American Families

by Suzanne M. Bianchi and Lynne M. Casper

Population Bulletin, Vol. 55, No. 4, December 2000

Understanding the ever-evolving American family requires taking the pulse of contemporary family life from time to time. This Population Bulletin analyzes the American family in the latter half of the 20th century to better understand what changes in the family portend for the first half of the 21st century.

Introduction

If there is one "mantra" about family life in the last half century, it is that the family has undergone tremendous change. No other institution elicits as contentious debate as the American family. Many argue that family life has been seriously degraded by the movement away from marriage and traditional gender roles. Others view family life as amazingly diverse, resilient, and adaptive to new circumstances.¹

Any assessment of the general "health" of family life in the United States and the well-being of family members, especially children, requires a look at what we know about demographic and socioeconomic trends that affect families. The latter half of the 20th century was characterized by tumultuous change in the economy, in civil rights, and in sexual freedom, and by dramatic improvements in health and longevity. Marriage and family life felt the reverberations of these societal changes. At the beginning of the 21st century, as we reassess where we have come from and where we are, one thing stands out. Our rhetoric about the dramatically changing family may be a step behind the reality. Recent trends suggest a quieting of changes in the family, or at least of the pace of change. There was little change in the proportion of two-parent or single-mother families during the 1990s. The living arrangements of children stabilized, as did the living arrangements of young adults and the elderly. The divorce rate had been in decline for more than two decades. The rapid growth in cohabitation among unmarried adults has also slowed.

Yet family life is still evolving. Age at first marriage rose as more young adults postponed marriage and children to complete college and settle into a labor market increasingly inhospitable to poorly educated workers. Accompanying this delay in marriage was the continued increase in births to unmarried women, though here, too, the pace of change slowed in the 1990s.²

Within marriage or marriage-like relationships, the appropriate roles for each partner are shifting as American society accepts and values more equal roles for men and women. The widening role of fathers has become a major agent of change in the family. There are an increasing number of father-only families, a shift toward shared custody of children by fathers and mothers after divorce, and increased father involvement with children in two-parent families.

Whether the slowing, and in some cases cessation, of change in family living arrangements is a temporary lull or part of a new, more sustained equilibrium will only be revealed in the first decades of the 21st century. New norms may be emerging about the desirability of marriage, the optimal timing of children, and the involvement of fathers in childrearing and of mothers in breadwinning. Understanding the ever-evolving American family requires taking the pulse on contemporary family life from time to time. This Population Bulletin describes the American family in the latter half of the 20th century to better understand what changes in the family portend for the first half of the 21th century.

A Changing Economy and Society

• Changing Family Norms
• An Aging Society
• Box 1: The 'Ups and Downs' of U.S. Fertility

Consider the life of a young woman reaching adulthood in the 1950s or early 1960s. Such a woman was likely to marry straight out of high school or to take a clerical or retail sales job until she married. She would have moved out of her parents' home only after she married, to form a new household with her husband. This young woman was likely to marry by age 20 and begin a family soon after. If she was working when she became pregnant, she would probably have quit her job and stayed home to care for her children while her husband had a steady job that paid enough to support the entire family.

Fast forward to the last few years of the 20th century. A young woman reaching adulthood in the late 1990s is not likely to marry before her 25th birthday. She will probably attend college and is likely to live by herself, with a boyfriend, or with roommates before marrying. She may move in and out of her parents' house several times before she gets married. Like her counterpart reaching adulthood in the 1950s, she is likely to marry and have at least one child, but the sequence of those events may well be reversed. She probably will not drop out of the labor force after she has children, although she may curtail the number of hours she is employed to balance work and family. She is also much more likely to divorce and possibly even to remarry compared with a young woman in the 1950s or 1960s.

Many of the changes in when women (and men) marry, have children, and enter the labor force reflect changed economic circumstances since the 1950s. After World War II, the United States enjoyed an economic boom characterized by rapid economic growth, full employment, rising productivity, higher wages, low inflation, and increasing earnings. A man with a high school education in the 1950s and 1960s could secure a job that paid enough to allow him to purchase a house, support a family, and join the swelling ranks of the middle class.

The economic realities of the 1970s and 1980s were quite different. The two decades following the oil crisis in 1973 were decades of economic change and uncertainty marked by a shift away from manufacturing and toward services, stagnating or declining wages (especially for less-educated workers), high inflation, and a slowdown in productivity growth. The 1990s were just as remarkable for the turnaround: sustained prosperity, low unemployment, albeit with increased inequality in wages, but with economic growth that seems to have reached many in the poorest segments of society.³

When the economy is on such a roller coaster, family life often takes a similar ride. Marriage was early and nearly universal in the decades after World War II; mothers remained in the home to rear children, and the baby-boom generation was born and nurtured. When baby boomers hit working age in the 1970s, the economy was not as hospitable as it had been for their parents. They postponed entry into marriage, delayed having children, and found it difficult to establish themselves in the labor market.

Many of the baby boomers' own children began reaching labor force age in the 1990s, when individuals' economic fortunes were increasingly dependent on their educational attainment. Those who attended college were much more likely to become self-sufficient and to live independently from their parents. High school graduates who did not go to college discovered that jobs with high pay and benefits were in relatively short supply. A high school graduate lucky enough to land such a job earned about 25 percent less than a comparable job would have paid 20 years earlier.⁴ The increasing benefits of a college education probably encouraged more young men and women to delay marriage and attend college.

Both men and women are remaining single longer, and are more likely to leave home to pursue a college education, to live with a partner, and to launch a career before taking on the responsibility of a family of their own. After a period of "no family" living, these young adults will increasingly form "new families."⁵ Many of these new families have increasingly egalitarian roles for men and women. The traditional, gender-based organization of home life (in which mothers have primary responsibility for care of the home and children, and fathers provide financial support) has not disappeared, but young women today can expect to be employed while raising children, and young men will likely be called upon to share in childrearing and household tasks.

Changing Family Norms

In 1950, there was one dominant and socially acceptable way for adults to live their lives. Those who deviated could expect to be censured and stigmatized. The idealized family was composed of a homemaker-wife, a breadwinner-father, and two or more children. Americans shared a common image of what a family should look like and how mothers, fathers, and children should behave. These shared values reinforced the importance of the family and the institution of marriage.⁶ This vision of family life showed amazing staying power, even as its economic underpinnings were eroding.

For this 1950s-style family to exist, Americans had to support distinct gender roles and the economy had to be vibrant enough for a man to financially support a family on his own. Government policies and business practices perpetuated this family type by reserving the best jobs for men and discriminating against working women when they married or had a baby. After 1960, with the civil rights movement and an energetic women's liberation movement, women and minorities gained legal protections in the workplace and discriminatory practices began to recede.

A transformation in attitudes toward family behaviors also occurred. People became more accepting of divorce, cohabitation, and sex outside marriage; less sure about the universality and permanence of marriage; and more tolerant of blurred gender roles and of mother's working outside the home.⁷ Society became more open-minded about a variety of living arrangements, family configurations, and lifestyles.

While the transformation of many of these attitudes occurred throughout the 20th century, the pace of change accelerated in the 1960s and 1970s. These years brought many political, social, and medical developments, including the highly publicized, although unsuccessful, attempt to pass the Equal Rights Amendment (ERA);⁸ the development of new, effective contraception; the legalization of abortion; and the dawn of the sexual revolution and an era of "free love."

A new ideology was emerging during these years that stressed personal freedom, self-fulfillment, and individual choice in living arrangements and family commitments. People began to expect more out of marriage and to leave bad marriages if their expectations were not fulfilled. These changes in norms and expectations about marriage may have followed rather than preceded increases in divorce and delays in marriage; however, such cultural changes have important feedback effects, leading to later marriage and more divorce.

An Aging Society

For Americans born in 1900, the average life expectancy was just below 50 years. But the early decades of the 20th century brought such tremendous advances in the control of communicable diseases of childhood that life expectancy at birth increased to 70 years by 1960. Rapid declines in mortality from heart disease — the leading cause of death — significantly lengthened life expectancy for those ages 65 or older after 1960.⁹ By 1998, life expectancy at birth was nearly 77 years. An American woman who reached age 65 in 1998 could expect to live an additional 19 years, on average, and a 65-year-old man would live another 16 years.¹⁰

Increased life expectancy translates into extended years spent in family relationships. A couple who marries in their 20s could easily spend the next 50 years together, assuming they remain married. Couples in the past were much more likely to experience the death of one spouse earlier in their older years. The increased life expectancy, in fact, may be implicated in the increased incidence of divorce. All family members today have more years together as adults now than they did during the early 1900s. Mothers and daughters spend nearly two-thirds of their years together as adults.¹¹ Siblings will spend a greater proportion of their relationship as adults now than when life expectancy was shorter.

Longer lives (along with lower birth rates) also mean people spend a smaller portion of their lives parenting young children. More parents live long enough to be part of their grandchildren's and even great grandchildren's lives. And, adults often are faced with caring for extremely elderly parents about the time they are beginning to experience their own health difficulties.

In sum, an aging society alters the context for family relationships. Longer life expectancy — combined with shifts in the economy and changing norms, values, and laws — influences individuals' life course trajectories. All these changes in individual lives and family relationships are transforming U.S. households and families.

Family Structure and Living Arrangements

• Figure 1: Trends in U.S. Households, 1960 to 2000
• Figure 2: Rates of First Marriage, Remarriage, and Divorce, 1960 to 1997
• Elderly Americans
• Figure 3: Living Arrangements of Elderly Men and Women, 1960, 1980, and 2000
• Young Adults
• Figure 4: Living Arrangements of Young Men and Women, 1970 to 2000
• Unmarried Couples
• Table 1: Unmarried Couples by Relationship Type in 1987 – 1988, and After Five to Seven Years
• Table 2: Selected Characteristics of Unmarried and Married Couples, 1998
• Box 2: Gay and Lesbian Families
• Box 3: Measuring Cohabitation

The term "family" carries rich social and cultural meanings, and it has deep personal significance for most people — but for statistical purposes a family is defined as two or more people living together who are related by blood, marriage, or adoption. Most households — which are defined by the U.S. Census Bureau as one or more people who occupy a house, apartment, or other residential unit (but not "group quarters" such as dormitories) — are maintained by families. But the social and economic transformation of the family in recent decades means that the family share of U.S. households has been declining, as shown in Figure 1. In 1960, 85 percent of households were family households; by 2000, just 69 percent were family households. At the same time, nonfamily households, which consist primarily of people who live alone or who share a residence with roommates or with a partner, have been on the rise. The fastest growth was among persons living alone. The proportion of households with just one person doubled from 13 percent to 26 percent between 1960 and 2000.

Figure 1
Trends in U.S. Households, 1960 to 2000

Source: U.S. Census Bureau, March supplements of the Current Population Surveys, 1960 to 2000.

Most of the decline in family households reflects the decrease in the share of married-couple households with children. Declines in fertility within marriage between 1960 and 1975 (see Box 1), later marriage, and frequent divorce help explain the shrinking proportion of households consisting of married couples with children. The divorce rate rose sharply between 1960 and 1980, and then eased, while the rate of first marriages declined steadily after 1970 (see Figure 2). Two-parent family households with children dropped from 44 percent to 24 percent of all households between 1960 and 2000.

Figure 2
Rates of First Marriage, Remarriage, and Divorce, 1960 to 1997

Note: First marriage rate is the number of first marriages per 1,000 women ages 15 and older (ages 14 and older in 1960 and 1965); Remarriage rate is remarriages per 1,000 divorced and widowed women ages 15 and older (ages 14 and older in 1960 and 1965); Divorce rate is the number of divorces per 1,000 married women ages 15 and older. Remarriage rates are not available after 1990.

Source: National Center for Health Statistics, National Vital Statistics Reports.

Change in household composition began slowly in the 1960s, just as society was embarking on some of the most radical social changes in our nation's history (see Box 2), and the leading edge of the huge baby-boom generation was reaching adulthood. The steepest decline in the share of family households was in the 1970s when the first baby boomers entered their 20s. By the 1980s, change was still occurring, but at a much less rapid pace. By the mid-1990s, household composition reached relative equilibrium, where it has been since.

Elderly Americans

Improvements in the health and financial status of older Americans helped generate a revolution in lifestyles and living arrangements among the elderly. Older Americans now are more likely to spend their later years with their spouse or living alone than with adult children. The options and choices are different for elderly women and men, however, in large part because women live longer than men yet have fewer financial resources and smaller pensions.

At the beginning of the 20th century, more than 70 percent of people ages 65 or older resided with kin.¹² By 1980, only 23 percent of elderly lived with relatives, and by 1998, the percentage had slipped to 20 percent.¹³ Meanwhile, living alone increased dramatically among the elderly in the latter half of the 20th century. Just 15 percent of widows age 65 or older lived alone in 1900, for example, while 24 percent lived alone in 1950, and more than 70 percent lived alone in 1998.¹⁴

The improvements in life expectancy among elderly Americans have meant more years in retirement and a greater likelihood of spending those years with a spouse. Accordingly, married couples without children under age 18 have made up an increasing share of American households. Both men and women ages 65 and older were more likely to be living with a spouse in 2000 than in 1960 (see Figure 3). After 1980, elderly men experienced a slight decline in the proportion living with a spouse, while elderly women experienced additional increases. Some demographers tie the decrease in the proportion of elderly men living with a spouse to the increase in divorce during the 1960s and 1970s: Divorced men are entering their older ages without a spouse. For women, the increase in divorce was offset by a countervailing decrease in the proportion of women entering their older years as a widow.¹⁵

Figure 3
Living Arrangements of Elderly Men and Women, 1960, 1980, and 2000

Note: Other living arrangements include people living with unrelated roommates or other relatives, and unmarried couples. People living in a nursing home or other institution are excluded. Source: U.S. Census Bureau, March supplements of the 1980 and 2000 Current Population Surveys, and the 1960 Census.

A woman is likely to spend more years living alone after a spouse dies than will a man because life expectancy is about four years longer for an elderly women than an elderly man, on average, and because women usually marry men older than themselves. Men age 65 or older were nearly twice as likely as women to be living with their spouse in 2000 (72 percent vs. 41 percent respectively). In sharp contrast, women were more than twice as likely as men to be living alone (40 percent vs. 17 percent).

Women were also almost twice as likely as men to be living with others (19 percent vs. 11 percent), in part because they tend to live longer and reach advanced ages when they are most likely to need the physical care and the financial help others can provide. Men generally receive companionship and care from their wives in the latter stages of life, while women are more likely to live alone, perhaps with assistance from grown children, to live with other family members, or to enter a nursing home.¹⁶

To explain trends in living arrangements among the elderly, researchers have focused on a variety of constraints and preferences that shape people's living arrangement decisions. These constraints and preferences fall under three general categories: availability and accessibility of relatives; feasibility; and preferences.¹⁷

The availability and accessibility of relatives with whom an elderly person might live are generally governed by the number and gender of their children. The greater the number of children, the greater the chances there will be a son or daughter who can take in an elderly parent. Daughters are more likely than sons to provide housing and care for an elderly parent, presumably as an extension of the traditional female caretaker role. Geographical distance from children is also a key factor; having children who live nearby promotes coresidence when living independently is no longer feasible for the elderly person.¹⁸

The feasibility of living alone, with relatives, or in a nursing home is tied to economic resources and health status. Older Americans with higher income and better health are more likely to live independently.¹⁹ Growth in Social Security benefits accounts for half of the increase in independent living among the elderly since 1940.²⁰ Similarly, elderly Americans in financial need are more likely to live with relatives.²¹

Finally, social norms and personal preferences determine the choice of living arrangements for the elderly.²² Many elderly individuals are willing to pay a substantial part of their incomes to maintain their own residence, which suggests strong personal preferences for privacy and independence. Social norms involving family obligations and ties also have an effect on residence patterns and may be especially important when examining racial and ethnic differences in the living arrangements of the elderly.

Young Adults

In 1890, one-half of American women had married by age 22 and one-half of American men had married by age 26. The ages of entry into marriage dipped to an all-time low during the post-World War II baby-boom years, when median age at first marriage reached 20 years for women and nearly 23 years for men in 1956. Age at first marriage then began to rise and reached 25 years for women and 27 years for men by the end of the 1990s.²³ In 1960, it was unusual for a woman to reach age 25 without marrying: Only 10 percent of women ages 25 to 29 had never married. In 2000, a women in her late 20s who had never been married was not typical, but she had many more friends like herself because two-fifths of women ages 25 to 29 had not been married. And, in 2000, the majority (52 percent) of men were still unmarried at these ages.²⁴

This delay in marriage has shifted the family behaviors in young adulthood in three important ways. First, it lengthens the period of time young adults have to settle into adult roles and stable relationships. Consequently, later marriage coincides with a greater diversity and fluidity in living arrangements in young adulthood. Secondly, delaying marriage has accompanied an increased likelihood of entering a cohabiting union before marriage. Third, the trend to later marriage affects fertility in two important ways. It tends to delay entry into parenthood and, at the same time, increases the chances that a birth (sometimes planned but more often unintended) may happen before marriage.

The young adult years have been referred to as "demographically dense" because these years involve many interrelated life-altering transitions.²⁵ Between the ages of 18 and 30, young adults usually finish their formal schooling, leave home, develop careers, marry, and begin families; but these events do not always occur in this order. Delayed marriage extends the period young adults can experiment with alternative living arrangements before they adopt family roles. Young adults may experience any number of independent living arrangements before they marry, as they change jobs, pursue education, and move into and out of intimate relationships. They may also return home during school breaks, if money becomes tight, or if a relationship breaks up.

Many demographic, social, and economic factors influence young adults' decisions about where and with whom to live.²⁶ Family and work transitions are influenced greatly by fluctuations in the economy as well as by changing ideas about appropriate family life and roles for men and women. Since the 1980s, the transition to adulthood has been hampered by recurring recessions, tight job markets, slow wage growth, and soaring housing costs, in addition to the confusion over roles and behavior sparked by the gender revolution.²⁷ Even though young adults today may prefer to live independently, they may not be able to afford to do so. Many entry-level jobs today offer low wages yet housing costs have soared, which has put independent living out of reach for many young adults. Higher education, increasingly necessary in today's labor market, is expensive, and living at home may be a way for families to curb college expenses. Even when young adults attend school away from home, they still frequently depend on their parents for financial help and may return home after graduation if they can't find a suitable job.

Annual data from the Current Population Survey (CPS) chronicle how the living arrangements of young American adults have changed over the decades. Married living declined dramatically between 1970 and 2000, among both young men and young women (see Figure 4). Thirty-one percent of men ages 18 to 24 lived with their spouses in 1970, for example, while only 9 percent lived with a spouse in 2000. A similar drop occurred for women — from 45 percent in 1970 to 16 percent in 2000.

Figure 4
Living Arrangements of Young Men and Women, 1970 to 2000

Note: Other living arrangements include unmarried couples and people living with unrelated roommates or other relatives. People living in military barracks or institutions are excluded.

Source: U.S. Census Bureau, March Supplements of the Current Population Surveys of 1970, 1980, 1990, and 2000.

As a declining share of young adults chose married life, a greater share lived with parents or on their own. The percentage of young men living in their parents' homes was 57 percent in 2000, about the same as in 1970, while the percentage increased for young women from 39 percent to 47 percent.The differences between young men's and women's living arrangements have declined. Traditionally, young men were much more likely to reside with their parents than were young women. Young women were much more likely to live with a spouse than young men, in part reflecting the older age of marriage for men. While these differences still persist, they have diminished somewhat over time.

Americans were leaving their parents' homes at increasingly younger ages throughout most of the 20th century. In the 1980s, however, this trend had reversed for both young men and young women. Not only are recent cohorts leaving home later, they are also more likely to return home — "backwards" transition out of the adult role and back into the role of a dependent. By the 1980s, about 40 percent of young adults who left home eventually returned for a time, a marked increase from less than 25 percent of those reaching adulthood before World War II.²⁸

Young adults who leave home to attend school, join the military, or take a job have always had, and continue to have, high rates of returning to the nest. The "return rate" is nearly as high for those who leave home to live with a partner or to form another type of nonmarital family. Those who leave home to get married have had the lowest likelihood of returning home, although returns to the nest have increased over time even in this group. Historically, moving away for independence was associated with very low rates of returning home — less than 20 percent prior to World War II. This pattern has changed over time so that currently about 40 percent of those who first leave for independence return to the nest.

American parents routinely take in their children after they return from the military or school, or when they are between jobs. In the past, however, many American parents apparently were reluctant to take children in if they had left home simply to gain "independence." This is not true today. Demographers Frances Goldscheider and Calvin Goldscheider argue that in the past, leaving home for simple independence was probably the result of friction within the family, whereas today, leaving and returning home seems to be part of a successful transition to adulthood. In the past, a young adult may have been reluctant to move back in with parents because a return home implied failure; there is less stigma attached to returning home these days.²⁹ Changes in the economy have also contributed to this trend. It may be more difficult to sustain an independent residence today than in the past.

Unmarried Couples

One of the most significant changes in the second half of the 20th century was the increase in men and women living together without marrying. The rise of cohabitation outside marriage appeared to counterbalance the delay of marriage among young adults and the general increase in divorce.

Most adults in the United States eventually marry. In 2000, 91 percent of women ages 45 to 54 had been married at least once.³⁰ An estimated 88 percent of U.S. women born in the 1960s will eventually marry.³¹ But the meaning and permanence of marriage may be changing. Marriage used to be the primary demographic event that marked the formation of new households, the beginning of sexual relations, and the birth of a child. Marriage also implied that an individual had one sexual partner, and it theoretically identified the two individuals who would parent any child born of the union. The increasing social acceptance of cohabitation outside marriage has meant that these linkages could no longer be assumed. Unmarried couples began to set up households that might include the couple's children as well as children from previous marriages or other relationships. Similarly, what it meant to be single was no longer always clear, as the personal lives of unmarried couples began to resemble those of their married counterparts.

Cohabiting and marital relationships have much in common — shared living space; emotional, psychological, and sexual intimacy; and some degree of economic interdependence. But the two relationships differ in important ways. Marriage is a relationship between two people of the opposite sex that adheres to legal, moral, and social rules. It is a social institution that rests upon common values and shared expectations for appropriate behavior. Society upholds and enforces appropriate marital behavior both formally and informally. In contrast, there is no widely recognized social blueprint to guide appropriate behavior between men and women who live together, or for the behavior of their friends, family, and the other institutions with whom they interact.

Because there is no legal bond, and because fear of social disapproval might discourage people from stating publicly that they live together, measuring trends in cohabitation has been tricky (see Box 3). There is little disagreement, however, that cohabitation has increased in U.S. society. Unmarried-couple households made up less than 1 percent of U.S. households in 1960 and 1970.³² This share rose to 2.2 percent by 1980, to 3.6 percent in 1990, and to nearly 5 percent by 1998. Unmarried-couple households also are increasingly likely to include children. In 1978, 29 percent of unmarried-couple households included children under age 18; by 1998, 43 percent included children.

The number of unmarried-couple households surged from 1.3 million in 1978 to 3.0 million in 1988, and to 4.9 million in 1998. These figures suggest that the growth in cohabitation from 1978 to 1998 could account for 38 percent of the decline in marriage over the period, assuming that all the cohabitors would have married.

Although a relatively small percentage of U.S. households consists of an unmarried couple — one in 20 households in 1998 — many Americans have lived with a partner outside marriage at some point, which means that cohabitation is a large and growing component of U.S. family life. The 1987–1988 National Survey of Families and Households found that 25 percent of all adults and 45 percent of adults in their early 30s had lived with a partner outside marriage. More than one-half of the couples who married in the mid-1990s had lived together before marriage, up slightly from 49 percent in 1985-1986, and a big jump from just 8 percent of first marriages in the late 1960s.³³

Why has cohabitation increased so much since the 1970s? Researchers have offered several explanations, including increased uncertainty about the stability of marriage, the erosion of norms against cohabitation and sexual relations outside of marriage, the wider availability of reliable birth control, and increased individualism and secularization. Youth reaching adulthood in the past two decades are much more likely to have witnessed divorce than any generation before them. Some have argued that cohabitation allows a couple to experience the benefits of an intimate relationship without committing to marriage. If a cohabiting relationship isn't successful, one can simply move out; if a marriage isn't successful, one suffers through a sometimes lengthy and messy divorce.

The increase in unmarried-couple households is slowing from the frantic pace of the 1970s and 1980s. The CPS estimates show the number of households with unmarried couples increasing 67 percent in the five years between 1978 and 1983, but just 23 percent in the five years between 1993 and 1998.

The pace of growth varied among the three largest racial and ethnic groups. In 1978, single non-Hispanic white women were least likely to cohabit outside marriage, while single Hispanic women were most likely to cohabit. By 1998, 10 percent of single non-Hispanic white women lived with a male partner, compared with 9 percent of single Hispanic and 7 percent of single non-Hispanic black women.

Cohabitation serves different purposes for different couples. It may be a precursor to marriage, a trial marriage, a substitute for marriage, or simply a serious boyfriend-girlfriend relationship.³⁴ In a 1987–1988 survey, 46 percent of cohabitors characterized their living arrangement as a precursor to marriage (see Table 1). Another 15 percent of these relationships were classified as a trial marriage and 10 percent as a substitute for marriage. Nearly 30 percent of the relationships were characterized as coresidential dating.

Some researchers believe that cohabitation closely resembles marriage. If so, family life as we know it is not likely to be altered much as a consequence of cohabitation because these cohabitors will either eventually marry (precursor) or are already in a relationship which functions like a marriage (substitute).

Table 1
Unmarried Couples by Relationship Type in 1987–1988, and After Five to Seven Years

Note: Couples were interviewed between 1987 and 1988 and again from 1992 to 1994.
1 Couple was still cohabiting at the time of the second survey.
2 Got married some time between the two surveys (may or may not be currently married).
3 No longer cohabiting.

Source: L.M. Casper and L.C. Sayer, "Cohabitation Transitions: Different Attitudes and Purposes, Different Paths." (Paper presented at the annual meeting of the Population Association of America, Los Angeles, March 2000.)

Other researchers maintain that cohabitation is more like being single. Cohabitation is seen as an enjoyable relationship of convenience that provides intimacy without the long-term commitment of marriage. This interpretation worries many people because it suggests that the increase in cohabitation signals a retreat from marriage. It allows for an intimate, but temporary, relationship without commitment or responsibility.

Still others argue that living together before marriage is somewhere between marriage and singlehood and that cohabitation provides a couple the opportunity to assess their compatibility before getting married. In this trial period, incompatible mates can easily end their relationship and presumably escape an unsuitable marriage. Thus, cohabitation might strengthen marriage and family life because some unsuited couples are weeded out before they marry.

About 40 percent of all unmarried couples surveyed in 1987-1988 were married within five to seven years. Their reasons for forming the relationship were strongly related to its eventual outcome.³⁵ Those with the strongest commitment to one another and to marriage were most likely to get married. More than one-half of couples who characterized their living together as a precursor to marriage did marry within five to seven years, compared with 33 percent of "dating" couples with no long-term expectations about their partner, their relationship, or marriage. About one-quarter of unmarried couples in "trial marriage" or "substitute marriage" married within the seven years.

How do unmarried couples compare with married couples? Although many cohabiting couples eventually marry, men and women who choose to live together outside marriage differ from married couples in some very interesting ways. In general, cohabiting couples tend to be more egalitarian and less traditional than married couples. Compared with a woman who is part of a married couple, for example, a woman in a cohabiting relationship is more likely to be older than the man, to be of a different race or ethnic group than the man, to contribute a greater percentage to the couple's income, and to have more education than the man.

American women tend to marry men a few years older than themselves and relatively few marry a much younger man. Yet nearly one-fourth of women in cohabiting couples were two or more years older than their male partner, compared with one-eighth of women in married couples (see Table 2). Cohabiting couples were also more than twice as likely to be of different races than married couples — 13 percent compared with 5 percent. About one-half of the interracial unmarried couples consisted of a white woman and a man who was African American, Hispanic, or of some other racial or ethnic group.

Table 2
Selected Characteristics of Unmarried and Married Couples, 1998

1 Unmarried couples who maintain a household together.
2 Calculated for couples in which both partners were employed.

Source: Authors’ tabulations of the March 1998 Current Population Survey.

In many married couples, the husband has a higher educational level than the wife. Women had a higher education level in 21 percent of cohabiting couples, compared with 16 percent of married couples. In almost four of five cohabiting couples, both partners were working in 1997, compared with only three in five married couples. Women in cohabiting couples contributed 41 percent of the couple's annual income in 1997, while married women contributed 37 percent. In addition, men earned at least $30,000 more than their partners in just 11 percent of cohabiting couples compared with 27 percent of married couples.

Parenting

• Table 3: Living Arrangments of Unmarried Fathers and Mothers With Children Under Age 18: 1978, 1988, and 1998
• Single Mothers
• Table 4: Trends in Single-Parent and Two-Parent Families With Children Under Age 18, 1950 to 2000
• Figure 5: Birth Rates for Married and Unmarried Women, and Percent of Births to Unmarried Women, 1960 to 1999
• Figure 6: Percentage of Births to Unmarried Women, 1998
• Fathering
• Figure 7: Percentage of Adulthood Spent as Parents With Children Under Age 18
• Father-Only Families
• Grandparenting
• Well-Being of Single-Parent Families
• Table 5: Socioeconomic Characteristics of Children in One-Parent and Two-Parent Families, 1998

Even with the rise in divorce and cohabitation, postponement of marriage, and decline in fertility, most Americans have children and most children live with two parents. In 2000, 73 percent of families with children were two-parent families. But the changes in marriage, cohabitation, and nonmarital childbearing over the past few decades have had a profound effect on American families with children and are changing our images of parenthood.

Changes in marriage and cohabitation tend to blur the distinction between one-parent and two-parent families. The increasing acceptance of cohabitation as a substitute for marriage, for example, may reduce the chance that a premarital pregnancy will lead to marriage before the birth.³⁶ A greater share of children today are born to a mother who is not currently married than in previous decades, but some of those children are born to cohabiting parents and begin life in a household that includes both their parents. Cohabitation has also become a widely accepted pathway into remarriage. It may effectively bring "stepfathers" into the picture before there is a formal remarriage.

Demographers Larry Bumpass and R. Kelly Raley show that the increase in cohabitation among Americans may be reducing the time children spend in a single-parent household. Bumpass and Raley found that the number of years a white mother spent as a single parent declined by one-fourth when they took into account the time she and her children shared a home with a partner.³⁷ Black women with children spent half as many years as a single parent after adjusting for the years they lived with an unmarried partner.

Many single-father families may also effectively be two-parent families because the father is living with his children and another woman. Demographers Steven Garasky and Daniel Meyer used census data from 1960 through 1990 to track the increase in the percentage of families that are father-only families. When they ignored the increase in cohabitation, Garasky and Meyer estimated that father-only families rose from 1.5 percent to 5.0 percent of families with children between 1960 and 1990. When they remove fathers who are likely to be cohabiting, the 1990 figure falls to 3.2 percent of all families with children.³⁸

The percentage of unmarried mothers who were cohabiting grew from 5 percent to 13 percent between 1978 and 1998, according to the CPS (see Table 3). Cohabitation increased for unmarried mothers in all race and ethnic groups but especially among whites. Cohabiting couples account for up to 16 percent of the white mothers classified as unmarried mothers in 1998, compared with 8 percent of black and 10 percent of Hispanic mothers.

Unmarried fathers living with children are much more likely than unmarried mothers to be living with a partner: 33 percent of the 2.1 million "single" fathers lived with a partner in 1998, more than twice the percentage for single mothers. About 1.4 million American men were raising their children on their own, without a wife or partner, in 1998.

Table 3
Living Arrangments of Unmarried Fathers and Mothers With Children Under Age 18: 1978, 1988, and 1998

1 Includes a partner of the opposite sex who is not married to the parent.
2 Other adults include nonrelatives other than a cohabiting partner and relatives other than parents.

Source: Authors’ tabulations of the March supplements of the Current Population Surveys of 1978, 1988, and 1998.

Single Mothers

Single mothers with children at home face a multitude of challenges: They usually are the primary breadwinner, disciplinarian, playmate, and caregiver for their children. They must manage the financial and practical aspects of a household, and plan for their children's and their own futures. Most mothers cope remarkably well, and many benefit from financial support and help from relatives and from their children's fathers.

Most single mothers are not poor, but they tend to be younger, earn lower incomes, and be less educated than married mothers. Women earn less than men, on average, and because single mothers are younger and less educated than other women, they are often at the lower end of the income curve. Single mothers often must curtail their work hours to care for their children. Many do not receive regular child support from their children's fathers.

Because single-mother families have lower incomes than other families, they have been linked with poverty and welfare receipt in the public's eye. Single mothers with children in poverty are particularly affected by major welfare reform legislation, such as the Personal Responsibility and Work Opportunity Reconciliation Act (PRWORA), signed into law by President Bill Clinton in August 1996. President Clinton claimed that the law would "reform welfare as we know it," and the changes embodied in PRWORA — time limits on welfare eligibility and mandatory job training requirements, for example — seemed far-reaching. Some argued that this legislation would end crucial support for poor mothers and their children; several high-level government officials resigned because of the law.³⁹ Others heralded PRWORA as the first step toward helping poor women gain control of their lives and making fathers take responsibility for their children. Many states had already begun to experiment with similar reforms.⁴⁰

Why has welfare receipt become such a contentious issue? In part, it was because many people were alarmed by the dramatic increase in the second half of the 20th century in the numbers of single mothers, in particular the increase in divorced and unmarried mothers. When legislation to protect poor women and children was enacted with the Social Security Act of 1935, most single mothers, poor as well as nonpoor, were widows. That changed dramatically in the 1960s and 1970s as the divorce rate soared. After 1980, the delay in marriage and increase in the proportion of births that occur to unmarried mothers meant that never-married mothers accounted for an increasingly significant component of the growth in single-parent families.

Why have mother-child families increased in number and as a percentage of American families? Explanations tend to focus on one of two trends. First is women's increased financial independence, either through their own wages as more women entered the labor force and women's incomes rose relative to those of men, or because of expanded welfare benefits for single mothers. Women today are less dependent on a man's income to support themselves and their children, and many can afford to live independently rather than stay in an unsatisfactory relationship.

Second, the job market for men has tightened, especially for less-educated men. As the U.S. economy experienced a restructuring in the 1970s and 1980s, the demand for professionals, managers, and other white-collar workers expanded while the jobs available for semiskilled and unskilled workers declined. The wages for men in lower-skilled jobs have declined in real terms over the past two decades. Men still earn more than women, on average, but the income gap has narrowed as women's earnings increased and men's earnings remained flat or declined. Many men without a college degree cannot earn enough to support a family of four.

Some analysts tie men's weakened economic position to an erosion of marriage and of fathers' involvement with their children — which promotes single-mother families. This line of reasoning assumes that men are less attractive marriage partners if they earn low salaries or do not have steady employment. Men's financial problems would affect their personal relationships by eroding men's self-esteem. Fathers may not feel they are a necessary part of their children's lives, which can make them more likely to live separately as well.

Other analysts interpret the general movement away from marriage as part of a broader cultural shift toward more individualism that is affecting Americans at all income levels.

Who is a single mother and for how long turns out to be a more difficult question to answer than it would first appear. Over time, it is easiest to calculate the number of single mothers who maintain their own residence. Between 1950 and 1998, the number of such single-mother families increased from 1.3 million to 7.7 million.⁴¹ The most dramatic increase was during the 1970s, when the number of single-mother families was increasing at 8 percent per year. The average annual rate of increase slowed considerably during the 1980s, and was near zero after 1994. By 2000, single-mother families accounted for 22 percent of all families with children, up from 6 percent in 1950 (see Table 4).

Table 4
Trends in Single-Parent and Two-Parent Families With Children Under Age 18, 1950 to 2000

Note: Includes only families who maintain their own households.

Source: U.S. Census Bureau, "Families, by Presence of Own Children Under 18:1950 to Present." Accessed online at www.census.gov/population/socdemo/hh-fam/tabFM-1.txt on Oct. 13, 2000; and PRB analysis of the March 2000 Current Population Survey.

In the early years of the 20th century, it was not uncommon for children to live with only one parent because of high mortality.⁴² As falling death rates reduced the number of widowed single parents, there was a counterbalancing increase in single-parent families because of divorce. Still, at the time of the 1960 Census, almost one-third of single mothers living with children under age 18 were widows.⁴³ As divorce rates rose precipitously in the 1960s and 1970s, most single-parent families were created through divorce or separation. By the end of the 1970s, only 11 percent of single mothers were widowed and two-thirds were divorced or separated. During the past two decades, the path to single motherhood has increasingly bypassed marriage. In 1978, about one-fifth of single-mothers had never married, but had a child and was raising that child on their own. By 2000, two-fifths of single mothers had never married.

The remarkable increase in the number of single-mother households with women who have never married was driven by a dramatic shift to childbearing outside marriage. The number of births to unmarried women grew from less than 90,000 per year in 1940 to more than 1.3 million per year in 1999. Less than 4 percent of all births in 1940 were to unmarried mothers compared with 33 percent in 1999. The rate of nonmarital births — the number of births per 1,000 unmarried women — increased from 7.1 in 1940 to 43.9 in 1999. The nonmarital birth rate peaked in 1994 at 46.9 and leveled out in the latter 1990s (see Figure 5).

A similar plateau in the early 1970s proved to be temporary, so demographers cannot predict whether the stability of nonmarital birth rates in the late 1990s is a temporary lull or an end to one of the most pronounced trends in the latter half of the 20th century.⁴⁴

Figure 5
Birth Rates for Married and Unmarried Women, and Percent of Births to Unmarried Women, 1960 to 1999

Source: National Center for Health Statistics, National Vital Statistics Reports 48, no. 16 (Oct. 18, 2000): table 1.

Trends in nonmarital fertility are connected to broad trends in marriage and fertility. The delay in marriage, for example, can lead to an increase in the number of births outside marriage even if the birth rate for unmarried women remains the same. When women remain single longer, they spend more years at risk of becoming pregnant and having a child outside marriage. At the same time, married women are having fewer children, which means that children born to unmarried women make up a greater share of all births.

In the 1960s and 1970s, the delay in marriage and decline in fertility within marriage were the major factors contributing to the increase in the proportion of births outside marriage. During the past two decades, the increase in the birth rates of unmarried women has been a much more important factor and, indeed, has raised concerns about the move away from marriage and a breakdown of social sanctions against out-of-wedlock childbearing.⁴⁵

The proportion of births that occur outside marriage is as high or higher in some European countries than in the United States. But the factors prompting unmarried childbearing may be different in Europe.⁴⁶ In the United States, the tremendous variation in rates of unmarried childbearing among population groups suggests that there may be a constellation of factors that determine whether women have children when they are not married. More than two-thirds (69 percent) of the babies born to African American mothers in 1998 were born to unmarried mothers, as were more than one-half (59 percent) of babies of American Indian mothers (see Figure 6). The percentage is relatively low for Cuban Americans (25 percent), and for Asian Americans and Pacific Islanders (16 percent). The rates are extremely low for Chinese Americans (6 percent) and Japanese Americans (10 percent). On the other hand, rates are extremely high for Hawaiians and for some Hispanic ethnic groups: 60 percent of Puerto Rican births were to unmarried women.

Figure 6
Percentage of Births to Unmarried Women, 1998

Source: National Center for Health Statistics, National Vital Statistics Reports 48, no. 3 (March 2000): tables 13 and 14.

Women form single-mother households through divorce, separation, or by having a child while unmarried. These households dissolve when the children move out or the mother marries or remarries. How many women will become single mothers and how long will they spend as single mothers? Using demographic and microsimulation techniques, Robert Moffit and Michael Rendall have estimated how the likelihood of becoming a single mother has changed across birth cohorts.⁴⁷ For women born around 1935, about 35 percent became a single mother at some point before all their children reached age 18. This percentage is projected to increase to 53 percent among women born around 1970 — women who were in the midst of their childrearing years at the end of the 1990s.

Single mothers spend a total of about nine years raising children without a partner present, on average, but some mothers have several shorter periods of single parenting. The average length of time a women spends as a single mother at one stretch is about six years. These averages have not changed much over time, even though single parenting has become more common.

The age pattern of single parenting has changed over time. As the source of single motherhood has shifted from widowhood (a transition that occurs later in life) to nonmarital childbearing (which occurs relatively early in life for most women), women enter single motherhood at younger ages than in the past. The peak in single motherhood occurs around age 30 for recent cohorts whereas for the 1935 birth cohort, the peak would have come for women in their late 30s and 40s, most often after the death of their husbands.

There are large racial differences in rates of single parenting, but the trends are similar for black and white women: Increasing percentages of women become a single parent and increasing percentages experience more than one episode of single mothering. Yet, the total number of years spent as a single mother has remained fairly stable. Among blacks, however, the likelihood of becoming a single parent increased from 65 percent of black women born in 1935 to 80 percent of black women born in 1970. The comparable estimates for whites were 31 percent and 45 percent, respectively, for women in the 1935 and 1970 birth cohorts. One caution about these estimates is that they rely on data collected in the mid- to late 1980s. If growth in single parenting is slowing, the high estimates of single parenting for younger cohorts may be leveling out and could decline somewhat, given the decrease in divorce rates since 1980.⁴⁸

Fathering

A new view of fatherhood emerged out of the feminist movement of the late 1960s and early 1970s. The new ideal father was a co-parent who was responsible for and involved in all aspects of his children's care. The ideal has been widely accepted throughout U.S. society; people today, compared with those in earlier times, believe that fathers should be highly involved in caregiving.⁴⁹ Fathers do spend more time with their children and are doing more housework than in earlier decades. In 1998, married fathers in the United States reported spending an average of 4.0 hours per day with their children compared with 2.7 hours in 1965.⁵⁰ Parallel findings emerge from data collected on children and who they spend time with. Studies of fathers' time with their children in other industrialized countries, including Great Britain and Australia, also indicate that fathers are becoming more involved in parenting.⁵¹

At the same time, other trends increasingly remove fathers from their children's lives. When the mother and father are not married, for example, ties between fathers and their children often are tenuous. Family demographer Frank Furstenberg uses the label "good dads, bad dads" to describe the parallel trends of increased commitment to children and childrearing on the part of some fathers at the same time that there seems to be less connection to and responsibility for children on the part of other fathers.⁵²

It is difficult to measure "father involvement" in the United States; until recently, there has been little scientific research that focuses on men as parents. Demographers Dennis Hogan and Frances Goldscheider have described some broad trends using census data from the late 1800s through 1990 that examine men's likelihood of sharing a residence with children under age 15, whether they are the fathers' own children (including stepchildren) or other children. Hogan and Goldscheider show that 42 percent of men ages 18 to 75 lived with their own biological children or stepchildren in 1880. This dropped to 33 percent in 1940 and to 28 percent in 1990 — a decline that parallels declines in the birth rate. The major interruption to this long-term trend was the baby boom years, captured in the 1950 through 1970 censuses. Almost one-half of adult men lived with an "own child" in these years, a higher percentage than in 1880.⁵³

How many years do men spend as parents? Demographer Rosalind King has recently estimated the number of years men and women will spend as parents of own (biological) children or stepchildren under age 18 if the parenting patterns of the late 1980s and early 1990s continue throughout their lives.⁵⁴ Almost two-thirds of the adult years will be "child-free" years in which the individual does not have biological children under age 18 or responsibility for anyone else's children. Men will spend about 20 percent of their adulthood living with and raising their biological children while women will spend more than 30 percent of their lives raising biological children (see Figure 7). Whereas women, regardless of race, spend nearly all of their parenting years rearing their biological children, men are more likely to live with stepchildren or a combination of their own and stepchildren. Among men, whites will spend about twice as much time living with their biological children as African Americans.⁵⁵

Figure 7
Percentage of Adulthood Spent as Parents With Children Under Age 18

Note: These estimates are for men and women who were ages 20 to 24 in 1987-1988, and estimate the number of years (until age 69) adults will spend living with biological children, stepchildren, or foster children.

Source: R.B. King, Demography 36, no. 3 (August 1999): 377-85, table 2.

Father-Only Families

One of the new aspects of the American family in the last 50 years has been an increase in the number of families maintained by the father without the mother present. Between 1950 and 2000, the number of households with children that were maintained by an unmarried father increased from 229,000 to nearly 1.8 million. An additional 250,000 unmarried fathers lived with their children in someone else's household, bringing the total count of single fathers to about 2 million for 2000.

While mothers generally get custody of children after divorce, shared physical custody — in which children alternate between their mother's and father's households — has become more common in recent years. While divorced fathers still rarely are granted sole custody of their children, shared custody promotes close involvement in all aspects of their children's lives.⁵⁶

In one of the few analyses of child custody trends, researchers Maria Cancian and Daniel Meyer examined the custody outcomes of divorce cases in Wisconsin between 1986 and 1994.⁵⁷ Their study found little change in the percentage of cases in which fathers were awarded sole custody of children; fathers got sole custody in about 10 percent of divorce cases throughout the period. The percentage of cases in which mothers were awarded sole custody declined, however, from 80 percent to 74 percent, and the percentage of cases with shared physical custody between the mother and father rose from 7 percent to 14 percent over the period. If trends in other states are similar to Wisconsin's, fathers are becoming more involved in the lives of their children after divorce, not as sole custodians but as increased participants in legal decisions (joint legal custody) and in shared physical custody arrangements.

Grandparenting

One moderating factor in children's well-being in single-parent families can be the presence of grandparents in the home. While the image of single-parent families is usually that of a mother living on her own, trying to meet the needs of her young child or children, many single mothers live with their parents. In 1998, about 17 percent of unmarried mothers with children lived in the homes of their parents, compared with 10 percent of single fathers (see Table 3). A much higher percentage of single mothers (36 percent) live in their parents' home at some point before their children are grown. Black single mothers with children at home are even more likely to live with a parent at some time. Never-married women are especially likely to live with their parents after they have a baby: 60 percent of white and 72 percent of black single mothers who had a child outside of marriage resided with their parent(s) for a period.⁵⁸

The involvement of grandparents in the lives of their daughters (and sons) who become single parents is receiving increased attention. Court cases are ruling on grandparents' visitation rights, and some welfare reform measures highlight the responsibilities of parents whose teenage daughters become mothers. The 2000 Census included a new set of questions, mandated as part of welfare reform, on grandparents' support of grandchildren.

Emerging research shows that grandparents play an important role in multigenerational households — which is at odds with the traditional image of grandparents as family members who require financial and personal support. While early studies assumed that financial support flowed from adult children to their parents, more recent research suggests that the more common pattern is for parents to give financial support to their adult children.⁵⁹

In most multigenerational households, the grandparents bring their children and grandchildren into a household the grandparents own or rent. In 1997, three-fourths of multigenerational households were of this type. In nearly one-third of the grandparent-maintained families, grandparents lived with their grandchildren without the children's parents.⁶⁰

Grandparents who own or rent homes that include grandchildren and adult children are younger, healthier, and more likely to be in the labor force than are grandparents who live in a residence owned or rented by their adult children. Grandparents who maintain multigenerational households are also better educated (more likely to have at least a high school education) than grandparents who live in their children's homes.

Parents who maintain a home that includes both dependent children and dependent parents have been referred to as the "sandwich" generation, because they provide economic and emotional support for both the older and younger generations. Although grandparents in parent-maintained households tend to be older, in poorer health, and not as likely to be employed, many are in good health and are, in fact, working. These findings suggest that the burden of maintaining a "sandwich family" may be overstated in the popular press. Many of these grandparents are capable of contributing to the family income and helping with supervision of children.

Well-Being of Single-Parent Families

The well-being of children, mothers, fathers, and grandparents is influenced by the type of family and household in which they live. In general, single-mother families have the lowest incomes while married-couple families have the highest.

Mothers who have never married tend to have many characteristics associated with low income and poverty. First, mothers who have never married are younger than divorced or separated mothers, single fathers, or married parents. Never-married mothers have less education and are less often employed than divorced mothers. The family income of children who reside with a never-married mother is less than one-fourth that of children in two-parent families (see Table 5). Almost three of every five children who live with a never-married mother are poor. Mothers who never married are much less likely to get child support from the father than are mothers who are divorced or separated. While 60 percent of divorced mothers with custody of children under age 21 received some child support from the children's father, less than 20 percent of never-married mothers reported receiving regular support from their child's father.

Table 5
Socioeconomic Characteristics of Children in One-Parent and Two-Parent Families, 1998

Note: Characteristics in this table are weighted by the number of children in each household type.

Source: U.S. Census Bureau, March 1998 Current Population Survey.

Children who live with a divorced mother tend to be much better off financially. Although the family income of divorced mothers is less than one-half the income of two-parent families and rates of poverty are greater, divorced mothers are substantially better educated and more often employed than are mothers who are separated or who never married. Homeownership is also significantly higher among families of divorced mothers, although not as high as among two-parent families.

How do single fathers who live with their children compare with married fathers? CPS data for 1998 show that married fathers were better educated, more likely to be employed full-time year-round, more often working in managerial or professional occupations, more often homeowners, and have higher earnings and income than single fathers. Never-married fathers were far more disadvantaged economically than divorced or separated fathers. In addition to their relatively young age, never-married fathers were the most likely to have dropped out of high school and the least likely to work full-time year-round or work in a managerial or professional job.⁶¹

Income and Poverty

• Figure 8: Per Capita Income of Selected Family Types Relative To Two-Parent, Dual-Earner Families, 1977, 1987, and 1997
• Poverty
• Figure 9: Poverty Rates by Age and Overall Poverty Rate, 1959 to 1999
• Children's Material Hardship
• Increase in Family Income Inequality
• Economics and Changes in Family Life

The amount of income available to each American depends on the type of family in which he or she lives, and how many people in the family earn incomes. For families with children, the income differences can be stark. Families with two parents who both work earn the highest family incomes, and their family members have the highest per capita incomes. Two-parent families in which just one parent works produce less income, and they have been losing ground to the two-earner families since the 1970s. For every $1 of per capita income in the two-earner family, the one-earner family — the traditional working dad and stay-at-home mom family — had just 82 cents in 1977, and just 66 cents in 1997 (see Figure 8). Perhaps because they tend to have fewer family members, father-only families supported by the father's income had higher per capita incomes than married-couple families supported by one employed parent. But they too have lost ground to families in which both the mother and father are employed.

Figure 8
Per Capita Income of Selected Family Types Relative To Two-Parent, Dual-Earner Families, 1977, 1987, and 1997

Source: L. Casper and S. Bianchi, Trends in the American Family (forthcoming).

Families with two working parents also have higher living expenses than families with a parent who is not employed outside the home, and many question whether the extra income really covers these expenses. Many dual-earner families pay dearly for child care that a parent would provide in a traditional one-earner family. They often spend more on convenience foods, household help, restaurant meals, and clothing than families with a parent at home. Nevertheless, an increasing proportion of two-parent families have both parents in the labor force, and the income-advantage of having two working parents has increased in recent decades. A single-parent family that relies on the mother's income had just 53 cents for every dollar of per capita income in the two-earner families in 1997. The children of single-mother families in which the mother is unemployed face especially bleak economic circumstances: they have just 17 cents for every $1 of income in the dual-earner families, down from 26 cents in 1977.

Poverty

Household and per capita income assesses how families are doing, on average, but an assessment of family economic well-being also requires a look at those families that end up at the bottom of the income distribution and experience substantial material hardship. The U.S. Census Bureau began calculating the poverty rate in 1959 and has tracked it annually ever since. During the 1960s, there was a dramatic decline in poverty rates as the economy expanded and the War on Poverty began (see Figure 9). Rates continued downward until the 1970s, and then began to rise. Poverty rates tended to fluctuate with recessions, reaching a peak in the early 1980s, falling off in the later 1980s, then peaking again in 1993. During the latter half of the 1990s, a period of strong economic performance in the United States, the poverty rate dropped and the number in poverty declined, although the rate has remained above the lows reached in the 1970s.

Figure 9
Poverty Rates by Age and Overall Poverty Rate, 1959 to 1999

Note: Data unavailable from 1960 to 1966 for those age 65 or older.

Source: U.S. Census Bureau, Current Population Reports P60-210 (2000): tables B-1, B-2.

The chances of falling into poverty vary greatly by family type, the employment status of the parents or grandparents, and the presence of children. Married-couple households, which have the highest average incomes, have the lowest poverty rates. But their poverty rate is twice as high if they have children than if they do not have children in the home. Families maintained by women with children have the highest poverty rates: in 1997, 42.5 percent of these households were poor, six times the rate for married couples with children.

Families in which both parents work have extremely low poverty rates; the rate was just 2.4 percent in two-parent, dual-earner families in 1997. Poverty rates rise considerably — to almost 15 percent in 1997 — if only the father works. Single-parent families with an employed mother had poverty rates of about 30 percent in 1997.

In 1978, sociologist Diana Pearce argued that poverty was "rapidly becoming a female problem," that poverty was "feminizing."sup>62 She suggested that female-headed families were losing ground vis-à-vis families with an adult male present in the household, noting that the ratio of income in female-headed families to other families had declined between 1950 and 1974. Between 1950 and 1980, poverty seemed to be affecting women disproportionately: The ratio of women's to men's poverty rates increased during the period among adult whites and blacks at all ages.⁶³

After 1988, the feminization of poverty appeared to cease among the working-age population, although it may have continued among the elderly. Women's increased salaries and job opportunities and men's decreased access to well-paying jobs in the 1980s coincides reasonably well with the halt in the feminization of poverty among working-age adults after 1980.⁶⁴

The feminization of poverty also led to a "juvenilization" of poverty in the late 1970s and 1980s, because many of the women falling into poverty had children at home. Child poverty fell dramatically between the Great Depression of the 1930s and the mid-1970s.⁶⁵ In 1959, when the official poverty figures were established, 27 percent of children were counted as poor. This declined to about 15 percent by the 1970s. Poverty declined even more sharply for the elderly: from 36 percent in 1959 to about 15 percent in the 1970s. Poverty rates for the elderly and children were about the same. After the mid-1970s, however, trends diverged for children and the elderly. The poverty rate for the elderly fell to about 11 percent by the mid- to late 1990s. Poverty rates for children rose after the mid-1970s. The rates fluctuated between 20 percent and 23 percent throughout most of the 1980s and early 1990s, and dropped below 20 percent only in 1998. Hence, children's relative risks of poverty skyrocketed vis-à-vis the poverty risk of the elderly. In the late 1960s, children's odds of being poor were only about 60 percent of those of the elderly; by the early 1990s, children's poverty rates were almost twice those of the elderly.

Children's Material Hardship

While there is little question that poverty rates have increased among U.S. children, living conditions for children have improved in many ways — especially in the poorest households.⁶⁶ Poor children are increasingly better housed over time: The percentage of low-income children living in homes without a complete bathroom, with leaky roofs, holes in the floor, no central heat, no electric outlets, or no sewer or septic system has declined substantially. Children are less likely to live in crowded housing conditions, though the safety and vibrancy of the neighborhoods surrounding these residences may have deteriorated over time.

Poor children today are also more likely to receive medical attention than in the past. The percentage of children who had not visited a doctor in the previous year declined, especially during the 1970s. Poor children are more likely to be immunized than in the past. Children at the bottom of the income distribution became more likely to live in families that owned an air conditioner and had telephone service. The improvement in these indicators raises the possibility that trends in official poverty based on money income do not adequately capture trends in material hardship of children. Children's housing, health, and access to certain items such as telephones and air conditioning seem to have improved over time, even among those with the lowest one-fifth of income. The most dramatic improvement took place during the 1970s, but conditions have not deteriorated since then and some indicators (access to air conditioning, for example) suggest continued improvement.⁶⁷

Increase in Family Income Inequality

In 1949, 15 percent of children lived in the poorest 20 percent of families, but by 1996, 28 percent of children were in the poorest 20 percent. The percentage of children in families at the bottom of the income distribution almost doubled over a 50-year period. This widening gap in income for families with and without children paralleled a notable shift in family structure. More children lived in mother-only families, which are more likely to have extremely low incomes: 15 percent of families in the bottom fifth of the income distribution had a female householder in 1949, but this rose to 42 percent by 1996. Reflecting this shift toward mother-child families, in 1949, 75 percent of the poorest families had at least one employed person in the household, while in 1996, 62 percent of these families had at least one person working.⁶⁸

At the other end of the income distribution — in the top 20 percent of families — another revolution was underway. The likelihood that married-couple families had earnings from both husband and wife grew tremendously. The likelihood of a wife being employed outside the home rose for married-couple families at all income levels. In 1949, for example, 32 percent of married-couple families in the richest 20 percent of families had an employed wife; this rose to 81 percent in 1996.⁶⁹

By any measure, income inequality has grown in the United States in recent decades. In 1970, the bottom 20 percent of households controlled only 4.1 percent of all household income, whereas the top 20 percent of households controlled 43.3 percent of income. By 1999, the gap had widened — the income share of the poorest 20 percent of families dropped to 3.6 percent, while it expanded to 49.4 percent for the families in the top 20 percent.

Why has the gap between the poorest and wealthiest families increased? The answer is complex and involves a mix of factors having to do with how well certain types of workers are doing in the labor force and with marriage and living arrangements. Inequality in the earnings distribution increased as the earnings of less-educated workers fell behind those of college-educated workers. Families relying on earnings from high school-educated (or less educated) householders are increasingly disadvantaged. The increase in single-parent households during this same period only amplified the income inequality.⁷⁰

Economics and Changes in Family Life

Changing family structure and changing economic opportunities of men and women interweave in complex ways to widen the gap in family and household income. Shifts in family structure and economic change interact and affect each other. A relatively large body of theory and empirical research assesses the ways in which economic changes may have promoted such changes in the family as later marriage, more cohabitation, less marriage (especially among the African American population), and more divorce and marital separation.

There have been two main thrusts in the literature on the economic causes of changes in marriage timing and divorce. First are theories that place major emphasis on male economic opportunities as determinants of when couples marry, whether couples marry at all, and whether marriages are disrupted. The second set of theories emphasizes the role of women's enhanced labor market opportunities in the delay in first marriage and the increase in marital disruption. Some argue that women's labor market opportunities should encourage marriage because women's high earnings may make them more "attractive" as a marital partner. Others assert that women's employment inhibits marriage and erodes marital stability to the extent that women's increased opportunities for employment and enhanced earnings make them more independent and less inclined toward marriage.

Men's breadwinning ability speeds the transition from cohabitation to marriage. Men with higher earnings are more likely to marry and less likely to divorce.⁷¹ Married men earn more than unmarried men, whether because higher-earning men are more likely to marry or because marriage changes the behaviors of men in ways that enhance labor market productivity. Men's unemployment is associated with an increased likelihood of divorce.⁷² All this is evidence for the importance of men's employment and earnings in facilitating marriage and enhancing marital stability.

The alternative hypothesis, that women's employment destabilizes marriage, is prominent in both sociological and economic theories of marital instability. Under this hypothesis, women's decreased specialization in childrearing and household maintenance, coupled with their increased labor force participation, reduce the benefits from marriage for men and women.⁷³ The effect of a wife's increased economic resources on marital disruption has been termed the "independence effect." Because of this effect, women who can support themselves and their children, either through their own earnings or via welfare payments, would have less incentive to marry and find it easier to exit unsatisfactory marriages.

The independence effect may delay marriage as men and women take longer to search for a good match when the job market is unstable, but there is little evidence that men or women are rejecting marriage because of women's increased economic opportunities.⁷⁴ The relationship among employment, earnings, and marriage is quite similar for men and women: Higher earnings and better job prospects enhance the likelihood of marriage, although the effects are larger and more consistent for men's than for women's employment and earnings.⁷⁵ Evidence relating a wife's greater economic independence to an increased likelihood of divorce is mixed. Studies that find that the wife's relative contribution to family income or her wages are positively associated with increased risks of marital disruption are counterbalanced by studies that find no relationship between a wife's economic independence and marital disruption.⁷⁶

Just as economic conditions affect who marries and divorces, decisions about whether to marry or remain married have economic consequences. The wide gap in economic conditions of husband and wife after divorce or separation has garnered considerable attention among the American public, and may affect the decision to separate, especially if dependent children are involved. When a couple with children separates, the children usually live with their mother. The mother is likely to have a lower salary than her husband, but much higher living expenses, especially if she must pay for child care. Many mothers do not receive regular child support. Fathers, however, may reduce their living costs when they separate from their family, while maintaining the same salary.

For couples who had children and who separated in the late 1980s or early 1990s, economic well-being declined for mothers by 36 percent, while the financial status of fathers improved by 28 percent. One-fifth of mothers improved their standard of living, but two-thirds of fathers benefited financially from the separation. In poor families, fathers are significantly more likely than mothers to rise out of poverty after marital separation. Only one-quarter of fathers who were poor when they separated remained in poverty, compared with almost three-quarters of mothers. Similarly, among couples whose income was above the poverty line before the separation, 19 percent of mothers compared with 3 percent of fathers fell below poverty as a result of the separation.⁷⁷

In sum, a complex set of interconnections among changes in family structure, decisions adults make about employment, and conditions in the labor market and the larger economy constrain, or at least set the context, for how well families do economically.

Blending Work and Family

• Table 6: Hours and Weeks of Paid Work for All Women, Married Women, and Married Women With Young Children, 1978 and 1998
• Housework
• Table 7: Average Hours Spent on Housework by Women and Men, Ages 25 to 64, 1965 and 1995
• Child Care
• Figure 10: Trends in Child-Care Arrangements for Preschool Children Whose Mothers are Employed, 1965 and 1994
• Attitudes About Women's Work and Family
• Box 4: Home Alone? Children Caring for Themselves
• Box 5: How Involved Are Fathers? It Depends on Whom You Ask…

A little over two decades ago, sociologist Rosabeth Kanter pointed out that, despite the "myth of separate worlds" of work and family, there were a myriad of ways in which work "spilled" over into family life.⁷⁸ The influence of work on family continues to be strong, probably stronger than the influence of family on work. Financial rewards from work define the opportunities families have and, hence, men, women, and children are called upon to adapt home life to the work schedules of the adults within the home. The dramatic increase in mother's labor force participation, particularly married mothers, has meant that more married couples commit more than 40 hours a week to market work because now both husband and wife are working for pay. A sizable reallocation of time is underway in American households with children.

Perhaps the most striking aspect of women's labor force participation over the past 20 years is how steadily the trends have moved upward.⁷⁹ In 1978, almost 66 percent of women in the prime "work and family" ages (25 to 54 years) worked during the year; this increased to 79 percent in 1998 (see Table 6). The percentage of women in these ages who worked full-time year-round increased from 32 percent to 50 percent between 1978 and 1998. The average annual hours of paid employment for all women increased 40 percent over the period, from 1,002 hours to 1,415 hours. Most of the increase occurred because more women were working in 1998 rather than because working women dramatically increased the number of hours they worked during the year.

Table 6
Hours and Weeks of Paid Work for All Women, Married Women, and Married Women With Young Children, 1978 and 1998

Source: Authors' tabulations based on March supplements of the Current Population Surveys in 1978 and 1998.

In recent decades, the most dramatic increases in labor force participation have been among married women, particularly those with young children. In 1998, 71 percent of married mothers of children under age 6 did some work for pay during the year. But just 35 percent worked full-time year-round, which means that nearly two-thirds of married mothers of preschoolers did not work full time in 1998. Most married mothers have not traded raising their own children for paid work. Although U.S. mothers of young children are much more likely to work in the 1990s than they were in the 1970s, which implies an increasing attachment of women to market work, married mothers tend to scale back their hours during their children's preschool years.

Housework

How have American men and women fulfilled their childrearing and job responsibilities, given women's increased hours in the labor market? One strategy has been to do less housework — or to do it faster. In 1965, women spent about 30 hours weekly doing unpaid household work, which included such core tasks as cooking meals, meal clean-up, housecleaning, and laundry, as well as more discretionary or less time-consuming tasks such as outdoor chores and repairs, gardening, animal care, and bill paying. By 1995, women spent just 17.5 hours per week on these tasks (see Table 7). While U.S. men spent much more time on household tasks in the 1990s than in the 1960s, they did not make up for the decline in women's time on housework. Women averaged 6.1 times more hours of housework than men in 1965, but this ratio fell to 1.8 in 1995 primarily because women spent so much less time on these chores. Men increased their time on weekly household chores from 4.9 hours to 10.0 hours between 1965 and 1995.

Table 7
Average Hours Spent on Housework by Women and Men, Ages 25 to 64, 1965 and 1995