FAST 10.1177/0002716204267959 TRACK THE ANNALS OF THETHE LONG ROAD TO AMERICAN ACADEMY
The Long Road to the Fast Track: Career and Family
By CLAUDIA GOLDIN
The career and family outcomes of college graduate women suggest that the twentieth century contained five distinct cohorts. The first cohort, graduating college from 1900 to 1920, had either "family or career." The second, graduating from 1920 to 1945, had "job then family." The third cohort, the college graduate mothers of the baby boom, graduated from 1946 to the mid1960s and had "family then job." Among the fourth cohort, graduating college from the late 1960s to 1980 and whose stated goal was "career then family," 13 to 18 percent achieved both by age forty. The objective of the fifth cohort, graduating from around 1980 to 1990, has been "career and family," and 21 to 28 percent have realized that goal by age forty. The author traces the demographic and labor force experiences of these five cohorts of college graduates and discusses why "career and family" outcomes changed over time. Keywords: college graduate women; career; family
ARTICLE
November 596
W
omen who graduated college at the dawn of the twentieth century were of a generation that contained two groups, equal in size yet highly dissimilar. One married and had children but had few jobs and even fewer careers. The other had no children and married at much 1 lower rates. Many in that group had jobs at some point in their postcollege lives. Some even had careers. Thus, half of this generation of college graduate women had "family." Some in the other half had "career." Few had both.
Claudia Goldin is Henry Lee Professor of Economics at Harvard University and director of the Development of the American Economy program of the National Bureau of Economic Research (NBER). She has been on the faculties of the University of Pennsylvania, Princeton University, and the University of Wisconsin. Goldin was president of the Economic History Association in 2000 and vice president of the American Economic Association in 1991. From 1984 to 1988, she was editor of the Journal of Economic History. Goldin received her B.A. from Cornell University and her Ph.D. from the University of Chicago. NOTE: This article draws on a public lecture presented at "Mommies and Daddies on the Fast Track," October 30, 2003. It relies heavily on Goldin (1992, 1997).
DOI: 10.1177/0002716204267959
20
ANNALS, AAPSS, 596, November 2004
THE LONG ROAD TO THE FAST TRACK
21
Long after this generation, college women began a struggle to "have it all," to have both "career and family." It would not be one or the other. Nor would it be one then the other. It would be both. My article is about the "long and winding road" from the cohort who had "family or career" to the latest generation of college graduate women who define their goal as "career and family." They are women who want to be "mothers on the fast track." It was about fifteen to twenty years ago that I first began to realize that college women as a group were talking about their futures in ways that would have been unimaginable to me when I was in college. They spoke, candidly and honestly, of desiring "CAREERANDFAMILY" or "FAMILYANDCAREER," as if the words were not three but one and as if the timing of the two goals would not be an issue. They were aware that only atypical and extraordinary college woman had been able to accomplish both career and family in the past. But in defense of their optimism, they noted that their generation was different. Barriers had fallen. They were as well trained and as able as their male friends, the majority of whom would achieve these two goals. I became curious about the evolution of this change in "attitude," and over the years I have compiled evidence concerning the labor force and demographic histories of college women from the late nineteenth century to the present.2 The sources range widely and include census materials; Current Population Surveys; U.S. Women's Bureau surveys; alumni records; and for the most recent cohorts, the National Longitudinal Survey (NLS) of Young Women (1968) and the NLS of Youth (1979) (http://www.bls.gov/nls/home.htm). As I looked at the data, I realized that college women across the past century had widely different attainments concerning family and job and career. I want to share these data with you.
Five Cohorts of College Graduate Women
The experiences of the cohorts of college women across the past century suggested to me that the full one hundred years contained five distinct cohorts--not necessarily of equal length by birth year. The cohort boundaries emerged from the data. They are not arbitrary and imposed. Each of these cohorts was faced with a different set of constraints, and each made its choices concerning family and career subject to these constraints. Table 1 will serve to introduce the five cohorts. To summarize, the first cohort that I am able to track graduated college from the beginning of the twentieth century to the close of World War I and had either "family or career."3 The second cohort graduated college from around 1920 to the end of World War II and had "job then family." This cohort was, as will soon be made clear, a transitional generation linking one of low marriage and low fertility rates to one of high marriage and high fertility rates. The third cohort graduated college from around 1946 to the mid-1960s and had "family then job." These women were the college graduate mothers of the "baby boom." The fourth cohort graduated
22
THE ANNALS OF THE AMERICAN ACADEMY
TABLE 1 FIVE COHORTS OF COLLEGE GRADUATE WOMEN Interval When Graduated from Four-Year College 1900-19 1920-45 1946-65 1966-79 1980-90 Approximate Birth Interval (Assuming College Graduation at Twenty-Two Years Old) 1878-97 1898-1923 1924-43 1944-57 1958-68 Characterization of Desired (or Achieved) Family and Career Path Family or career Job then family Family then job Career then family (13 to 18 percent attained) Career and family (21 to 27 percent attained)
Cohort 1 2 3 4 5
college from the late 1960s to the late 1970s and aspired to have "career then family." The fifth cohort graduated from around 1980 to 1990 and is the most recent one that can be studied. Its goal has been "career and family." These are five distinct cohorts. Yet each one's achievements and ability to attain its goals built on both the accomplishments and frustrations of the previous cohorts. I should note at the outset that I must truncate the groups studied with the fifth cohort. The reason is that I can categorize and assess the achievements of these women only if they are now old enough to have been given a chance to have "family and career." I have used a cutoff of about forty years old, and for that reason, the last cohort I can study--cohort 5--is that born in the 1960s. Cohort 5, whose members proclaimed that they would achieve both family and career (and not career first and then family), is now old enough for the researcher to observe if its members have broken through to the "fast track" in substantial numbers. Researchers can also assess how this cohort differed from (or was similar to) previous generations of college women in its ability to achieve career and to combine it with family. And for this particular cohort, researchers can compare the "family and career" success of college graduate women with that of comparable men. The NLS Youth (1979) data set, which I will use to assess the success of cohort 5, has a large sample of both females and males traced over time, something absent in data sets for previous cohorts. The NLS Young Women (1968) data, used to evaluate the career and family accomplishments of cohort 4, began with a large group of male respondents, but attrition led to the discontinuation of that portion of the survey. I will trace the demographic and labor force experiences of these five cohorts of college graduates and discuss some of the constraints they faced, the trade-offs they made, and their "career and family" outcomes. I will first give the "broadbrush" details for each of the cohorts separately (given in tabular form in Table 2) and then link the cohorts together in a more continuous fashion, using graphs.
THE LONG ROAD TO THE FAST TRACK
23
TABLE 2 DEMOGRAPHIC AND ECONOMIC CHARACTERISTICS OF THE FIVE COHORTS College Graduation Nonmarriage Years (by age fifty) Cohort 1, 1900-19 Cohort 2, 1920-45 Cohort 3, 1946-65 Cohort 4, 1966-79 Cohort 5, 1980-90 >30% 15%-20% 8% 12% (but deferral) (too young) No Children (by age forty) >50% 30%-35% 17% 28% 26% Work at Age Thirty (if married) Low 25% 25%-30% 65% 80% Work at Age Forty-Five (if married) 20% 50% 75% 80% (too young) Dominant Occupation Teaching, office work Teaching, office work Teaching Varied professions Varied professions
SOURCE: Goldin (1997); Current Population Survey data (http://www.bls.census.gov/cps/ cpsmain.htm). NOTE: Labor force participation rates, nonmarriage rates, and childlessness rates are approximations based on the decennial censuses.
Cohort 1
Recall that this cohort graduated college during the first two decades of the twentieth century and was born during the last two decades of the nineteenth century. More than 30 percent of this cohort never married by age fifty, a rate that was four times that for their female counterparts who attended no college at all. College graduate men in this cohort, furthermore, had about the same marriage rates as did noncollege men. Among the women who married, about 30 percent had no children. Putting these two facts together (that is, weighting the numbers by their relative proportions in the relevant population) reveals that more than 50 percent of the group did not have children. This rather high rate of childlessness led some contemporaries, in an era of rampant nativism, to ruminate about "race suicide," for a substantial number of these college women were from upper-class families. Many questioned the appropriateness of college for women. Despite their low rate of family formation, these women also had surprisingly low rates of labor force participation when young and married. Even when they were around age forty-five, their participation rate was around 20 percent, on average, for those who had married. Among those who had jobs, teaching was, by far, the main occupation. These women had made a clear choice between family and career. Given the constraints of their day, they could not easily have had both. Many of this generation wrote and spoke of their careers--as teachers, librarians, social workers, and nurses, among other professions--as higher callings. Their careers liberated them from the constraints of marriage and household duties.
24
THE ANNALS OF THE AMERICAN ACADEMY
Cohort 2
Cohort 2 graduated college mainly during the interwar years of the twentieth century. The generation was transitional in yet another sense. The fraction of the cohort who never married by fifty years old was about 15 to 20 percent, a decrease from cohort 1, and the fraction who never had a first birth was about 30 to 35 percent, also a decrease from that experienced by cohort 1. Although the rate of family formation was higher than that for cohort 1, it was much smaller than that for cohort 3.
It was about fifteen to twenty years ago that I first began to realize that college women as a group were talking about their futures in ways that would have been unimaginable to me when I was in college.
The fraction of cohort 2 who worked when young and married was much higher than that for cohort 1. About 25 percent of the group was in the labor force when they were about age thirty (conditional on being married). This cohort was the first whose members did not exit the labor force at the precise moment of marriage but, in general, waited until they were pregnant with their first child. As a group, therefore, they had "job then family." Among those with occupations, teaching was, once again, the most customary. Mary McCarthy's semiautobiographical novel The Group (1954) best epitomizes this cohort. In its description of the "group's" mothers, it also characterizes the previous cohort of college graduate women (cohort 1). The novel depicts a group of eight Vassar women, graduates of the class of 1933, who resolve to be different from their conventional parents. Each embarks on a career ranging from journalist, to veterinarian, to airplane pilot, although most marry and eventually exit the labor force. Their mothers achieve some vicarious satisfaction that their daughters occupied exciting positions, if only for a while. That is, cohort 2 built on the frustrations of cohort 1.
Cohort 3
Cohort 3 graduated from college during the era of the "baby boom"--from the end of World War II to the turbulent and socially transformative era of the mid-
THE LONG ROAD TO THE FAST TRACK
25
1960s. The cohort married and had children at exceptionally high rates. Just 8 percent never married--a rate that was almost as low as it was for women who attended no college at all. Just 10 percent of those who married did not have a first birth in their lifetime. Once again, this is an extremely low rate. Therefore, putting together these numbers, correctly weighted by the proportions in the population, yields just 17 percent of the entire group who were childless in their lifetime. Not only was their rate of marriage high, but their age at first marriage was extremely low by historical standards for college graduate women. The median age at first marriage was less than twenty-three years old. Of those who ever married, a substantial fraction--57.2 percent--married within one year of graduating college or during college (the drop-out rate because of marriage and often pregnancy was also substantial). Given the timing of marriage relative to college graduation, a considerable portion of these women must have met their future husbands while in college, raising interesting issues regarding the role of college as a "marriage market." The fraction of this cohort that was in the labor force when young and married was 25 to 30 percent, or almost the same level as for cohort 2. But the fraction in the labor force at age forty-five was 75 percent, considerably higher than for the previous cohort. These college graduate women had "family then job." Family came first both in terms of priority and timing. College women had gained considerably in "family" but had not advanced much in terms of "career." Cohort 3 became the frustrated group described by Betty Friedan in her influential volume The Feminine Mystique (1963). As a group, its members became increasingly discontent with a labor market that offered college women little in the way of career advancement and with employment officers who often asked them just one question: "Can you type?"
Cohort 4
The women of cohort 4, the "baby boom generation," graduated college during the heady days of the late 1960s and early 1970s. A substantial fraction put off getting married for several years after college. In consequence, the age at first marriage rose by more than two years--from twenty-three years old for the cohort born in 1950 to more than twenty-five years old for that born in 1957.4 But even though this generation married late, its members deferred marriage rather than avoiding it altogether, and just a small fraction--12 percent--remained single by their midforties. Even though a substantial fraction married at some point in their lives, the delay in marriage, together with several other factors, led 19 percent of the ever-married group to have no births by age forty. Together with the group that never married and had no children, about 28 percent of the entire group remained childless by age forty. The labor force participation rate of the group when it was young and married soared relative to that of the previous cohort. About 65 percent of its members were in the labor force at age thirty (conditional on being married). About 80 per-
26
THE ANNALS OF THE AMERICAN ACADEMY
cent were in the labor force at age forty-five (again, conditional on being married). The dominant occupation for college graduate women in this cohort shifted from teaching to a variety of professions, including those at the very top of any occupational prestige scale.
The dominant occupation for college graduate women in [cohort 4, the baby boom generation] shifted from teaching to a variety of professions, including those at the very top of any occupational prestige scale.
College women gained in careers but lost in family. As a group, I will show that about 13 to 18 percent of this cohort achieved "career and family" by age forty. Among those who placed their careers first, children were put "on hold," sometimes forever.
Cohort 5
This cohort, graduating from college in the 1980s--the "decade of greed"--and the last one that can be tracked to around age forty, looked to previous generations for direction. Putting career first and then trying for a family, as was attempted by cohort 4, had led to substantial "childlessness." Putting family first and then getting a job, as cohort 3 did, would not be the route either. Cohort 5 would try for both together. But did it succeed? College graduate women in cohort 5 deferred marriage, just as did those in cohort 4. They achieved a slight decline in the fraction with no births--26 percent by around age forty rather than 28 percent for the previous cohort. Their labor force participation when young and married was extremely high--around 80 percent--and their dominant occupation was a group of diverse professions, as was the case for cohort 4. A larger fraction of this cohort managed to achieve career and family by around age forty. The range is about 21 to 27 percent for cohort 5, as opposed to 13 to 18 percent for cohort 4. The answer, then, is that cohort 5 did succeed to a greater extent than did its predecessors in large part because those with children did better at attaining a career.
THE LONG ROAD TO THE FAST TRACK
27
Cohort Summary
In sum, cohort 1 had either family or career, rarely both. Cohort 2 was a transition cohort who married and had children to a greater degree but who also had a greater variety of jobs when young. By cohort 3, college women had discovered family to an almost equal degree as noncollege women. But these women eventually became deeply frustrated by their treatment in the labor market and by potential employers who did not take them seriously. Cohort 4 delayed marriage and childbirth and aimed to have career first and then family. Cohort 5, recognizing the problems with the biological clock, claimed it could balance the two together. It has achieved more success in managing family and career than cohort 4 and probably has had the greatest achievement in this regard among all cohorts of college graduate women in U.S. history. Whether the fraction of cohort 5 that has attained family and career is large is a question I will address in a moment. Before I that, do I would like to confront an important issue concerning the selectivity of women into the group of college graduates in each of the cohorts and how that might influence my findings. I would also like to give the more continuous details of marriage and childbearing using all years of college graduates to show how I demarcated the cohorts.
College Graduation Rates by Cohorts
The rate of college graduation (from four-year institutions) increased over time for both men and women, as can be seen in Figure 1, but at different moments for much of the century. Graduation rates were not much different by sex for those born early in the twentieth century, such as cohort 1. In fact, the rate of college going (not shown here), rather than graduation from a four-year institution, was about the same for men and women until the birth cohorts of the 1920s. The cohorts of men who fought in World War II and Korea (those born from around the early 1920s to the early 1930s) had vastly increased college going and graduation rates. By the cohorts born in the 1930s, and thus who graduated from college in the 1950s (such as cohort 3), the ratio of men to women in college was two to one or more. But the rate of graduation for young women also began to rise for cohort 3. Although the rates for women rose, college graduation rates for the men of cohort 4 increased precipitously with the draft deferments of the Vietnam War. But by the close of cohort 4, graduation rates were on par by sex and have remained so for cohort 5.5 Thus, the earliest cohorts of college graduates studied here must have been from a highly select group of families by income and social standing. By midcentury, however, the middle class had found its way to college, and by the end of the period, college graduation included individuals from most types of families. The selection of young people into college and the families from which they came could affect the conclusions I have made about marriage, family, and career goals.
28
THE ANNALS OF THE AMERICAN ACADEMY
FIGURE 1 GRADUATION RATES FROM FOUR-YEAR COLLEGES AND UNIVERSITIES FOR MEN AND WOMEN, BY BIRTH COHORT (AS OF THIRTY-FIVE YEARS OLD)
0.4
Cohort 1
Cohort 2
Cohort 3
4
5
Males
0.3
0.2
0.1
Females
0.0 1880 1900 1920 Birth Year 1940 1960 1980
SOURCE: De Long, Goldin, and Katz (2003). NOTE: Data for cohorts born after 1970 are based on extrapolations.
Although selectivity might affect many issues of importance concerning college graduate women, the characterizations of each of the cohorts I have offered remain robust to selectivity concerns. When the group is limited to a particular college, such as a private women's college like Wellesley and Radcliffe, or a coeducational, public institution such as the University of Michigan, the portrayals of each of the cohorts with respect to marriage, family, occupation, and career remain the same. That is, women from particular colleges, and even those from elite colleges, conformed to the trends described.
Marriage and Childbearing for College Graduate Women during the Past One Hundred Years
To connect the various cohorts I have delineated and to give more information about the demarcations selected, I have constructed continuous series for mar-
THE LONG ROAD TO THE FAST TRACK
29
FIGURE 2 PERCENTAGE NEVER MARRIED, COLLEGE GRADUATE WOMEN (WHITE)
30 55-64 years 25-34 years 20 35-44 years 10 45-54 years Cohort 1 0 1880 1890 1900 1910 1920 1930 1940 1950 1960 1970 Cohort 2 Cohort 3 Cohort 4 Cohort 5
Birth Year
SOURCE: Goldin (1992, 1997).
riage rates and childbearing that encompass as many of the cohorts as the data would allow. The data on marriage rates, given in Figure 2, show the very high rate of nonmarriage--about 30 percent for even the oldest age group--for cohort 1.6 In sharp contrast, the extremely low rates of nonmarriage for cohort 3 reveal why cohort 2 was a transitional cohort. The rate of nonmarriage among the members of cohort 3 bottomed out around the 1940 birth cohort when the rate dropped to about 7 percent for women forty-five to fifty-four years old. Note as well that although the rate of marriage for cohort 4 by the time it was thirty-five to forty-four years old was almost as high as for cohort 3, it deferred getting married until it was considerably older. By the start of cohort 5, more than 30 percent of the twentyfive- to thirty-four-year-old group was not yet married, whereas that figure was 17 percent for cohort 3. For the birth data, which are measured when the cohort reached thirty-five to forty-four years, only three of the cohorts can be precisely studied. Cohort 1 was born too early to be tracked with the 1940 population census, and cohort 5 is too young to be studied with any surveys.7 But related data suggest what the childlessness rates of these cohorts were or will be.
30
THE ANNALS OF THE AMERICAN ACADEMY
FIGURE 3 PERCENTAGE OF WOMEN (WHITE) OF ALL MARITAL STATUSES WITH NO BIRTHS BY AGE THIRTY-FIVE TO FORTY-FOUR
50
College graduate
40
H.S. grad, no coll.
30
20
No college
10 Cohort 2 0 1900 1910 1920 1930 1940 1950 1960 Cohort 3 Cohort 4 Cohort 5
Birth Year
SOURCE: Goldin (1992, 1997).
Figure 3 gives the fraction of women by cohort (and educational attainment) who had not yet had a first birth by ages thirty-five to forty-four. Although the trend for college graduate women mimics that for the two other educational groups, it has a more extreme dip and later rise.8 Note that the fraction without children for the youngest in cohort 1 was around 50 percent. Alumni records for various colleges suggest that the fraction may have been higher for all of cohort 1. Also note that by the midpoint of cohort 3, about 15 percent of college graduate women did not have a first birth by ages thirty-five to forty-four, the lowest figure on record here.
Career and Family among Cohorts 4 and 5
I can now return to the more recent cohorts and assess the extent to which they have achieved family and career. First of all, I have to define "family" and "career." No single definition will satisfy everyone, and thus I have employed several.
THE LONG ROAD TO THE FAST TRACK
31
I define "family" as ever having a first birth (adoptive children can easily be added, but they do not change the results by much). Note that having a "family" does not necessarily mean that the individual in question is currently married or that the individual was ever married. "Career" is a more difficult concept. The word comes from the French for "racecourse" and means a person's progress through life. In common parlance, it means a success that is not ephemeral but that exists over some period of time. To assess "career," one needs longitudinal data, and these are, luckily, available for both cohorts 4 and 5 in two of the NLSs. I define "career" as reaching an annual income (or hourly wage) level greater than that achieved by a comparable college graduate man who was at the 25th percentile of the male annual income (or hourly wage) distribution. This accomplishment, moreover, must exist for two or three years running when the woman in question was in her late thirties or early forties. Thus, the definition of "career" entails earning more than a college graduate man whose income is well below that of the median man (but about equal to the median of the female earnings distribution) for several consecutive years.
Many of [the women in cohort 1] wrote and spoke of their careers--as teachers, librarians, social workers, and nurses, among other professions--as higher callings.
To assess both career and family for cohort 4, I use the NLS of Young Women. When the survey began in 1968, these women were fourteen to twenty-four years old. That is, the women interviewed were born from 1944 to 1954 and were thirtyfour to forty-four years old in 1988, around the year when I assess career. I restrict the sample to (white) college graduate women and employ four definitions of career. Two use hourly earnings and two use annual earnings. The latter definition of income will include more individuals, such as the self-employed and those who did not list usual weekly hours. For each income measure, I require that the individual earn at least the cutoff amount for either two years or three years. Because the NLS Young Women skipped some years, these are not all consecutive. The results given in Table 3 show, first of all, that about 29 percent of the group did not have a first birth, consistent with the aggregate data, and that about 25 to 28 percent attained career. The numbers in boldface give the fraction that had both family and career, which ranges from about 13 to 18 percent, depending on the def-
32
THE ANNALS OF THE AMERICAN ACADEMY
TABLE 3 CAREER AND FAMILY FOR THE COLLEGE GRADUATE WOMEN OF COHORT 4 (IN PERCENTAGES) Career No Yes No Yes Kids Family No Kids
a
Kids
Family
No Kids
Hourly Wage: Three Years 55.3 16.7 14.0 14.0 Hourly Wage: Two Yearsc 51.8 13.8 17.8 16.6
Annual Income: Three Yearsb 58.9 15.8 12.9 12.4 Annual Income: Two Yearsd 53.5 11.5 18.4 16.7
SOURCE: National Longitudinal Survey of Young Women (1968) (http://www.bls.gov/nls/ home.htm). NOTE: The "hourly wage" columns use only observations that give annual earnings and usual hours. The income measure uses annual earnings and includes the self-employed. "Kids" are biological and exclude adoptive children and stepchildren. The inclusion of adoptive children does not change the results much. Some of these results differ slightly from those in Goldin (1997) and correct minor errors in them. See text and Goldin (1997) for the definition of "career." The numbers in boldface give the fraction that had both family and career. a. Uses hourly wage measure for 1985, 1987, 1988 (N = 443). b. Uses income measure for 1985, 1987, 1988 (N = 550). c. Uses hourly wage measure for 1987, 1988 (N = 477). d. Uses income measure for 1987, 1988 (N = 550).
inition of career used. Whether one judges this to be high or low will depend on a reference group, and there is none in this survey. But there is one in a subsequent survey. Luckily, the NLS 1979 has followed both men and women who were fourteen to twenty-one years old in 1979. These individuals, born from 1957 to 1964, are members of cohort 5. I employ the same definitions of career and family, except that career is now calculated with respect to the college graduate man at the 25th percentile in a later year. Also, the NLS 1979 was performed biennially, so the years chosen could not be consecutive. Table 4 gives the results for the women, and Table 5 gives them for the men. The results for the women show, first of all, that the fraction with children rose ever so slightly, consistent with the aggregate data. The real change is that the fraction with careers rose to around 35 to 40 percent, up by about 10 percentage points from the previous cohort. Thus, the fraction with both family and career (in boldface) increased to around 21 to 27 percent. Whether this is considered large or small can be judged relative to the men in the sample. The fraction of the men who claimed never to have fathered a child is a bit higher than the fraction of the women who stated they never had a birth--from 30
THE LONG ROAD TO THE FAST TRACK
33
TABLE 4 CAREER AND FAMILY FOR THE COLLEGE GRADUATE WOMEN OF COHORT 5 (IN PERCENTAGES) Career No Yes No Yes Kids Family No Kids
a
Kids
Family
No Kids
Hourly Wage: Three Years 47.8 17.9 23.7 10.6 Hourly Wage: Two Yearsc 44.9 15.4 26.7 13.0
Annual Income: Three Yearsb 50.4 14.9 21.4 13.3 Annual Income: Two Yearsd 48.9 13.6 22.9 14.6
SOURCE: National Longitudinal Survey of Youth (1979) (http://www.bls.gov/nls/home.htm). NOTE: The "hourly wage" columns use only observations that give annual earnings and usual hours. The income measure uses annual earnings and includes the self-employed. "Kids" are biological and exclude adoptive children and stepchildren. The inclusion of adoptive children does not change the results much. See text and Goldin (1997) for the definition of "career." The fraction with both family and career is given in boldface. a. Uses hourly wage measure for 1996, 1998, 2000 (N = 536). b. Uses income measure for 1996, 1998, 2000 (N = 611). c. Uses hourly wage measure for 1998, 2000 (N = 570). d. Uses income measure for 1998, 2000 (N = 611).
to 34 percent. But the fraction of men that had a career was much higher than that of women, ranging from 61 to 74 percent.9 In consequence, the fraction with career and family for the male sample ranged from 45 to 55 percent or about double that for the women. But even though men managed to achieve career and family about two times as often as women, this is probably the lowest that figure has been in U.S. history.
Why Did Change Occur?
What were the main factors that led college graduate women to take the long and winding road to the fast track? The transitions occurred, by and large, because constraints were loosened on women's ability to work in fulfilling careers, first after marriage and later after childbearing. Some of these changes were rooted in the labor market--such as the growth of a wide variety of white collar jobs (as was the case in the 1920s and again in the 1950s) and the greater ability of women in various periods to hold certain professional jobs. These changes were also rooted in schools--such as the increase in labormarket-relevant college majors for women beginning in the 1970s and the related increase in the enrollment of women in professional schools.
34
THE ANNALS OF THE AMERICAN ACADEMY
TABLE 5 CAREER AND FAMILY FOR THE COLLEGE GRADUATE MEN OF COHORT 5 (IN PERCENTAGES) Career No Yes No Yes Kids Family No Kids
a
Kids
Family
No Kids
Hourly Wage: Three Years 25.2 14.1 44.6 16.2 Hourly Wage: Two Yearsc 17.7 12.5 48.0 21.8
Annual Income: Three Yearsb 17.9 13.8 51.4 17.0 Annual Income: Two Yearsd 14.3 11.3 55.0 19.4
SOURCE: National Longitudinal Survey of Youth (1979) (http://www.bls.gov/nls/home.htm). NOTE: The "hourly wage" columns use only observations that give annual earnings and usual hours. The income measure uses annual earnings and includes the self-employed. "Kids" are biological and exclude adoptive children and stepchildren. The inclusion of adoptive children does not change the results much. See text and Goldin (1997) for the definition of "career." The definition for men is the same as for women. The fraction with both family and career is given in boldface. a. Uses hourly wage measure for 1996, 1998, 2000 (N = 505). b. Uses income measure for 1996, 1998, 2000 (N = 582). c. Uses hourly wage measure for 1998, 2000 (N = 527). d. Uses income measure for 1998, 2000 (N = 582).
But many of these changes occurred in the personal lives of college women. For example, cohort 2 was able to be married and have a job, at least for a short while. Cohort 4 achieved a later age at marriage and delay of childbirth because of better contraceptive methods, such as "the pill," that enabled them to control their fertility and thus plan for "career then family."
Conclusion
I have described the path to the fast track that college graduate women have taken starting with cohort 1, which graduated in the first two decades of the twentieth century and had "family or career," to the latest group, cohort 5, which has achieved a modicum of success in combining career and family. Each generation built on the successes and frustrations of the previous ones. Each stepped into a society and a labor market with loosened constraints and shifting barriers. The road was not only long, but it has also been winding. Some cohorts of college graduate women gained "family," whereas others gained "career." Only recently has a substantial group been able to grasp both at the same time.
THE LONG ROAD TO THE FAST TRACK
35
Notes
1. Throughout this article, having no children and the term "childless" will mean having no biological children. In the analysis of more recent cohorts, the fraction adopting can also be included, but it is small. 2. The data I use here are primarily for white college graduate women because the numbers of African American college graduate women in the early years of this study are too small. I hope that a future project will tap into the alumni records of historically black colleges to get a larger sample of black female college graduates. 3. Historians often call this group the "second cohort" of female college graduates to distinguish it from the "first cohort" that attended college just after the establishment of many of the fine women's colleges in the 1870s and 1880s. The fraction of women in cohort 1 that graduated from women's colleges was only about 20 percent, far lower than most presume. In 1897, it was 18 percent, and in 1924, it was 23 percent (these calculations use the data sources given in Goldin and Katz [1999, e.g., Table 1]). 4. The age at first marriage among college graduate women remained fairly constant from the cohort born in 1930 to that born in 1950 (see Goldin and Katz 2002). The age at first marriage continued to rise after the cohort born in 1957, although more slowly. 5. The more recent increase in college enrollments for women relative to men is reflected in the extrapolations in Figure 1 for college graduation rates achieved by thirty-five years old. 6. It should be noted that nonmarriage means that the woman was never married by the age given. 7. The 1940 federal population census was the first to include information on educational attainment. 8. The other two educational groups are no college and no college but a high school graduate. The latter group is given for the earlier years when college graduation was less common and high school graduation was not yet the norm. 9. The income levels for the man at the 25th percentile come from the Current Population Survey (http:// www.bls.census.gov/cps/cpsmain.htm) and are a bit lower than those from the NLS. Thus, the fraction of men with income exceeding this mark for two years can even exceed 75 percent. Since the same cutoff is used for both men and women, this should not affect the comparisons for cohort 5.
References
De Long, J. Bradford, Claudia Goldin, and Lawrence F. Katz. 2003. Sustaining U.S. economic growth. In Agenda for the nation, edited by H. Aaron, J. Lindsay, and P. Nivola. Washington, DC: Brookings Institution. Friedan, Betty. 1963. The feminine mystique. New York: Norton. Goldin, Claudia. 1992. The meaning of college in the lives of American women: The past hundred years. NBER Working Paper no. 4099, National Bureau of Economic Research, Cambridge, MA. . 1997. Career and family: College women look to the past. In Gender and family issues in the workplace, edited by Francine Blau and Ronald Ehrenberg. New York: Russell Sage Foundation. Goldin, Claudia, and Lawrence F. Katz. 1999. The shaping of higher education: The formative years in the United States, 1890 to 1940. Journal of Economic Perspectives 13 (Winter): 37-62. . 2002. The Power of the pill: Oral contraceptives and women's career and marriage decisions. Journal of Political Economy 110 (August): 730-70. McCarthy, Mary. 1954. The group. New York: Harcourt, Brace, Jovanovich.