Journey Across a Century
of Women


claudia goldin, a past president of the American Economic Association, is a professor of economics at Harvard. This article is adapted from her NBER 2020 Martin S. Feldstein Lecture, available at

illustrations by sean mccabe

Published April 26, 2021


This article will take us on a 120-year odyssey of generations of college-graduate women from a time when they were only able to have either a family or a career, to now, when they anticipate having both a family and a career. More women than ever before are within striking distance of these goals. Fully 45 percent of young American women today will earn bachelor’s degrees, and more than 20 percent of them will eventually obtain an advanced degree above a master’s. In addition, more than 80 percent of 45-yearold college-graduate women have children, either biological or adopted. More women than men now graduate from college, and there is greater similarity in their ambitions and achievements than ever before. This should all make for a very pleasant ending to the journey. But that happy ending doesn’t seem to be happening.

Before we delve into why that is, a few clarifications. My evidence concerns the United States and the history of its college-graduate men and women. I focus on women with college degrees because they have the greatest opportunities to achieve a “career.” Career is achieved over time, as the etymology of the word — meaning to run a race — implies. A career generally involves advancement and persistence and is a long-lasting, sought-after employment for which the type of work — writer, teacher, doctor, accountant, religious leader — often shapes one’s identity. A career needn’t begin right after the highest educational degree and can emerge later in life (just not too late).

A career is different from a job. Jobs generally do not become part of one’s identity or life’s purpose. They are often solely taken for generating income and generally do not have a clear set of milestones.

Before the Coronavirus Era

I recently finished most of a book on this century- long journey. But my book, like the Old Testament, was written in a BCE world — in this case, that is, Before the Coronavirus Era. Many inequities have been exposed by the Covid-19 economy and society, most notably those concerning social justice and our criminal justice system. The Covid-19 economy has also magnified gender differences at work and in the home. Women are essential workers. But impossibly suddenly now at home and at work simultaneously. The burden of school closings on working parents threatens to erase years of career gains by young women in a way we have rarely seen.

That is where this article will take us. But, first, we must journey to that moment.

I’ll begin 120 years ago, when collegegraduate women were faced with a stark choice: family or career (or sometimes a job). Five distinct groups of women can be discerned across the past 120 years, according to their changing aspirations and achievements.

• Women in Group 1, which graduated college between 1900 and 1919, aspired to attain family or career. Few managed both. In fact, they split into two groups: 50 percent never bore a child, 32 percent never married. In the portion of Group 1 that had family, just a small fraction ever worked for pay.

• Group 2 was a transition generation between Group 1, which had few children, and group 3, which had many. More Group 2 women aspired to careers, but the Great Depression intervened, and this transitional generation got a “job then family” instead.

• Group 3, the subject of Betty Friedan’s Feminine Mystique, graduated college between 1946 and 1965. As America was swept away in a tide of early marriages and a subsequent baby boom, Group 3 women shifted to planning for a family then job. Just 9 percent of the group never married, and just 18 percent never bore a child. Even though their labor force participation rates were low when they were young, those rates rose greatly — to 73 percent — when they (and their children) were older. But by the time these women entered the workplace, it was generally too late for them to develop their jobs into fullfledged careers.

• Career then family became a goal for many in group 4, my generation, which graduated between 1966 and 1979. This group, aided by the Pill, delayed marriage and children to obtain more education and a promising professional trajectory. Consequently, the group had high employment rates when young. But the delay in having children led 27 percent to never have children.

• Group 5 continues to today. At present, for group 5, the goal is career and family — and although they are delaying marriage and childbirth even more than Group 4, just 21 percent don’t have children.

You may be thinking that, because of large increases in college graduation rates, most of the differences across the groups concern selection in who goes to and graduates from college. The surprising finding is that selection is not that important. I’ve tracked college entrance classes among women who have similar ability and parental resources from the 1890s to the 1990s. Their marriage ages and birth fractions track those of the total college-graduate group astoundingly well. “Treatment,” not “selection,” dominates.

College for Group 1 had a treatment effect by enabling women to be financially independent — they didn’t have to marry. After Group 1, as women’s potential earnings rose and as substitutes for household goods became cheaper, husbands’ expectations rose rather than necessarily changing. Though family came first for Group 3, women planned their home confinement — and their eventual escape. They trained to be teachers, nurses, social workers, librarians and administrators after the kids were sufficiently grown. For Group 4, the Pill and its dissemination to young, single women enabled delay of marriage and family and helped boost their investment in careers. But the biological clock ran out for many of these women. Group 5 has pushed back marriage and family even further, but birth rates have been up, in part due to assisted reproductive technologies that have enabled the group to beat this clock.

The Times They Are a-Changing

The transition wasn’t swift, and it wasn’t due mainly to dissent. Instead, it was often due to technological advances, increased earnings and greater education.

The aspirations and achievements of women greatly changed across the past century, with increased income, the mechanization of the household, and technological improvements in fertility control and assisted reproductive methods. But the way work is structured and the persistence of social norms, no matter how much weaker they have become, have limited the success of college- graduate women in achieving career and family.

An important accompaniment to the transition across the groups concerns changes in customs and norms. For the past 50 years, the General Social Survey (conducted by an organization affiliated with the University of Chicago) has asked respondents whether they agreed more or less strongly with the statement: “Preschool children are likely to suffer if their mother works.”

The responses are depicted by the respondent’s birth year. As can be seen, agreement is always less for women than for men and decreases for both men and women by birth cohort. (It also increases with age, since earlier birth cohorts were generally older when interviewed than more recent birth cohorts.) Norms became more expensive to sustain, and they changed. At the same time that the cost of not working — calculated in terms of foregone earnings — rose, child care became more available, commonly used and generally acceptable.

To measure the degree of “success” at both career and family these groups achieved, I created some definitions. Family means having a child (biological or adopted), but not necessarily a husband or partner (sorry, dogs do not count as surrogate children). Career is achieved by exceeding a certain level of income for three years in each five-year period — where the level in question is the income of men of the same age and education at the 25th percentile of the male distribution.

Passing the Baton

Interestingly, “success” at career and family for women has increased both across and within cohorts. Success at both for the latest group (born 1958 to 1965) is around 30 percent for women in their early 50s, or half of what it is for comparable men. But success for that birth group is just 22 percent for women in their late 30s, or still just half the success rates of comparable men (at 40 percent).

Even though a succession of women, group after group, advanced on this journey, women’s careers still often take a back seat to those of their husbands. Group 5 has expressed its disappointments and frustrations by focusing on issues such as bias, pay inequity, salary transparency and sexual harassment.

But, as each group progressed and passed the baton to the next, and as actual barriers fell and social norms changed, the real underlying problem that fuels differences in occupations, promotions and pay has been revealed. There is no question that there are cases of classic discrimination, bad actors, sexual harassment and biased workers and supervisors. But most of the difference is due to something else.

To paraphrase Betty Friedan, the new “problem with no name” is the notion of “greedy work” — the idea that the most prestigious, highest-paying jobs absorb ever more time and pay more even on an hourly basis. To have a family takes the time of at least one parent: there is no way to contract out all child care. And one wouldn’t want to do that, anyway. Why have children in the first place? Parents have children to spend time with them. For college graduates, the gender gap in earnings is an indication and a symptom of career blockage.

Greedy Work

Women earn less than men, on average. The ratio of women’s to men’s earnings, often adjusted by hours of work, is referred to as the gender earnings gap. The ratio for all workers narrowed considerably from the early 1960s until today, but remains around 0.8. The ratio for college-graduate women to college-graduate men followed a similar path until the late 1980s, when it flattened out.

Some of the clues to why the ratio is still substantial, and why the ratio for collegegraduate women to men became smaller than for the aggregate after the late 1980s:

• The gap exists for both annual earnings and those paid on an hourly basis, so it is not just due to the fact that women work fewer hours.

• Women with children earn less than women without children.

• Earnings gaps increase with age (to a point), and they increase with (joyous) events like births and often marriage.

• Gaps are greater at the upper end of (male) earnings and education levels.

• The more unequal earnings are within occupations.

• The gap is greater in occupations that make greater demands on employee time, and where “face-time” and client relationships matter most.

The flip side to gender inequality is couple inequity. Working mothers are “on-call at home,” whereas working fathers are “on-call at work.” The reasons for both gender inequality and couple inequity are the same. The issues are the two sides of the same problem mentioned above: that is, greedy work.

Many jobs, especially the higher-earning ones, pay far more on an hourly basis when the work is long, on-call, rush, evening, weekend and unpredictable. And these time commitments interfere with family responsibilities. The problem is illustrated in below.

What can we expect to happen to the child care and home-schooling burdens placed on parents? For most mothers, the AC/ DC world will be the BCE world on steroids.

One job is flexible (orange line), and wages grow proportionately with hours worked. The other is not-so-flexible (black line) and has a wage rate that rises with hours. A couple with children can’t both work at the blue dot. They could both work at the orange dot. But if they did, they would be leaving a lot of money on the table — the distance between the two dots. So, one works the flexible (less remunerative) orange job, and the other works the less flexible (more remunerative) black job. More often than not, the man takes the less-flexible, higher-paying job.

And that gives rise to a gender gap in earnings. It also produces couple inequity. If the flexible job were more productive, the difference would be less, and family equity would be cheaper to purchase. Couples would acquire it and reduce both the family and the aggregate gender gap. They would also enhance couple equity.

Note that even if these were same-sex couples, there could still be couple inequity but without gender inequality. And even if couples wanted a 50-50 relationship, high earnings for the position that had less controllable hours could entice them to engage in a new version of an old division of household labor.

What are the solutions? First off, any solution must involve lowering the cost of temporal flexibility. The simplest is to create good substitutes. Clients could be handed off seamlessly. Information technology could be used to pass information with little loss in fidelity.

Teams of substitute workers, as opposed to complementary workers, could be created — as they have been routinely in pediatrics, anesthesiology, veterinary medicine, personal banking, trust and estate law, software engineering and primary care. The cheaper the solution to job inflexibility, the more proportional pay becomes by hours worked.

Child Care in the Time of Coronavirus

But the tale I have been telling has been set in the Before-the-Coronavirus-Era (BCE) world. In March 2020, suddenly and swiftly, we descended into a During Coronavirus (DC) world. Most workers sheltered in place and worked from home. Fortunate children had on-line schooling and at-home help. Less fortunate workers were deemed essential and often worked in unhealthy circumstances. Less fortunate children lost valuable schooling.

How does our understanding of the BCE world of career and family help us understand the impact of the new DC era and what will come after? Again, I focus on collegegraduate, employed adults and their families.

These parents and their children are clearly in the more-fortunate category. I use the U.S. Bureau of Labor Statistics’ American Time Use Survey from 2010 to 2019 to compute child care hours of mothers and fathers by the age of their youngest child.

The child care hours of the mothers in the BCE world are shown by the turquoise bars in the graph to the left, and the percentage of the total parental child care hours that they performed is shown above the bars. In the BCE world, college-graduate employed mothers (with college-graduate employed husbands and school-aged children) were doing around 60 percent of total child care hours. That percentage was higher for mothers with the youngest children and lowest for those with the oldest children.

The descent into the DC world almost doubled total child care hours for working couples with children. The orange bars denote the hours mothers contributed to child care and are derived from various surveys done in the United States and elsewhere in April, combined with plausible assumptions.

In mid-March 2020, almost 90 percent of school-aged children were not physically in a school, and most child care facilities for younger children were shuttered. Many families had also temporarily furloughed care workers who worked in their homes. That greatly increased the child care demands on mothers. But there was also more parental sharing, since many households had both parents at home full-time. Consequently, the fraction of child care performed by women fell, even as the absolute number of hours greatly increased. (For those with sick relatives, other care hours also increased, and for single mothers, care hours must have been overwhelming.)

We are now moving to an entirely new After Coronavirus/During Coronavirus (AC/DC) world. Draconian pandemic restrictions have been partially lifted in some schools and daycare facilities. But, by the same token, fulltime work has returned to many offices, stores, workplaces, construction sites and factories. What can we expect to happen to the child care and home-schooling burdens placed on parents? For most mothers, the AC/ DC world will be the BCE world on steroids.

One possible scenario is that total child care and home-schooling hours will be halfway between what existed in the BCE and DC worlds. That makes sense if schools and child care are available half the week. But because of rigidities in work schedules, one member of the couple could go back to work full-time, and the other could work part-time from home and take care of the kids whenever inperson school is out and virtual school is in. If history is any guide, men will go back to work full-time and revert to their BCE child care levels. Women will take up the slack and do a greater share of the total.

The bottom line may be that there will be no net gains for working women in the AC/ DC world. What they gain from minimal school and daycare re-openings, they lose from reduced help at home from dad. Because jobs requiring long hours pay more per hour, couple equity will remain an expensive luxury. That expense persists from the BCE world, and the careers of many young women will take a back seat on this journey.

What Might Be Done

The corrective in the BCE world was to “change the workplace”: that is, drive down the price of flexibility. The corrective in the AC/DC world must add “change the careplace”: drive down the cost of child care (and other family demands). But how can one do that safely and equitably?

When public and free elementary schools spread in the United States in the 19th century, and when they expanded during the high school movement of the early 20th century, a coordinated equilibrium was provided by good government. Good government today could do the same thing. We need to find safe ways to have classes for children — for their future and for their parents’.

As in the Great Depression, we now have high rates of unemployed labor. Today, many of the unemployed are highly educated recent college graduates and gap-year students with little to do. They could be harnessed in a new Works Progress Administration (WPA) manner and put to work educating children, especially those from lower-income families. This could free parents, especially women, to return to work. I’ll repurpose an old name and call them the “Civilian College Corps” — a new CCC.

Some of the Corps’ educational work could be done remotely. The Corps could support beleaguered parents too exhausted to correct their children’s essays and too confused to help their children with algebra. Other Corps members could be in the classroom, helping districts cope with reduced numbers of older teachers who don’t want to return to a school building. The Corps would employ those without jobs, providing meaning and direction and giving them something worthy to do: educate the next generation and assist women to go back to work fulltime, either in their homes or on site.

In the BCE world, if the cost of flexibility were much lower, we could have solved the problem of greedy work and achieved both gender equality and couple equity. In the AC/ DC world, we must also reduce the cost to parents of educating and caring for children of all ages. The original journey was from career or family to career and family. The journey continues.

main topic: Gender