Schumpeter was disturbed by the idea of subjecting parent­hood to modern cost-benefit analysis because, as he saw it, the resulting balance sheet was incomplete, if not fundamentally wrong, especially when it came to the benefits of motherhood. He explained that “the contribution made by parenthood to physical and moral health—to ‘normality’ as we might express it—particularly in the case of women, almost invariably escapes the rational searchlight of modern individuals who, in private as in public life, tend to focus attention on ascertainable details of immediate utilitarian relevance and to sneer at the idea of hidden necessities of human nature or of the social organ – ism.”31 The trade-off for women between rearing children in a traditional family home and living a childfree postmodern lifestyle, in other words, involves a calculated choice in which everything that counts is rarely counted accurately.

Capitalist activity inculcates a cost-benefit mind-set that poses a knotty problem for calculating choices that affect many aspects of family life, particularly childrearing activities that hinge on voluntary sacrifice and altruistic behavior. To what extent does the rational-choice approach explain family deci­sions, and how does it account for the “hidden”benefits, which Schumpeter perceived as being outside the purview of rational calculations based on ascertainable details of immediate utili­tarian relevance? Or do hidden benefits even exist? Nobel Prize-winning economist Gary Becker’s systematic applica­tion of economic analysis to the hard choices faced by family members offers some interesting insights—and raises some additional questions.32

According to rational choice, the traditional division of labor has been widely adopted because it offers a highly effi­cient way for most people to achieve their family objectives, by drawing on the comparative advantages of both men and women. These comparative advantages stem in part from spe­cialized investments made by men and women, and in part from intrinsic differences between the sexes, particularly sur­rounding the reproductive process.33 (The relationship between biology and human nature is a veritable minefield, through which Becker deftly maneuvers on the bland vernacular of comparative advantage, specialized investments, marginal util­ity, and an arcane array of algebraic equations.) On a good day, men’s biological contribution to the production of children may involve twenty minutes or so of physically pleasurable ac­tivity. Women, on the other hand, carry and grow the fetus for nine months, endure the risks and severities of childbirth, and produce milk to nurture the child once it is born. In terms of sheer time and physical effort invested, the differences are in­comparable. Thus, it is reasonable to expect that the substan­tial investment by women would make them more prone than men to devote themselves to the care of their children.

Biology not only encourages women to go to greater lengths in ensuring their high investment in reproduction, it also affords a comparative advantage after childbirth when, for example, the sound of an infant’s crying releases oxytocin into its mother’s bloodstream, stimulating the flow of her breast milk.34 There are numerous benefits associated with breast­feeding, from lower rates of pneumonia and meningitis in children to lower rates of cancer and osteoporosis in mothers. Women are more responsive to their infants’ usual crying than men (though both are equally attentive to cries of extreme dis­tress).35 But while women’s comparative advantage may be strong, it is not absolute. In practice, less than half of all infants are breastfed, and only 18.5 percent are breastfed for at least six months.36 And just as some mothers are capable of child abuse, many fathers are capable of providing excellent care for young children. Yet in most cases Alice Rossi’s 1978 statement still holds true: “The mother-infant relationship will continue to have greater emotional depth than the father-infant rela­tionship because of the mother’s physiological experience of pregnancy, birth and nursing.”37 She went on to note that while a society could try to override these biological propensi­ties by training boys and men in infant care, it was not entirely clear how well such efforts might work out.

Women working in the household can nurture and care for older children at the same time that they produce addi­tional children more easily than those employed in the market economy. This complementarity between childbearing and childrearing augments the scale of efficiency in the traditional sexual division of labor. Moreover, the biological differences between men and women have typically overlapped with and reinforced differences in the kinds of human capital invest­ments they experience.38 Thus, until recently, the efficiencies and comparative advantages of the traditional division of labor, which related largely to childrearing, resulted in special invest­ments in education and training that would prepare women to assume the traditional homemaker’s role and men to parti­cipate in the paid labor force. In i960 relatively few college women were enrolled in prelaw, premed, or Ph. D.-oriented

programs. Instead, they studied the liberal arts and sought to graduate with a B. A. Many also graduated with an “M. R.S.”— the median age of marriage for women was 20.3 at that time.

All that has changed, however. Today more women than men are going to college, and they represent an increasing pro­portion of graduates from professional schools and Ph. D. pro­grams. The biological propensities in support of the tradi­tional sexual division of labor in childrearing have collided with the cultural forces of women’s liberation. Women are in­vesting substantially in increasing their human capital not for household productivity but for paid employment in the mar­ket economy. Accordingly, the median age for a first marriage has climbed to over twenty-five. Such changes have altered the calculations of efficiency and comparative advantage. When the income a female doctor or lawyer can earn pursuing a pro­fessional career outweighs the productive value of her house­hold work, her time is worth more in the labor force than at home. For the lawyer’s family members to get the most eco­nomic value out of the investment of their time, either they should outsource as much cleaning, cooking, and care work as possible or the husband should increase his participation in household work, particularly if he earns less than his wife. It might even make sense from a cost-benefit perspective for the husband to remain at home and take on all of the childrearing and household work—assuming that the intangible costs and benefits to the parents and children of this arrangement re­mained constant (a thought I shall revisit shortly).

As educational achievement advanced and employment opportunities expanded, many women expected to have a ca­reer, children, and a husband who would help clean, cook, and care for the kids—in sickness and in health and particularly in the middle of the night. This view emerged somewhat in re­sponse to the perception of shifting efficiencies and compara­tive advantages of engaging in household versus market ac­tivities. And it was reinforced by the idea that the traditional sexual division of labor was essentially a social construct that had very little to do with biological differences. From the 1960s to the mid-1980s the second-wave feminists, led by what Christina Hoff Sommers identifies as “gender” feminists, imagined that men could be convinced to renounce their so­cially constructed roles and step up to the plate—or, more ac­curately, the sink, the washing machine, and so on—to share equally in the traditionally female household and child-care duties.39 Indeed, studies of the changing sexual division of labor suggest that some of this is happening—but perhaps not entirely as the gender feminists might have hoped.

There is persuasive evidence that despite all the changes in education and paid employment since the 1960s, women continue to assume the brunt of household and child-care re­sponsibilities, resulting in what Arlie Hochschild dubs the “stalled revolution.”40 In countries where parental leave allows for fathers and mothers to share the time off from paid em­ployment, for example, women consistently use more of the leave than men. And there is strong indication that when men earn less than their wives, they share even less in household work then those who earn more than their wives.41 This beha­vior is explained as an effort to compensate for the damaged ego associated with failure to assume the normal (socially con­structed) male identity as the primary breadwinner. Whether due to biological indispositions, inadequate socialization, in­sufficient ego, level of outside earnings, sluggish character, or some combination thereof, these findings suggest a degree of intractability in the roles men are willing to assume on the home front.

“Intractability” may be too strong a word, however, be­cause the sexual division of labor has not remained constant over the decades. But precisely how much time and effort spouses invest in household and child-care duties is difficult to gauge. Even when examining nine-to-five employment, it is not easy to take an accurate reading of how people spend their time, except perhaps for a few occupations where time and activities are tightly framed by the job—for example, trolley drivers who stay on a track and lawyers who charge for their services by the minute. (A professor’s work, by contrast, is never done—some say their best ideas have come while sleep­ing.) When it comes to family life, studies that derive estimates of unpaid household labor based on time-use diaries probably yield the most accurate information. Among recent studies using this method, two of the most thorough analyses reveal several prominent trends in the changing investments in em­ployment, household work, and child-care activities since the mid-1960s.

Findings on the balance between paid and unpaid work from 1965 to 1999 show a changing pattern of behavior among men.42 Although women did about 40 percent more unpaid work than men in 1999, the amount of time men devoted to housework and child care had more than doubled since 1965, rising from an average of 6.2 to 14.7 hours per week. However, 50 percent of men’s housework activities involved what are typically considered masculine chores, such as maintenance and outdoor work, whereas only 21 percent of women’s house­work concentrated on these activities.43 Most of the increase in men’s unpaid work took place between 1976 and 1985, after which it slowed to a crawl. Women substantially reduced their housework from an average of 30.4 hours per week in 1965 to 16.8 hours in 1999; however, their average investments in child care showed almost no difference, going from 7.4 hours per week to 7 hours per week. The latter finding bears further comment since this average includes women with and without children and does not distinguish between employed and stay – at-home mothers.

The amount of time devoted to child care is an ambigu­ous subject that lends itself to varying interpretations. Is a mother playing soccer with her fifteen-year-old daughter en­gaged in child care (noted in the time diary under the child­care category of time spent on outdoor play), or is this just a leisure activity? What about the parents watching from the bleachers while their children play Little League baseball or soccer (and watching the teams play even when their children are not on the field)? Under the category of time spent talking to kids, does a mother talking to her son while preparing din­ner qualify to the same extent as sitting at the table having a cup of coffee and talking to her child? Then there is the issue of whether all child-care activities are socially equivalent. Since 1965 there has been a significant increase in the amount of time spent driving preschool children to and from day care. How does spending five hours a week driving to and from day-care centers (which is identified as a child-care activity in time­diary codebooks) measure up against reading and playing with a child for that amount of time?44 If a mother spends two hours a week volunteering in her son’s first-grade classroom, does it count as child care? Finally, there is the question of how to count “stand-by” time—when, for example, a two-year-old is napping but might get up at any moment hungry, wet, or ready to play. Babysitters and child-care-center employees are paid to stand by while children nap.

The ambiguity surrounding child care is complicated by the fact that today caring for children is a sensitive topic, par­

ticularly for the majority of employed women with young children. Mothers do not want to feel that by altering their tra­ditional roles they have in any way shortchanged their kids. Thus, reporting on the 2006 findings from a highly detailed and methodologically rigorous study of parental time devoted to child care, Ann Hulbert offers the soothing interpretation that the latest data do not support “the concern that kids have been shortchanged as women have flocked to the work force.” As evidence, she explains that “the sociologist Suzanne M. Bianchi and her associates at the University of Maryland have noted that the average employed mother in 2000 recorded the same amount of primary child-care time (roughly 10 hours a week) as the average at-home mom did in 1975.”45 Indeed, most stories in the national press emphasize, as Robert Pear did in the New York Times, that “despite the surge of women into the work force, mothers are spending at least as much time with their children today as they did 40 years ago.”46 This widely cited finding is accurate as far as it goes—but it is not the entire story.

According to Bianchi and her colleagues, in 2000, em­ployed mothers devoted as many hours to primary child care (that is, activities focused predominantly on children—such as reading to and dressing them) as nonemployed mothers did twenty-five years earlier (Table 1). But in both 1975 and 2000, the number of hours spent on primary child care by nonem­ployed mothers was more than 60 percent higher than that of employed mothers. If by children being “shortchanged” Hul – bert is referring to the amount of time spent with their moth­ers in primary care, then it is true that children of working mothers today are not at a disadvantage compared to kids of nonemployed mothers twenty-five years ago. By the same token, compared to their peers today with stay-at-home mothers, those children are receiving significantly less primary care.

Table 1. Average Time (in Hours) Devoted to Child Care Weekly for Children under Eighteen Years of Age







Primary care




Secondary care









Primary care




Secondary care








Source: Suzanne Bianchi, John Robinson, and Melissa Milkie, Changing Rhythms of American Family Life (New York: Russell Sage Foundation, 2006), table 4.4.

The gap widens when the amount of time spent in secondary child care is considered.47 Secondary care involves the periods of time during the day in which respondents were engaged in an activity such as shopping, cooking, and watch­ing television while with their children. The researchers “as­sume that children and parents can benefit from the interac­tion that takes place when care is secondary, even though there may be great variability in the intensity of the interaction.”48 Secondary care might include, for example, folding the laun­dry while keeping a watchful eye on a child engaged in inde­pendent play. Day-care centers after all do not distinguish be­tween primary and secondary care; keeping a watchful eye on children while they are engaged in independent play or during nap time is just part of the job. In 1975 nonemployed mothers spent more than twice as much time involved in secondary child care as employed mothers did, a difference that increased to almost three times as many hours in 2000.

Combining primary and secondary care, in 2000 stay-at-home mothers spent twice as much time with their children as em­ployed mothers did.

But even these figures do not tell the entire story. Ac­cording to the study, a stay-at-home mother in 2000 spent fewer than four hours a day on total child care—a figure that would probably baffle any stay-at-home mother with a two- year-old. When the amount of care is averaged for all children, from newborn babies through eighteen-year-olds, the results mask an important disparity between the daily effort required to care for children in the developmentally crucial years of pre­school life and that required for school-aged children who are away from home most of the day—regardless of whether the mother is employed elsewhere. Thus, it is reasonable to imag­ine that the considerable difference between employed and nonemployed mothers in average time devoted to care of all children under eighteen years of age increases when the focus shifts just to preschoolers.

The findings on the changing sexual division of house­hold labor are less ambiguous and socially charged than those on child care. What accounts for the increase in household labor among men and the decrease among women? Some of the shift is no doubt attributable to changes in certain charac­teristics of the male and female populations between 1965 and 1999, including age, education, employment, and marital and parental status. Obviously, more women were at home caring for larger families in 1965 than in 1999, which meant there was both more time available and greater demand for cooking, cleaning, and the like. Also in 1999 there were more single men living on their own and thus responsible for household main­tenance than in 1965. The other major factor is changes in behavior—for example, men in 1999 spent less time on the couch drinking beer and watching football and more time working in the garden or the kitchen.49

The degree of behavioral change is estimated by compar­ing the number of hours spent on household work by men and women in 1965 with the number of hours spent in 1999 by men and women with matching characteristics. For example, the housework habits of a woman (or man) who was a college graduate, employed, and married with two children in 1965 are examined against those of a woman (or man) in 1999 living in the same circumstances. According to this analysis, behavioral modifications accounted for most of the increase in men’s housework up through 1990. But after 1990, the small uptick in men’s housework was essentially due to changing characteris­tics among men—specifically, more men were living alone and cleaning bachelor pads. The analysts interpret the declin­ing magnitude of behavioral adjustments as a sign that the ex­tent to which men and women can rearrange their lives to accommodate both paid and unpaid work has hit a ceiling.50 Ultimately, there are only so many hours a day available for re­allocation by people who are employed full-time.

These findings shed empirical light on the extent to which modification of the traditional division of labor benefits capi­talism and, more broadly, on how organizing one’s life accord­ing to rational calculation of costs and benefits shapes family choices. From the rational-choice perspective it comes as no surprise that as women invest more time and effort in devel­oping human capital to boost their market value, they devote less time and effort to household productivity associated with the long-established sexual division of labor. Although men have taken up some of the slack, the average weekly hours spent on housework by men and women combined fell by more than 20 percent between 1965 and 1999. Much of the decline was in discretionary work, such as mopping floors, shampooing carpets, dusting furniture and ironing under­wear, which simply no longer gets done or is done less fre­quently.51 And some of the reduction represents time-saving products—wrinkle-free clothing, self-cleaning ovens, and au­tomatic sprinklers. A substantial amount of household produc­tion, however, has been outsourced to the market. People now hire experts to organize their closets and arrange their daily schedules. Food from supermarkets as well as precooked meals are purchased online and delivered to the door—sometimes with a discount if the delivery is not prompt. People also eat out more often. In 2000, forty-one cents of every dollar Ameri­cans spent on food was consumed in restaurants, compared to twenty-nine cents in 1987.52 As women’s labor has shifted from the home to the market, the art of good housekeeping has ap­parently fallen on hard times.

The family serves not only as a haven that nurtures its members’ physical and emotional lives but also as an economic unit that produces all sorts of goods and services. Educating the young, cleaning the house, preparing food, even changing lightbulbs are all services that contribute to the standard of liv­ing. The economic value of these services is their market price when produced by others—teachers, chefs, housecleaners, and electricians. “Even if you buy a chocolate bar,” says Stein Rin – gen, “it takes a tiny bit of additional work to remove the wrap­ping before you can enjoy it. That’s work. Make no mistake about it: some people pay servants to do it for them. The Prince of Wales is said to have his servant squeeze out tooth paste for him on to his tooth brush.” Based on careful quanti­tative analysis, Ringen estimates that in 1986 half of the per – person value of the total output of goods and services in Britain came from economic activities in families—more than half if the family’s nonpaid care work is included.53 Gary Becker also concludes that nonpaid family work accounts for half or more of society’s economic activity.54

These estimates of the immense amount of family pro­duction underscore the huge transfer of goods and services to the market economy spurred by the 20 percent decrease in un­paid housework. Not only did changes in the traditional divi­sion of labor open the door to capitalist activity in the house­hold, but they also diminished the mutual exchange of services among kin for reasons of emotional attachment and social ob­ligation. As new modes of production emerged, the process of creative destruction (which Schumpeter took as an essential fact of capitalism) struck an intangible, yet profound, blow to family life.

In determining whether the rational cost-benefit ap­proach has influenced women’s roles in the division of family labor, I would conclude that probably neither men nor women are entirely responsive to rational choice as economists mea­sure it, because of norms and values as well as biological propensities. That being said, they have responded palpably to the changing costs and benefits of the traditional sexual divi­sion generated by women’s increasing investment in human capital. The response has been mostly in the direction of women investing their labor in the market instead of the household, for a better return and a higher material standard of living.