs. joint custody arrangements. Making a particular custody decision turn entirely, or even primarily, on an individually identifiable child’s expression of preference, however, does not.

The complete History of Custody Law is now available in paperback and as a Kindle e-book:

Purchase at

  1. See Christopher Peterson & Martin E.P. Seligman, Causal Explanations as a Risk Factor for Depression: Theory and Evidence, 91 Psychol. Rev. 347 (1984); see also Gall v. Gall, 336 So. 2d 10 (Fla. Ct. App. 1976) (holding that allowing a child to choose where he wants to live at a particular time is not in children’s best interests because it invites disciplinary problems); Ralston v. Ralston, 396 S.W.2d 775 (Ky. 1965) (same.)

Primary Attachment Figures

Harlow's surrogate mother experiment

Harlow’s surrogate mother experiment

Because children may suffer emotional or psychological damage if they either are not allowed to form an attachment to a caregiver, or are separated from a person to whom they have formed an attachment, it could make sense to apply a custodial preference for the person to whom a child has formed an attachment.

Attachment theory

Attachment, in the psychological context, simply means a strong and lasting emotional bond between two people.1 It may be unilateral or bilateral. A person may form an attachment to a child who has not formed an attachment to that person; and a child may form an attachment to a person who has not formed an attachment to the child.

Attachment theory was first formally propounded by John Bowlby.2 He theorized that separating infants from their mothers impaired children’s social, emotional and cognitive development. Observing that children separated from their mothers experience anxiety even when they are provided adequate food from another source, he concluded that a special bond between a child and a mother exists, and that it cannot be explained in terms of behavioral conditioning linked to the mother’s nursing of the baby.3 He postulated that children are born with an innate need to be close to a primary caregiver, a need that evolved in humans and other animal species because its existence tended to improve a child’s chances for survival. According to Bowlby, people are born with an innate desire stay close to a caregiver, at least when under stress or threatened.4 He and other proponents of the evolutionary theory of attachment have suggested, further, that people are innately pre-programmed to engage in certain kinds of behaviors (crying and smiling, for example) that will stimulate pre-programmed caregiving responses in human adults. A child’s attachment, therefore, is not necessarily to the person who provides her food, but to the person who responds to the child’s needs – which can include, in addition to food, such things as protection and comfort.

To determine whether the source of an infant’s attachments is nature or nature, twentieth century researcher Harry Harlow conducted a series of experiments on rhesus monkeys. He found that monkeys who were raised in isolation from birth tended to die, behaved abnormally, and had difficulty interacting with other monkeys. They also tended to lack caregiving skills upon becoming parents themselves, many of them even abusing and neglecting their own babies. This demonstrated how attachment could be an evolutionary adaptation that would enhance a species’ chances of survival.5

Harlow’s most famous experiments involved separating rhesus monkeys from their mothers at birth, and raising them with different kinds of surrogate mothers. Some of these were made of cloth; others were made of wire. Some gave milk; others did not. He found that the monkeys would go to the wire surrogates only for milk, but would go to the cloth ones the rest of the time. When frightened, the monkeys would go to the cloth surrogate, not the wire one, even if the wire one gave milk and the cloth one did not. Neither the cloth nor the wire surrogates provided any actual protection for the infant monkeys. And monkeys raised with surrogates developed more normally as compared to those raised in isolation.

Harlow’s findings tend to support Bowlby’s evolutionary theory of attachment. Since the baby monkeys seemed to have the closest attachment to cloth surrogate mothers even when it was only the wire surrogate mother that provided them with food, this seems to suggest that sensitive response and security have more to do with attachment than conditioning does.

Konrad Lorenz’s experiments with goslings provides further support for the evolutionary theory of pre-programmed attachments. Lorenz placed half a batch of goose eggs under a goose mother, and kept the other half beside himself as they hatched. Goslings hatched by the mother followed her and looked to her for sustenance and protection. The others followed Lorenz and looked to him for sustenance and protection. In other words, geese follow the first moving object they see during the first few hours after hatching. When the goslings were mixed together in a box and then released, the ones that had been hatched near the goose mother went to her; the ones that had hatched near Lorenz went to him. This behavior occurred prior to any feeding. This strongly suggests that attachment is a pre-programmed response, not something that is learned through conditioning.6

A Freudian overlay

In 1973, Joseph Goldstein and Albert Solnit teamed up with Sigmund Freud’s daughter, Anna Freud, to publish a book called Beyond the Best Interests of the Child.7 It espoused a psychoanalytic approach to determining what is in a child’s best interests, one that essentially applied Sigmund Freud’s psychoanalytic theories to John Bowlby’s attachment theory. This book has had a tremendous impact on custody evaluations, and therefore on the practical development of custody law.

Freud and her colleagues observed that an infant develops a psychological attachment to an adult who is personally and emotionally involved with him. They theorized that the attachment consists of the baby’s superimposition of its libidinal interests on the events of bodily care (feeding and diaper-changing, for example.) They postulated that because of this early attachment, it is vitally important to a child’s psychological well-being for this first bond to remain uninterrupted.

Such primitive and tenuous first attachments form the base from which any further relationships develop. What the child brings to them next are no longer only his needs for body comfort and gratification but his emotional demands for affection, companionship, and stimulating intimacy. Where these are answered reliably and regularly, the child-parent relationship becomes firm, with immensely productive effects on the child’s intellectual and social development.8

“Unlike adults, who are generally capable of maintaining positive emotional ties with a number of different individuals,” they asserted, “Children lack the capacity to do so.”9 Since mothers are the ones who are biologically equipped to nurse their babies, it is easy to see how Freud’s theoretical framework could be used to support a preference for making awards of sole custody to mothers.

One difficulty with using Freud’s psychoanalytical libido-superimposition theory as a justification for a judicial preference for mothers is that Harlow’s experiments with surrogates demonstrated that primate attachment does not correlate directly with the meeting of a child’s libidinal need for food. The monkeys in his experiments demonstrated a greater attachment to cloth surrogates that provided no milk than to wire surrogates that did. Moreover, it is not necessarily true that the mother is the one upon whom the child superimposes its libidinal needs in every case. If a particular father does a greater share of the feeding, diapering and comforting of the baby than the mother, then the theory would support a preference for making an award of sole custody to the father rather than the mother in that case. The theory is not sufficient, by itself, to support a preference or presumption specifically in favor of mothers.

Significantly, researchers H. Rudolph Schaffer and Peggy Emerson found that attachment correlates closely with the person who responds most accurately to the baby’s needs and desires, not the person who spends the most time with the baby.10 This suggests that applying a presumptive preference for a child’s primary caregiver may actually be detrimental to a child’s health and development, in those cases where the child’s primary caregiver and primary attachment figure are not the same person. Such a presumption or preference may be in the best interests of a real or assumed primary caregiver (i.e., mothers), but custody law is supposed to be focused on what is in a child’s best interests, not what is in one or the other parent’s best interests. For this reason, a presumption or preference for a child’s primary attachment figure would seem to be an improvement over the presumptive preference for a child’s primary caregiver.

Unlike the primary caretaker presumption, a primary attachment preference associates directly with a child’s interests, not the interests of the parents. Because they take a quantitative approach to the evaluation of parental care, the primary caretaker and approximation principles reflect a greater concern with ensuring that a harder working parent’s efforts are rewarded than with the child’s psychological and emotional needs. A primary attachment presumption, by contrast, would focus directly on the child’s psychological and emotional needs rather than simply attempting to “reward” parents in proportion to the extent of their efforts to perform specified caregiving tasks.

Multiple attachments

A major difficulty with using attachment theory as a basis for establishing a preference for the person to whom the child has formed an attachment is that there appears to be no real basis for the assumption that children are incapable of forming psychological attachments to more than one person. Nor does there appear to be any evidence for the assumption that multiple attachments are detrimental to a developing child’s emotional psychological health and well-being. To the contrary, available empirical evidence contradicts both assumptions. Schaffer and Emerson, for example, found that from birth to three months of age, babies respond and attach equally to any caregiver; between four and nine months, a baby tends to have a primary attachment to one person, but also maintains secondary attachments to others; and by nine months babies form multiple attachments. Further, they reported that at eighteen months, a child’s main attachment figure was the mother for only about half of the children studied, and the father was the main attachment figure for nearly all of the rest. Also – and contrary to Freud’s libidinal-superimposition theory — they found that the most important factor in forming attachments is not who feeds and diapers a child, but who plays and communicates with her.11

Parental rights

Because a child may have a strong attachment to someone who is not the child’s parent, a presumptive custodial preference for a child’s primary attachment figure conflicts with the doctrine of parental rights. Under a primary-attachment rule, if a live-in nanny can prove that children have established a stronger attachment to him or her than to their mother or father, then the nanny would have a presumptive right to custody of their children notwithstanding the parents are both fit and capable, and have not abandoned their children. While some states may be prepared for a brave new world in which natural biological relationships have no special legal significance, it is doubtful that very many legislators and judges are prepared to abolish the notion of parental rights completely, particularly since the United States Supreme Court has granted them recognition as constitutionally protected fundamental rights. Accordingly, while the principle that children normally should be placed in the custody of the person(s) to whom they have formed the closest attachment(s) might be considered as a factor in custody cases, it may be questioned whether it should be elevated to the level of a presumption.

Attachment theory as an argument for presumptive joint custody

Since children are capable of forming multiple attachments, and since most children form attachments to both of their parents, attachment theory would seem to support a presumptive award of shared custody at least as much as it would support a preference for sole custody.

Moreover, the United States Supreme Court has declared that parents must be presumed to act in the best interest of their children. Most child development experts agree that strong, healthy bonds between a child and each of his parents are in a child’s best interests. Therefore, attachment theory arguably may support a legal presumption that joint parental custody is in the best interests of children.

The complete History of Custody Law is now available in paperback and as a Kindle e-book:

Purchase at

  1. JOHN BOWLBY, 1 ATTACHMENT (1969) (“lasting psychological connectedness between human beings”); M.D.S. Ainsworth, The development of infant-mother attachment, in REVIEW OF CHILD DEVELOPMENT RESEARCH (B. Cardwell & H. Ricciuti eds., 1973.)
  2. John Bowlby, The Nature of the Child’s Tie to His Mother, 39 INT’L. J. PSYCHOANALYSIS 350-71 (1958)
  3. The prevailing belief at the time had been that a mother’s act of nursing children conditions children to form attachments to their mothers. Children, it was thought, could expect to be rewarded with milk for seeking physical closeness with their mothers; and could expect to be punished with hunger if they did not. See J. DOLLARD AND N.E. MILLER, PERSONALITY AND PSYCHOTHERAPY (1950).
  5. H.F. Harlow and R.R. Zimmermann, The development of affective responsiveness in infant monkeys, 102 PROC. AMER. PHIL. SOC’Y 501-09 (1958)
  6. Konrad Lorenz, Der Kumpan in der Umwelt des Vogels. Der Artgenosse als auslösendes Moment sozialer Verhaltensweisen, 83 J. ORNITHOLOGIE 137–215, 289–413 (1935).
  8. Id. at 18.
  9. Id. at 13.
  10. H. Rudolph Schaffer and Peggy E. Emerson, The Development of Social Attachments in Infancy, 29 MONOGRAPHS OF THE SOC’Y FOR RES. IN CHILD DEV. no. 94 (1964)
  11. Id.

The approximation rule, revisited


Because the primary caretaker presumption granted sole custody to a parent on the basis of which parent spent the majority of time caring for the child, it has been criticized for discounting the importance of the 49% caregiver, and giving the 51% caregiver a windfall. Some have suggested that a fairer approach would be for a court to  apportion time between the parties on the basis of how they apportioned time between themselves before the court became involved. Of course, a court cannot determine with exact certainty how much time each parent spent caring for the child in the past. Proponents argue, though, that a court should strive for a result that comes as close to the pre-divorce allocation as possible.1

This approach, known as the “approximation rule,” was first suggested by Elizabeth Scott.2  It has been adopted as the rule of decision in West Virginia.3 The American Law Institute endorses it.4

The approximation rule is an improvement over the primary caretaker presumption in the sense that it better allocates rewards to the parents in proportion to the child-raising services they have provided. Instead of giving 100% of the custody “reward” to a parent for doing 51% of the caregiving, while giving 0% of the custody “reward” to the parent who did 49% of the work, the approximation rule attempts to assign parental responsibility to the parties in approximately the same proportion they exercised prior to the breakup of their relationship. This seems to be fairer to each parent than the primary caretaker and presumptive maternal sole custody rules are. Instead of applying a winner-takes-all approach, each party “wins” only that amount of time which he or she has earned through the performance of child-raising services.

The rule also seems — at first blush, anyway — to be likely to promote children’s best interests by giving parents an added incentive to spend as much time with their children as possible. Knowing the amount of time they will get to spend with their children after a divorce or relationship breakup is directly tied to the amount of time they spend caring for their children gives each parent an incentive to spend as much time with the children as possible.


The approximation rule is vulnerable to the same kinds of criticisms as the primary caretaker presumption is.

Definitions of caregiving, for the purpose of determining how much time each parent has spent “caring” for a child typically are sex-biased. Judges and legislators tend to include tasks that traditionally are performed by women in the definition, while omitting tasks traditionally performed by men from the definition.

Even if a purely time-based quantitative measure were applied, it may be questioned whether it makes sense to lock a couple in to caregiving patterns that were established during a child’s infancy. Children’s needs, and parents’ capacities to meet them, vary over the course of a child’s life. It may not make a whole lot of sense, from a child development point of view, to decide how much time an adolescent or teen boy spends with his father on the basis of how much time the mother, as compared to the father, spent nursing him when he was a baby.

Moreover, a couple’s circumstances rarely are the same after a divorce as they were before the divorce. The fact of separation, in itself, necessarily alters the allocation of responsibilities. Economic realities are such that it often is necessary for both parents to work outside the home following a divorce, as there will be two households to maintain following a divorce, not just one. Accordingly, it may not be feasible – let alone desirable, or fair – to try to “lock” a couple into an arrangement that existed at a time when their circumstances were radically different from what they are after the divorce.

Like the primary caretaker presumption, the approximation rule is subject to the criticism that it ascribes greater importance to quantity than to the quality of caregiving, and it overlooks the importance of the strength of the attachment between a parent and a child.5

Finally, because the rule gives parents an incentive to spend as much time as possible with their children relative to the time spent by the other parent, it also gives them an incentive to prevent, limit or even interfere with the other parent’s ability to spend time with the child. This is a critical concern in cases involving unmarried parents, where state laws automatically give the mother an exclusive right to sole custody at the time of birth, and the father has no legally enforceable right to spend time with his child until he obtains a court order addressing custody and visitation. In a state that has adopted an unqualified version of the approximation rule, an unmarried mother who wished to completely excise the father from the child’s life would have the absolute power to do so simply by exercising her exclusive custodial rights to prevent the father from spending any time at all with his child before he is able to get into court. If she succeeds in preventing the father from having any contact at all with his child, then a consistent application of the approximation rule would require the court to reward her with 100% of the custody rights and parenting time, and to allow the father to have no rights of access to his child at all.


The complete History of Custody Law is now available in paperback and as a Kindle e-book:

Purchase at


  1. American Law Institute, Principles of the Law of Family Dissolution: Analysis and Recommendations (2003); Elizabeth S. Scott, Pluralism, Parental Preference and Child Custody, 80 Cal. L. Rev. 615, 619 (1992)
  2. Scott, supra note 1 at 617
  3. W. Va. Code § 48-11-106 (2000).
  4. American Law Institute, supra note 1.
  5. See, e.g., Shelley A. Riggs, Is the Approximation Rule in the Child’s Best Interests? A Critique from the Perspective of Attachment Theory, 43 Fam. Ct. Rev. 481 (2005) (critiquing the approximation rule from the point of view of attachment theory); see also Joan B. Kelly, Developing Beneficial Parenting Plan Models for Children Following Separation and Divorce, 19 J. Am. Acad. Matrim. Law. 237, 241 (2005) (arguing that the approximation rule is not an improvement over the “best interests of the child” standard.)

“This State Treats Women Badly”



“Welcome to Ohio, one of 31 states that allow rapists to sue women for child custody rights. This state treats women badly.” This is what Nita Chaudhary says is the message her organization, Ultraviolet, intends to convey in a sign that greets visitors as they arrive in Ohio.

Signs like this are showing up at or near airports in several states. The list of states, so far, includes Ohio, Louisiana, Texas, North Carolina, Nevada and Kentucky. Each sign is tailored to a specific issue affecting women in the particular state in which the sign appears. The Louisiana sign, for example, asserts that women in that state earn less than 70% of what men earn.

Turning to the Ohio sign, I was ready to shout, “Right on, sister!” until I thought about the second part of the message. I still believe the law needs to be changed, but my interest in joining forces with Ultraviolet or other feminist organizations to bring about that change has gone out the window. Here’s why.

Misrepresentation to promote sexism does not excite interest in fighting sexism

The statement that Ohio is one of 31 states that allow rapists to sue women for child custody rights is true, but it is so incomplete as to be misleading. The complete statement would be: Ohio is one of 31 states that allow rapists and others who force people to submit to sexual acts to sue women and men for child custody rights. These may seem like unnecessary clarifications to make, but not if you stop and think about it for a moment.

At common law, rape was defined in such a way that it was legally impossible for anyone other than a man to be guilty of it. A woman who forced a man or a boy to engage in sex with her was excluded from the definition of a rapist. Most states today, though, have criminalized the commission of sex acts by force, or upon minors, even when the act is perpetrated by a female and the victim is male. States that allow male sex offenders to sue women for child custody rights also allow female sex offenders to sue men for child custody rights. There is no rational, non-discriminatory reason that the thought of a male sex offender seeking custody of a child should be more abhorrent than the thought of a female sex offender seeking custody of a child.

Moreover, there are two kinds of rape: forcible and statutory. If an underage male has sexual intercourse with an underage female, he may be guilty of statutory rape even if the entire encounter appeared to be consensual. It doesn’t matter that no force was used. The law says people under a specified age are simply incapable of consent. If two seventeen-year-olds have what would otherwise be consensual sex if not for their ages, and a baby results, why should the female have an exclusive right to custody of the child?

It is only by concealing this information that it is possible to jump to the conclusion that “this state treats women badly.” Once the facts are correctly stated, the only inference that can logically be drawn from the state of the law is that these states  treat male and female victims of sexual coercion badly, not that they specifically target women for maltreatment.

Knowing this dampens my enthusiasm for the message because it makes it seem as if the message is more about promoting pro-female and anti-male discrimination than it is about protecting children from sexual deviants.

 False attribution of discriminatory intent

“Welcome to Ohio, one of 31 states that allow rapists to sue women for child custody rights. This state treats women badly.” The juxtaposition of these two statements suggests that sex offenders are allowed to take women’s children away from them because the state of Ohio exalts men and hates women. Of course, this is not true. To see that it is not, consider whether the following syllogism would also be logical:

“Welcome to Ohio, one of 50 states (or whatever the current number is) that allow women who molest boys to sue the boys they molest for child support rights. This state treats boys badly.”

The same logic is being applied in each proposition. Yet, does it really make sense to say that states are engaged in wars of discrimination against both men and women? Maybe. But if so, then the correct statement would be, “This state treats women and men badly,” not “This state treats women badly.”

And this is the fundamental problem with feminism as it has been practiced for the past thirty or more years: It seeks solutions only for those aspects of unfair laws that are unfair to women, while insisting that unfairness to men and boys must be ignored. This is why we have laws against female genital mutilation instead of laws against genital mutilation; challenges to the exclusion of women from combat positions but no challenges to the exemption of women from Selective Service registration requirements (at least, not any coming from women); a Violence Against Women Act instead of a Violence Against People Act; and so on and so on.

A growing number of people are beginning to realize that the truth is that some laws have been, and are unfair to women, while other laws have been, and are unfair to men. Unfairness is neither a contest nor a zero-sum game. There is plenty of it to go around. Portraying men as evil-doers and women as innocent victims in order to capitalize on societal expectations that men should serve and protect women, and should never be heard to complain about it, can be an effective strategy for making short-term gains. At the same time, though, it damages the credibility of women who talk about equality, or who say that feminism is “about equality.” It would be prudent to consider whether the short-term advantages that women gain from this approach outweigh what is lost in the long-term.

The primary caretaker presumption: a critical analysis


A merit-based alternative to sex discrimination

Rather than using a sex-based preference to decide custody, it is sometimes suggested that courts should apply a works-based preference; that is to say, that courts should apply a preference for the parent who has done most of the work of caring for the child.

In the 1980’s, Minnesota and West Virginia courts did just that. These states established a legal presumption that children should be placed in the sole custody of the parent who has been their primary caretaker, which they defined as the one who has assumed the most responsibility for child-care tasks like feeding, clothing, bathing, transporting to school and medical appointments, and the like.


The West Virginia court reasoned that implementing a uniform presumption would provide greater predictability of results and therefore would reduce litigation.1 In practice, though, the presumption had the opposite effect, resulting in what one commentator described as an “explosion” in litigation.2 Rather than reducing contested litigation, the rule caused parents to start scrambling to marshal evidence and construct arguments to prove they had done “most of the work,” or that the kind of work they had done was more important than the kind of work the other parent had done.

Another justification that is sometimes offered for the presumption is that it is felt to be fairer to the parent who performed most of the caregiving during a marriage. In Burchard v. Garay,3 for example, the California Supreme Court said that an award of custody to anyone other than the mother is “particularly unfair when, as here, the mother has in fact been the primary caregiver.” Concurring in the result, Chief Justice Rose Bird suggested that not only should there be a presumption in favor of the primary caretaker, but there should also be a presumption that mothers are always the primary caregivers and fathers are not. “Typically, it is the mother who provides most day-to-day care, whether or not she works outside the home. A presumption which ignores this fact is likely to lead to erroneous and unfair decisions,” she wrote.4

It is somewhat disconcerting that even as late as 1986, the top judicial official of the most populous state in the nation was advocating for the perpetuation of sex role stereotypes in the law, at a time when those very stereotypes were handicapping women who were seeking equal opportunities in workplace roles traditionally performed by men. To an individual learned in the law, it should make no sense to say that employers should be entitled to apply a presumption in favor of male applicants when making hiring and promotion decisions with respect to medical, technical, law enforcement, labor-intensive, or other jobs which traditionally have been performed mostly by men rather than women. By the same token, it should make no sense to say that judges should be entitled to apply a presumption in favor of females when making custody decisions simply because custody-related jobs traditionally have been performed mostly by women rather than men – especially when judges themselves, by establishing legal doctrines like the tender years presumption and the maternal preference, were largely responsible for creating and enforcing the practice of rigid sex-role stereotyping in the first place.

Feminist commentators, however, argue that the law should apply sex-based preferences in favor of women. The standard rationalization for allowing “benign” discrimination (sometimes called “affirmative action”) is that doing so is necessary to undo, or balance out, the effects of past discrimination. Since American custody law has always favored women, that argument is of no avail in the family law context. What possible argument could there be for re-establishing a standard that favors women simply because they are women? Well, according to Laura Sack, doing so would “alleviate much of the post-divorce suffering of women.”5 This, she says, will help promote “gender equality.”

The “equality” she is referencing, of course, is not equality of rights, but equality of economic outcome. The idea is that the economic benefit to a parent of sole custody, in terms of higher child support and so on, should be used to offset perceived differences in the relative class-economic-statuses of men and women.6


The principal feminist argument in support of a “primary caretaker” presumption that includes a biased presumption that women are the primary caretakers of children is that women should receive custody because they suffer more financially after a divorce. This argument, though, is a circular one. The principal reason cited for women’s purported greater financial suffering after a divorce is that women more often have custody of the children after a divorce, so their living expenses are greater.7 This would be an argument for making more awards of custody to fathers, unless it is assumed that only women should be awarded custody of children. Hence the circularity.

Moreover, the notion that women suffer more, economically, after a divorce than men do is dubious, at best. The claim that women suffer financially after a divorce, while men gain financially, was propounded by Lenore J. Weitzman in her 1985 book, The Divorce Revolution: The unexpected social and economic consequences for women and children in America. In support of this claim, she set out a series of statistical “findings” demonstrating what appeared to be extraordinarily severe financial losses suffered by women post-divorce, and nothing but financial gains experienced by men. These have been described as “ranking among the most cited demographic statistics of the 1980s.”8 The problem is that her statistics were not valid. They were not even consistent with her own information, something which Weitzman herself eventually acknowledged.9 In reality, it is estimated that men suffer a 10-40% financial loss following a divorce.10 Unless a man contributed more than 80% of a family’s income before the divorce, he will actually suffer more financially from divorce than his wife will.11

Claims that women suffer more financially than men post-divorce even after transfers of income from men to women in the form of child support are accounted for are based on comparisons of gross (pre-tax) income, and do not take other tax consequences into consideration. This is a significant omission because child support is not deductible by the payor, and is received tax-free by the recipient. In addition, custodial parents enjoy several tax-reducing advantages that noncustodial parents do not, i.e., taxation at the lower, head-of-household rate; a presumptive right to claim children as exemptions; child tax credit; child-care expense deduction or credit; and, in some cases, an earned income credit. Moreover, people who make these claims entertain an assumption that a custodial mother pays 100% of the children’s expenses, while the noncustodial parent has absolutely no child-related expenses at all. In other words, they assume either of two things: (1) that noncustodial parents do not need to provide space in their homes for children during visitation; do not purchase food or anything else for the children during visitation; do not share in the cost of health insurance or expenses; and do not have any transportation costs; or (2) that the custodial parent gives money to the noncustodial parent to cover all child-related expenses during visitation. Neither assumption is valid. Most states mandate a contribution by the noncustodial parent to the child’s medical insurance and expenses; and also require the noncustodial parent to contribute to the custodial parent’s child-care expenses. No state requires the custodial parent to provide money to the noncustodial parent to support children during visitation periods. While visitation percentages vary, there is some evidence that most noncustodial parents today typically have children in their care over 24% of the year.12 The assumption that custodial parents fund the noncustodial parent visitation for 24% of the year has no foundation in reality. The truth is that noncustodial parents incur and pay child-related expenses that are not reimbursed by custodial parents, and these expenditures can be quite substantial.13 When data is corrected for these errors, and when account is taken of the existence of male custodial parents (rare, but not non-existent), mothers’ post-divorce standard of living not only is not inferior to fathers’, but it is actually higher than fathers’.14

Even if it were true that women suffer more economically after a divorce than men do, the claim that giving women the benefit of discriminatory stereotypes “promotes gender equality” requires a definition of equality that is expressed exclusively in terms of equality of economic outcome. It may be questioned why other kinds of equality, i.e., equality of rights and equality under the law, must be sacrificed for the sake of achieving absolute equality of economic outcomes. If, by some happenstance, it became possible to ensure that the average woman had exactly the same level of income as the average man simply by limiting freedom of religion to Protestants and not Catholics, or by prohibiting black people from voting, would we do it? If not, then why should we be willing to sacrifice other fundamental rights, such as the fundamental right of parents to the custody and care of their children, to achieve that outcome?

There are also ethical problems. Should children’s lives and well-being be sacrificed in the interest of achieving gender-class-outcome-equalization goals? Should children be treated as items of property to be distributed among the populace on the basis of adult financial statuses?

Some feminists have argued that courts should always award sole custody to women as “a reward in consideration for the caretaking she has performed.”15 If this is true, then why shouldn’t male caretakers be rewarded for the same reason?

Does it even make sense to view custody law as a sort of Skinnerian system of rewards and punishments, though? That approach certainly conflicts with the view of custody law as an exercise of the state’s parens patriae authority to protect the best interests of children. Applying the “reward for services rendered” approach would mean routinely awarding custody to parents who, although buying all the groceries and making all the meals for their children, also subject them to repeated beatings.

Finally, all of these arguments are based on sexist generalizations about men and women and their appropriate roles in society. Establishing presumptions on this basis entrenches sex-role stereotyping as a social norm. Unless feminists are prepared to concede that employers may favor male job applicants in traditionally male-dominated fields simply because they are male, they should think twice before advocating for any legal standard on the basis that it produces outcomes that are more favorable to one sex than to the other.

Discriminatory definitions of caregiving

The primary caretaker presumption has been criticized because it leaves a judge with the power to make a finding that the father in a particular case is the primary caretaker, or to award custody to a father on the basis of a finding that the mother is unfit.16 Of course, no solution short of a return to an explicit maternal preference can avoid that kind of criticism. It seems doubtful, however, that returning to a system in which courts explicitly rule in favor of mothers solely because they are women could, or should, withstand constitutional scrutiny.

Others argue that the problem with the standard is not that it fails to discriminate enough, but that it discriminates too much, albeit in a less overt way than the tender years and maternal preference doctrines did. The crux of this argument is that legislators and judges often define tasks traditionally performed by mothers as “caregiving,” while omitting from the definition those parenting tasks that traditionally are performed by fathers. For this reason, several commentators have observed that it is really little more than the maternal preference in disguise.17

There is some merit to this criticism. The genesis of the presumption, after all, was the West Virginia Supreme Court’s concern, expressed in Garska v. McCoy,18 that it would be “terrifying” to the primary caretaker of a child for a court to take the child away if the primary caretaker in question was a mother. No court has ever expressed a concern, or felt a need, to protect a grown man from becoming “terrified” about being made a mere visitor to his children. To the contrary, courts have been doing that to men for centuries. The purpose of establishing a primary caretaker doctrine was to ensure that mothers would continue to receive sole custody of their children notwithstanding the gender-neutralization of custody law.19

That being the case, it would make sense for courts to define caregiving in terms of functions traditionally performed by women, while omitting caretaking functions traditionally performed by men. And that is exactly what they have done. Functions traditionally performed by women – such as cooking; cleaning; driving children to medical appointments; and shopping – typically are included in lists of “caregiving” functions that courts and evaluators use to decide which parent is the primary caregiver. Meanwhile, functions that historically were more often performed by males – taking out garbage; washing dishes; doing household repairs; mowing lawns; earning the money to make shopping for groceries and clothing possible; maintaining and repairing the vehicle that is used to drive children to medical appointments; and developing a child’s intellectual, social and motor skills through physical activities — typically are omitted from the list.

Penalizing working women?

Feminists have criticized the rule, too, though for different reasons. They argue that establishing a preference for stay-at-home mothers penalizes working mothers for pursuing careers rather than staying at home to care for their children.20 They say that the standard places too much emphasis on the quantity of caregiving (“counting hours”), and not enough on quality. This, it is said, overvalues the performance of caretaking tasks while undervaluing the strength of the bond between a parent and the child.21

It is this kind of concern that led the Minnesota legislature to prohibit courts from applying the primary caretaker presumption, not long after having required them to apply it. Largely at the urging of feminists, the legislature concluded that courts should have discretion to choose to favor the primary caretaker for custody in some cases, such as those involving stay-at-home mothers, and to favor other factors in other cases, such as those involving working women.

Unreasonably cementing caregiving patterns

Another criticism of the primary caretaker standard is that it fails to take account of the fact that caregiving patterns can, and should, change as children grow and change. Child development professionals generally agree that it is normal and healthy for a child’s attachments and needs to change over the course of his or her childhood. One parent may have a stronger attachment bond, or a better ability to meet a child’s needs, when the child is younger; but the child may have a stronger attachment or need for the other parent when he is older. Needs and attachments can alternate several times throughout childhood. A custody arrangement for a fourteen-year-old boy that is based on which parent nursed him the most when he was a baby may not necessarily be in his best interests.

Caregiving vs. attachment

Historically, the motivation for establishing a primary caretaker presumption was to ensure that mothers would continue to receive sole custody of children; and this was also the motivation for abolishing it. Despite the inauspicious motives behind them, these conflicting changes in the law brought a larger question into sharper relief: Why should the relative quantity of time a parent spends performing certain specified functions be more important than the kind of attachment, or bond, the child has formed with the parent? Why should the law automatically favor the parent who is cooking dinner while the other parent is teaching the child how to perform a house or car repair, or developing his motor skills and self-esteem through play? Why shouldn’t the law favor the parent to whom the child turns for solace, comfort and protection, over the cold, aloof parent who happens to drive the child to 51% of her doctor appointments?

It is difficult to find any good answers to these questions.


States considering a primary caretaker presumption should take note of the Minnesota experience. They should also consider whether they are defining caregiving functions in a sexist manner, and whether it is even possible to quantify the many tangible and intangible ways that parents contribute to the care of their children. They should consider the potential impact of a caregiving presumption on the interests of working mothers (and fathers.) They should ask whether it makes sense to “reward” the parent who can establish that he or she provided 51% of the child’s care with a judgment giving him or her 100% of it (sole custody.) They should consider whether establishing a presumption favoring awards of sole custody to the parent performing the largest quantity of specified caregiving functions might be harmful to the bond a child may have established with the other parent, or with both parents. They should consider whether focusing exclusively on early patterns of caregiving entrenches gender roles, or impedes growth and change. They should question whether deciding custody in this way really has much at all to do with children’s best interests, or whether it is really more about protecting parents – specifically, parents of the more fragile sex — from the “terror” of not winning the title of sole custodian. Most importantly, they should ask whether and why the comparative performance of specified caretaking tasks should be weighted more heavily than other, arguably more important factors, such as the quality of care a parent provides and the strength of the bond between child and parent.


The complete History of Custody Law is now available in paperback and as a Kindle e-book:

Purchase at


  1. Garska v. McCoy 278 S.E.2d 357 (W. Va. 1981); see also Pikula v. Pikula, 374 N.W.2d 705 (Minn. 1985)
  2. Gary Crippen, Stumbling Beyond Best Interests of the Child. Reexamining Child Custody Standard-Setting in the Wake of Minnesota’s Four Year Experiment with the Primary Caretaker Preference, 75 MINN. L. REV. 427, 452 (1990); see also Paul L. Smith, The Primary Caretaker Presumption: Have We Been Presuming Too Much? 75 IND. L. J. 731, 743 (2000)
  3. 724 P.2d 486, 492 (Cal. 1986)
  4. Id. at 494 (Bird, C.J., concurring)
  5. Laura Sack, Women and Children First: A Feminist Analysis of the Primary Caretaker Standard in Child Custody Cases, 4 YALE J.L. & FEMINISM 291, 302 (1992).
  6. Id.
  7. See L.C. Sayer, Economic aspects of divorce and relationship dissolution, in HANDBOOK OF DIVORCE AND RELATIONSHIP DISSOLUTION 385–406 (M. A. Fine & J. Harvey eds., 2006.)
  8. S.D. Hoffman & G.J. Duncan, What are the economic consequences of divorce, 25 DEMOGRAPHY 641 (1988). Other researchers purport to have proven a gender gap between mothers and fathers in terms of post-divorce economic well-being, too, though none nearly as extreme as the one Weitzman had claimed. See, e.g., S.M. Bianchi & J.R. Kahn, The gender gap in the economic well-being of nonresident fathers and custodial mothers, 36(2) DEMOGRAPHY 195-203 (1999).
  9. Associated Press, Huge Gap Reported in Post-Divorce Standard of Living a Mistake (May 16, 1996); R.R. Peterson, A re-evaluation of the economic consequences of divorce. 61 AM. SOC. REV. 528-36 (1996); Lenore J. Weitzman, The economic consequences of divorce are still unequal, 61 AM. SOC. REV. 537-38 (1996).
  10. Sayer, supra note 7.
  11. P.A. McManus & T.A. DiPrete, Losers and winners: The financial consequences of separation and divorce for men, 66 AM. SOC. REV. 246–268 (2001).
  13. See W.V. Fabricius & S.L. Braver, Non-child support expenditures on children by nonresidential divorced fathers: Results of a study, 41(3) FAM. CT. REV. 321-336 (2003).
  14. Sanford L. Braver, Jenessa R. Shapiro & Matthew R. Goodman, The Consequences of Divorce for Parents, in HANDBOOK OF DIVORCE AND RELATIONSHIP DISSOLUTION 313-38 (Mark A. Fine & John H. Harvey eds., 2006).
  15. Pamela Laufer-Ukeles, Selective Recognition of Gender Difference in the Law: Revaluing the Caretaker Role, 31 HARV. J.L. GENDER 1, 47 (2008); see also MARTHA ALBERTSON FINEMAN, THE ILLUSION OF EQUALITY: THE RHETORIC AND REALITY OF DIVORCE 180-85 (1994); David L. Chambers, Rethinking the Substantive Rules for Custody Disputes in Divorce, 83 MICH. L. REV. 477, 501 (1984); Jon Elster, Solomonic Judgments: Against the Best Interest of the Child, 54 U. CHI. L. REV. 1, 16–21 (1987); Martha Fineman, Dominant Discourse, Professional Language, and Legal Change in Child Custody Decisionmaking, 101 HARV. L. REV. 727, 771–73 (1988)
  16. cf. Krista Carpenter, Why Are Mothers Still Losing: An Analysis of Gender Bias in Child Custody Determinations, 1996 DET. C.L. REV. 33 (1996); see also Sack, supra note 5 at 292-93 (arguing that it establishes a sexist double standard unfavorable to women.)
  17. Crippen, supra note 2 at 494, n.227; see also Ronald Henry, ‘Primary Caretaker’ Is It a Ruse? FAM. ADVOC. 53 (Summer 1994); Judith Band Jennison, The Search for Equality in a Woman’s World: Fathers’ Rights to Child Custody, 43 RUTGERS L. REV. 1141, 1153 (1991); Cynthia A. McNeely, Lagging Behind the Times: Parenthood, Custody, and Gender Bias in the Family Court, 25 FLA. ST. U.L. REV. 891, 916 (1998)
  18. 278 S.E. 2d 357, 360 (W.Va. 1981)
  19. See, e.g., Burchard v. Garay, 724 P.2d 486, 494 (Cal. 1986) (Bird, C.J., concurring) (“Typically, it is the mother who provides most day-to-day care, whether or not she works outside the home. A presumption which ignores this fact is likely to lead to erroneous and unfair decisions.” (citation omitted)). Although no longer giving it presumptive effect, courts today still attempt to ascertain which parent has been the primary caregiver. And they still continue to use the arguably sexist definition of caregiving functions that was established in the 1980’s when courts sought to establish a substitute for the maternal preference doctrine that would have the outward appearance of being gender-neutral.
  20. Mary Becker, The Sixties Shift to Formal Equality and the Courts: An Argument for Pragmatism and Politics, 40 WM. & MARY L. REV. 209, 256 (1998) (describing it as entrenching gender roles in society.)
  21. Crippen, supra note 2 at 489-92; see generally Sherri A. Ahl, A Step Backwards: The Minnesota Supreme Court Adopts a “Primary Caretaker” Presumption in Child Custody Cases: Pikula v. Pikula, 70 MINN. L. REV. 1344 (1986).