AMERICAN FAMILIES

Many long-standing assumptions about American families have been challenged by family scholars. Among these assumptions is the belief that in colonial times the American family was extended in its structure, with three generations living together under one roof. It has been commonly believed that the nuclear family came about as a result of industrialization, with smaller families better able to meet the demands of an industrialized economy. However, historical data show that the extended family was not typical in the colonial era and that the earliest families arriving from Great Britain and other western European countries were already nuclear in structure (Demos 1970; Laslett and Wall 1972).

Some commonly held views about contemporary families also have been debunked in the research literature. For example, it has been commonly thought that American families neglect their elder members and are quick to place them in nursing homes. However, research shows that most older Americans have frequent contact with one or more members of their family, and that families typically provide extensive, long-term care to older persons when such care is needed. In most cases, families turn to placement in a nursing home as a last resort rather than a first option when elder members grow ill or frail.

More generally, family scholars have successfully challenged the notion of "the American family." As Howe (1972, p. 11) puts it, "the first thing to remember about the American family is that it doesn't exist. Families exist. All kinds of families in all kinds of economic and marital situations." This review will show the great diversity of family patterns that characterize the United States of the past, the present, and the foreseeable future.

HISTORICAL OVERVIEW

It is unfortunate that textbooks intended for courses on the family rarely include a discussion of Native Americans, for a historical examination of these groups shows a striking range of variation in family patterns. In fact, virtually all the variations in marriage customs, residence patterns, and family structures found the world over could be found in North America alone (Driver 1969). Though some of these traditional family patterns have survived to the present day, others have been disrupted over the course of U.S. history. It is important to note however, that research has not confirmed the commonly held assumption that Native-American societies were placid and unchanging prior to European contact and subsequent subjugation. Important changes were taking place in Native-American societies long before the arrival of Europeans (Lurie 1985).

As has been noted, European immigrants to the American colonies came in nuclear rather than extended families (and also came as single persons—for example, as indentured servants). It was long believed that colonial families were very large, with some early writers claiming an average of ten to twelve children per family, and that most people lived in extended rather than nuclear families. Family scholars, however, have cited evidence showing somewhat lower numbers of children, with an average of eight children born to colonial women (Zinn and Eitzen 1987). Scholars also have distinguished between number of children born per woman and family size at a given point in time. Average family size was somewhat smaller than the average number of children born, due to high infant mortality and because the oldest children often left home prior to the birth of the last child. Evidence suggests an average family size of five to six members during colonial times (Nock 1987). Thus, although the average size of colonial families was somewhat larger than today's families, they were not as large as has been commonly assumed. Furthermore, most people in colonial America lived in nuclear rather than extended family settings.

To understand the size and composition of colonial households, consideration must be given to nonrelated persons living in the home. Servants often lived with prosperous colonial families, and other families took in boarders and lodgers when economic conditions required such an arrangement (Zinn and Eitzen 1987). Households might also include apprentices and other employees. The presence of nonfamily members has important implications for family life. Laslett (1973) has argued that the presence of nonkin meant that households offered less privacy to families and hence provided the opportunity for greater scrutiny by "outsiders." Colonial homes also had fewer rooms than most American homes today, which also contributed to the relatively public nature of these households.

Colonial communities placed great importance on marriage, particularly in New England, where sanctions were imposed on those who did not marry (for example, taxes were imposed on single men in some New England colonies). However, historical records indicate that colonists did not marry at especially young ages. The average age at marriage was twenty-four to twenty-five for men and twenty-two to twenty-three for women (Leslie and Korman 1989). Older ages at marriage during this era reflect parental control of sons' labor on the farm (with many sons delaying marriage until their fathers ceded land to them) and also reflect the lower relative numbers of women (Nock 1987). Parents also typically exerted strong influence over the process of mate selection but did not control the decision. Divorce was rare during this period. The low divorce rate cannot be equated with intact marriages, however. Spousal desertion and early widowhood were far more common experiences than they are today.

The population for the American colonies came primarily from Great Britain, other western European countries, and from western Africa. Initially brought to the colonies in 1619 as indentured servants, hundreds of thousands of Africans were enslaved and transported to America during the colonial period (by 1790, the date of the first U.S. census, African Americans composed almost 20 percent of the population; Zinn and Eitzen 1987). It has been commonly assumed that slavery destroyed the cultural traditions and family life of African Americans. The reasoning behind this assumption was that slave families often were separated for sale to other masters, males were unable to provide for or protect their families, and slave marriages were not legal. The stereotype of "matriarchal" black families, in which women are the family heads and authorities, usually assumes that slavery produced this family form. Empirical research challenges these assumptions. Though slave families lived in constant fear of separation (Genovese 1986), many slave marriages were strong and long-lasting (Gutman 1976). Marriages were legitimized within the slave community (symbolized, for example, by the ritual of "jumping over a broomstick"; Boris and Bardaglio 1987), and two-parent families were common among slaves as well as among free blacks in the North and the South (Gutman 1976). A variety of family structures, including female-headed households, were found in the slave community and attested to the importance placed on kin ties. Rather than the "absent family" assumed to characterize slave life, slaves were connected to one another through extensive kinship networks (Genovese 1986). Extended kin ties continue to be an important aspect of African-American families today.

For decades, the heritage of slavery and its presumed effects on family life have been invoked to explain social problems in poor black communities (e.g., Moynihan 1965). The historical evidence described above does not lend support to this explanation. Most writers today argue that social problems experienced in poor black communities can more accurately be attributed to the effects of discrimination and the disorganizing effects of mass migration to the urbanized North rather than to the heritage of slavery (e.g., Staples 1986).

Societal changes associated with the Industrial Revolution profoundly affected all types of American families, though the specific nature and extent of these effects varied by social class, race, ethnic origins, and geographic region. Prior to the Industrial Revolution, family members worked together as an economic unit. Their workplace was the home or family farm, with families producing much of what they consumed. Family life and economic life were one and the same, and the boundaries between "private life" in the family and "public life" in the community were blurred. With the development of a commercial economy, the workplace was located away from the family unit. Separation of work and family life created a sharp distinction between the "public" realm of work and the "private" realm of family. Particularly for women's roles, changes initiated by the Industrial Revolution have been long-lasting and far-reaching. Increasingly, women's roles were defined by activities assumed to be noneconomic, in the form of nurturing and caring for family members. This was especially true for middle-class women, and married women were excluded from many jobs. Poor and working-class women often participated in wage labor, but this work was generally seen as secondary to their family roles. Men were viewed as having primary responsibility for the economic welfare of their families. No longer an economically interdependent unit, families were transformed such that women and children became economically dependent on the primary wage earner.

Thus, children's roles and family relationships also changed with industrialization. In contrast to earlier times, in which children were viewed as miniature adults and engaged in many of the same tasks they would also perform as adults, childhood came to be seen as a special stage of life characterized by dependence in the home. And although children in working-class homes were more likely to work for pay, the evidence suggests that these families also viewed childhood as a stage of life distinct from adulthood (Zinn and Eitzen 1987). Overall, the family became increasingly defined as a private place specializing in the nurturance of children and the satisfaction of emotional needs, a "haven in a heartless world" (Lasch 1977).

Family structures also changed during the 1800s. Family size declined to an average of four to five members. (Of course, average numbers obscure variation in family sizes across social classes and other important dimensions such as race and ethnicity.) Though the average size of nineteenth-century American families was close to that of today's families, women bore more children during their lifetimes than do American women today. Infant and child mortality was higher and births were spaced further apart, thus decreasing the average size of families at a given point in time. Household size also declined, with fewer households including nonrelated persons such as boarders or apprentices. The average ages at which women and men married were similar to those of colonial times, with an average of twenty-two for women and twenty-six for men. However, greater life expectancy meant that marriages typically lasted longer than they did during the colonial period (Nock 1987).

From 1830 to 1930 the United States experienced two large waves of immigration. The first wave, from 1830 to 1882, witnessed the arrival of more than ten million immigrants from England, Ireland, Germany, and the Scandinavian countries. During the second wave, from 1882 to 1930, over twenty-two million people immigrated to the United States. Peoples from northern and western Europe continued to come to the United States during this second wave, but a large proportion of immigrants came from southern and eastern Europe as well (Zinn and Eitzen 1987). Immigrants' family lives were shaped by their ethnic origins as well as by the diverse social and economic structures of the cities and communities in which they settled.

Ethnic traditions also helped Mexican-American families adapt to changing circumstances. Annexation of territory from Mexico in 1848 and subsequent immigration from Mexico produced sizable Mexican-American communities in the Southwest. Immigrants from Mexico often reconstructed their original family units within the United States, typically including extended as well as nuclear family members. Extended family households are more common today among Mexican Americans than among non-Hispanic whites, reflecting Mexican Americans' strong family orientation (or "familism") as well as their less advantaged economic circumstances (Zinn and Eitzen 1987).

Imbalanced sex ratios among Chinese and Japanese immigrants greatly influenced the family experiences of these groups. First coming to the United States in the early 1900s, Chinese immigrants were predominantly male. The Chinese Exclusion Act of 1882 barred further immigration, and only wealthy Chinese men were able to bring brides to the United States (Boris and Bardaglio 1987). As of 1910, there were 1,413 Chinese men in the United States to every 100 Chinese women. This sex ratio was still skewed in 1940, when there were 258 men to every 100 women. In contrast to the extended family networks typical of traditional Chinese culture, many Chinese-American households consisted of single men living alone (Marden and Meyer 1973).

Substantial immigration from Japan took place between 1885 and 1924. Like traditional Chinese families, Japanese families were based on strong extended kin networks. As was true for Chinese immigration, most Japanese immigrants were male. In addition to immigration restrictions, Japanese-American families (especially those on the West Coast) were disrupted by property confiscation and the mass relocations that took place during World War II (Marden and Meyer 1973).

In addition to the "old" immigration of the mid-nineteenth century and the "new" immigration of the late nineteenth and early twentieth centuries, a third wave of large-scale immigration to the United States began in the mid-1960s. In contrast to the earlier waves, when most immigrants came from European countries, most immigration in this third wave has been from Latin America and Asia. However, as has been true of earlier periods of immigration, public controversies surround the economic and social absorption of these new groups (Marger 1991). In addition to occupational and economic challenges facing immigrant families, social challenges include the continuing debate over whether schools should provide bilingual education to non-English-speaking children.

MARRIAGE AND FAMILY TRENDS IN THE UNITED STATES

The separation of paid work and family life, associated with the transition to an industrialized society, gave rise to profound changes in family life. Over the course of the twentieth century, women's roles were defined primarily by family responsibilities within the "private sphere" of the home, but except for a brief period following World War II, women's labor-force participation rose steadily. As of 1997, nearly half (46 percent) of all employed workers were women. Increases in labor-force participation were especially great among married women. In 1900, less than 4 percent of married women were in the labor force. By 1997, that figure had risen to 62 percent (U.S. Bureau of the Census 1998). In contrast to earlier eras of American history, when African-American women were more likely to work for pay than white women, rates of labor force participation are now nearly the same for women in these racial groups, for both married and unmarried women. Married Hispanic women are slightly less likely to be employed than married African-American and non-Hispanic white women (U.S. Bureau of Labor Statistics 1991). As discussed below, women's labor-force participation has important implications for many dimensions of family life.

Though American families have changed in important ways over the past 100 years, examination of historical trends also reveals continuation of some family patterns begun long ago. Notably, the period of the 1950s is commonly thought to mark the end of a golden age of family life. However, historical data show that for a number of family patterns, the 1950s was an unusual period. Lower rates of divorce, lower ages at marriage, and higher rates of childbearing observed during the 1950s have been attributed mainly to greater economic prosperity following the Great Depression and World War II (Cherlin 1992).

Age at First Marriage. According to U.S. census data, the average (median) age at first marriage in the United States was twenty-five years for women and 26.7 years for men in 1998 (U.S. Bureau of the Census 1998). These average ages are higher than for most U.S. population censuses of the past century, but the current median age for men is comparable to that documented for (white) men near the turn of the twentieth century. In 1890, the median age at first marriage was twenty-two for women and 26.1 for men. The average age at marriage declined until 1956, when the median age at first marriage was 20.1 for women and 22.5 for men. The average age at marriage subsequently began to rise (Saluter 1996). Comparison of non-Hispanic whites, blacks, and Hispanic whites shows that age at first marriage climbed more rapidly for blacks than for non-Hispanic whites, with blacks marrying later than non-Hispanic whites. In contrast, Hispanics marry at younger ages than do other groups. It is difficult to assess whether Hispanics' lower age at marriage reflects longterm trends within the United States due to the large numbers of Mexican Americans who immigrated in the 1980s and 1990s (Sweet and Bumpass 1987).

Factors promoting later age at marriage include greater societal acceptance of singlehood and cohabitation as well as greater emphasis on educational attainment. The relationship between age at marriage and level of education is nearly linear for non-Hispanic white men and women, with more education associated with later age at marriage. This relationship is more complex for minority groups, and especially for black and Hispanic men. For these men, later age at marriage is associated both with lower and higher educational levels, producing a U-shaped relationship between education and age at marriage. Minority men with low education are likely to have especially poor job prospects, which in turn affect prospects for marriage. Overall, less racial and ethnic diversity in age at marriage is shown for those with higher educational attainment (Sweet and Bumpass 1987).

Interracial Marriage. Most American marriages are homogamous with regard to race. In 1997, there were 1.3 million interracial marriages, representing only 2.3 percent of all marriages. Many Americans equate interracial marriage with black-white marriage, but only one-quarter of American interracial marriages are between blacks and whites. Currently, 4 percent of black men and 2 percent of black women are married to a white partner. Marriage between African Americans and Asian Americans is even less common than black-white marriages (U.S. Bureau of the Census 1998). Asian Americans are more likely than African Americans to marry a partner of another race. A 1990 study found that 23 percent of Asian Americans intermarry, and that over 40 percent of Japanese-American women are married to a man of a different race (Lee and Yamanaka 1990).

Although the total number of interracial marriages is quite small, the rate of growth in these marriages has been increasing. The number of interracial marriages in the United States increased fourfold between 1970 and 1998. During this period, the total number of marriages grew by only 24 percent (U.S. Bureau of the Census 1998).

Singles. Throughout U.S. history a proportion of individuals have remained unmarried. High levels of singlehood were recorded in the late 1800s, when 15 percent of women and 27 percent of men had not married by the age of thirty-four. In 1940, 15 percent of women and 21 percent of men had not married by this age, but by 1970, these proportions had dropped dramatically. By 1970, just 6 percent of women and 9 percent of women had not married by the age of thirty-four (Kain 1990). The size of the unmarried population has been increasing since the 1970s. In 1994, 20 percent of women and 30 percent of men had not married by age thirty-four (Saluter 1996). Women's changing roles have been linked with the rise in singleness: Women with higher education and higher personal income are less likely to marry or have children. Also, in contrast to earlier eras, there is greater societal tolerance of singlehood, providing greater freedom for both women and men to choose a single lifestyle.

Since few individuals marry for the first time at the age of sixty-five or older, a more accurate picture of the never-married population is provided when attention is restricted to the older population. At present, 3.8 percent of men and 4.7 percent of women age sixty-five and older have never married (Lugaila 1998). Due to the continuing stigmatization of homosexuality, it is difficult to ascertain the numbers of single persons who are gay or lesbian. Researchers have estimated that 4 percent of men and 2 percent of women are exclusively homosexual (Collins 1988). Though homosexual marriages are not legally recognized, many gay and lesbian couples form lasting unions.

Childbearing. Childbearing patterns have varied somewhat over the past century. Women born in 1891 had an average of three children. Women born in 1908, who bore children during the Great Depression, had an average of two children. This figure increased to three children per mother during the 1950s and has since declined to two children per mother on average today (Sweet and Bumpass 1987; U.S. Bureau of the Census 1998). In addition to fewer numbers of children born, current trends in childbearing include higher age at first childbirth and longer intervals of time between births. These trends are interrelated. Waiting longer to have a first child and spacing births further apart decrease the average number of children born per mother. The timing of childbearing also has important effects on other life experiences, including educational and occupational attainment. Lower rates of childbearing are associated with higher educational levels and higher incomes.

Fewer married couples are having their first child in the period immediately following marriage, but there are some important differences by race. In 1960, 54 percent of non-Hispanic white couples had children within twelve to seventeen months of marriage. In 1980 this figure dropped to 36 percent. Little change was shown over this period for black couples, who had children within twelve to seventeen months of marriage. Compared to the total population, black couples are likely to have more children on average and to have a child present at the time of marriage. For whites (Hispanics and non-Hispanics) as well as blacks, nine-tenths of all couples have children within seven to eight years of marriage (Sweet and Bumpass 1987).

Childbearing among single women has increased greatly over the past several decades. While just 5 percent of all births were to single women in 1960, approximately one-third (32.2 percent) of all births were to single women in 1995 (Saluter 1996; U.S. Bureau of the Census 1998). As is true for many family patterns, there are substantial variations across racial-ethnic groups. Asian Americans and Pacific Islanders show a relatively low rate of nonmarital births, at 16.3 percent of all births in this group. In contrast, 25.3 percent of all births among whites are to unmarried women, and the rate for Hispanics (who may be of any race) is 40.8 percent. Nearly 70 percent of African-American children are born to single women (U.S. Bureau of the Census 1998). Although socioeconomic factors do not account completely for births to single women, nonmarital childbearing in the United States tends to be higher among those who are poor. Socioeconomic factors can help to explain why African Americans, who are disproportionately likely to be poor, have had higher rates of nonmarital childbearing.

Divorce. A rising divorce rate has been a feature of U.S. society since the Civil War. At that time, the divorce rate per 1,000 existing marriages was just 1.2 (Jacobson 1959). By the early 1990s, the rate had increased to more than 20 divorces per 1,000 existing marriages (Cherlin 1996). This long-term trend does not show a smooth and progressive rise, however. Divorce rates have risen more sharply after every major war during this century. Divorce also increased following the Great Depression, apparently reflecting stresses associated with unemployment and economic deprivation. Economic prosperity as well as a greater emphasis on family life have been linked to the lower divorce rate observed from 1950 to 1962. Following 1962, dramatic increases in the divorce rate occurred, with a 100 percent increase between 1963 and 1975 (Cherlin 1992). By the early 1970s, the chance of eventual divorce reached almost 50 percent. The divorce rate has more or less stabilized since that time, such that approximately 50 percent of all first marriages are projected to terminate eventually in divorce. The chances of divorce are higher for second marriages, of which 60 percent are projected to end in divorce (Olson and DeFrain 1994).

Population trends have been linked with the increased rate of divorce. Among these trends is greater longevity, with an average life expectancy at birth of approximately eighty years for women and seventy-three years for men who were born in 1991. When they reach the age of sixty-five, women who were born in 1991 can expect to live an additional nineteen years, while men of that birth cohort can expect to live fifteen more years (U.S. Bureau of the Census 1996). (These figures are for the total population. Life expectancies are lower for members of racial and ethnic minorities.) In contrast, the life expectancy at birth for those born in 1900 was forty-eight years for women and forty-six years for men (Grambs 1989). Unsatisfactory marriages that formerly may have been terminated by the death of one partner are now more likely to be dissolved by divorce (Uhlenberg 1986). Scholars have also noted the apparent connection between women's increasing levels of labor-force participation and the increased rate of divorce in the United States. Divorce is more likely to occur in couples where the wife is able to support herself financially.

The risk of divorce also varies with age at marriage, duration of the marriage, education, race, and ethnicity. Age at marriage is one of the most important factors, with the likelihood of divorce twice as great among couples where the wife was seventeen or younger than among couples where the wife was in her early twenties. Further, most divorces take place within the first few years of marriage. The longer a couple has been together, the less likely they will be to get divorced. This pattern also holds among couples in which one or both partners have remarried. Education also seems to be an important factor, with a higher divorce rate observed among high school dropouts than among college graduates. However, the effect of education is due in large part to the fact that college graduates tend to marry at later ages. Looking across racial and ethnic groups, the risk of divorce is greater among African Americans than among whites, and especially high divorce rates are observed for Hispanics (Puerto Ricans in particular), Native Americans, and Hawaiians. Divorce is less common among Asian Americans (Sweet and Bumpass 1987).

Widowhood. Rising life expectancies have increased the average age of widowhood. Among women, the median age at widowhood was fiftyone in 1900, compared to sixty-eight in 1979. The median age at widowhood for men was forty-five in 1900 and seventy-one in 1979 (Grambs 1989). Lower average ages of bereavement for men compared to women in 1900 are linked with women's risks for death in childbirth in that era. Women today can expect to live longer in a widowed status compared to widowed men. This gender gap is explained primarily by higher female life expectancy and lower rates of remarriage among women, and also because women tend to marry men several years older than themselves. Nearly one-half (45 percent) of all American women who are age sixty-five or older are widowed, while 15 percent of men in this age group are widowed. Among the oldestold—those who are eighty-five years of age and older—these figures rise to 77 percent for women and 42 percent for men (Lugaila 1998).

Remarriage. Due in large part to the fact that widowhood tends to occur later in life, fewer men and women remarry following the death of a spouse, compared to those who remarry following a divorce. Most people who divorce eventually remarry, but the likelihood of remarriage varies greatly according to gender, age, and race. Approximately five out of six men eventually remarry following a divorce, compared to two out of three women who do so (Cherlin 1992). As noted above, men are also more likely than women to remarry following the death of a spouse. The probability of remarriage declines with age, especially among women. Only one in four women who divorce at age forty or older eventually remarry (Levitan, Sar, and Gallo 1988). Race differences also are observed for remarriage. The proportion of women who remarry following divorce is approximately 50 percent for African Americans and 75 percent for whites (Bumpass, Sweet, and Castro Martin 1990).

Increased rates of divorce and remarriage are transforming American families. "Blended" families or stepfamilies are becoming increasingly common, whereby one or both spouses bring children into a remarriage. One in four children will spend some time in a blended family (Furstenberg and Cherlin 1991). Nearly all of these children live with their biological mothers.

Household Structure. As defined by the U.S. Bureau of the Census, "family households" contain persons who are related to the household head (the person in whose name the home is owned or rented) by blood ties, marriage, or adoption. "Nonfamily households" consist of individuals who live alone or with one or more unrelated persons. Historically, most American households have been family households, and most of these have included married couples. In 1910, 80 percent of all households included married couples. By 1998 this percentage had declined to 53 percent (Casper and Bryson 1998). In contrast, the proportion of single-person households has risen dramatically over the century. In 1890, only 4 percent of all households were of this type (Sweet and Bumpass 1987). As of 1998, single-person households accounted for one-quarter of all U.S. households. Nearly 12 percent of all households consist of men living alone, 15 percent consist of women living alone (Lugaila 1998).

Breakdowns of family structure by race and ethnicity have shown that Americans of Korean, Filipino, Vietnamese, and Mexican heritage are most likely to live in family households (for each group, about 84 percent reported living in family households). African Americans and non-Hispanic whites are somewhat less likely to live in family households. Also, compared to other racial and ethnic groups, Puerto Ricans are most likely to live in a household consisting of a mother and one or more children (with 23 percent living in this type of household), followed by African Americans, Native Americans, and Hawaiians (Sweet and Bumpass 1987).

Type of household is tied closely with economic status. While the "typical" dual-earner couple with children earned an average annual income of $46,629 in 1991, the average income for mother-only households was only $13,012 (McLanahan and Casper 1998). Of all household types, those headed by a woman with no husband present have the highest poverty rate. In 1997, nearly 32 percent of these households had incomes that fell below the poverty line (Dalaker and Naifeh 1998).

Due largely to women's risks for poverty and the rise in female-headed households, children are more likely to be poor today than they were several decades ago. The proportion of American children living in poor families declined during the 1960s—from 26.5 percent in 1960 to 15 percent in 1970—but has since increased. Nearly one-quarter of all American children live in poor families. For children who live in a female-headed household, the chances of living in poverty rise to 46 percent for white-Hispanic and non-Hispanic children, and to 82 percent for African-American children (Ollenburger and Moore 1998).

IMPACT OF FAMILY TRENDS ON CHILDREN AND OLDER PERSONS

The majority of the family patterns described here represent long-term trends in the United States and are unlikely to represent the demise of American families, as has been decried by some social observers. Indeed, marriage remains as "popular" as ever. Although Americans are marrying somewhat later on average than they did in the 1950s and 1960s, the majority of women and men continue to marry and have children. Those who divorce tend to remarry. Although these trends did not originate in the late twentieth century, it is nonetheless true that divorce, single-parent households, employed mothers, and nonmaternal childcare are more typical features of American life today than they were in the past. The impact of these family patterns on children's development and well-being has been a matter of great concern to researchers and policymakers. The impact of women's employment and changes in family composition also have given rise to concerns regarding the provision of informal care to elderly parents.

Divorce. Parents' divorce has been linked with a range of negative outcomes for children in the areas of psychological adjustment, life satisfaction, academic achievement, and social relationships. These effects are strongest in the first year or two following the divorce, but some long-term consequences also have been found. The experience of parents' divorce can continue to have negative effects on a child's well-being as she or he grows into young adulthood (Cherlin, Chase-Lansdale, and McRae 1998).

Although divorce typically is a stressful experience, studies have found a great deal of variation in how children adapt to parents' divorce. Among the factors identified as important to consider are the family's socioeconomic status, race-ethnicity, the child's gender and his or her age at the time of the parents' divorce or separation. For example, the effects of divorce apparently are more acute for children of school age than for preschool children. Marital disruption also brings greater risks to children when the parent-child relationship suffers as a result of the divorce and when one or both parents experience multiple divorces (Amato and Booth 1991).

The level and type of conflict present in the family prior to a divorce also is important in understanding the effects of divorce on children. Children whose families were highly conflictual may show increased well-being following a divorce. Conversely, problems of adjustment have been found among children whose parents did not divorce, but whose family lives were characterized by high levels of conflict (Furstenberg and Cherlin 1991).

Individuals who experienced "low-stress" parental divorces do not appear to differ significantly from those who grew up in happy, intact families (Amato and Booth 1991). In general, negative consequences of divorce are not found for children when parents maintain a positive relationship with the child and with one another, and when the child is provided with adequate social and economic resources. As discussed below, economic difficulties pose a central challenge for single-parent (usually single-mother) households.

Single-parent households. Just as the impact of divorce on children depends upon multiple factors, no one pattern characterizes how children's well-being is influenced by living in a single-parent household. The circumstances of single parenthood are diverse—single parents can be divorced, widowed, or never-married—and they have access to varying levels and types of social and economic resources. Households may include only the single parent and one or more children, or may include extended family or other household members. Noncustodial parents may or may not be part of the child's life.

It has been noted that single-parent households are more likely than dual-parent households to be poor, and that mother-headed households are especially likely to be poor. Following a divorce, women's subsequent income declines 27 percent on average, while men's average income increases by 10 percent (Peterson 1996). Some of this income gap is due to the fact that children are more likely to live with their mothers than their fathers following a divorce. Currently one child in four lives in a single-parent family, and women head 83 percent of these families (U.S. Bureau of the Census 1998).

Many of the problems associated with single-parent households are associated with problems of poverty. The economic difficulties faced by single parents are compounded when child support payments from the noncustodial parent are not provided regularly. About one-half of custodial mothers were awarded child support payments in 1992. Of the women who had been awarded child support, 76 percent received full or partial payment. A significant amount of awarded child support is not received by the custodial parent: one-third of all awarded child support was not paid in 1991. Among custodial parents, mothers have higher child support award rates and payment rates than fathers, but they are also much more likely to be poor. Single custodial mothers are two and one-half times more likely to be poor than single custodial fathers (U.S. Bureau of the Census 1995).

In addition to economic strains, single parents report problems arising from role and task overload, in coordinating a social life and parenting responsibilities, and difficulties with former spouses. Strains experienced by the parent can in turn impact the well-being of children (Richards and Schmiege 1993). Some evidence also suggests that single parents provide less supervision of children compared to two-parent families (married or cohabiting) (e.g., Astone and McLanahan 1991).

Although research has focused primarily on problems of single-parent households, some potential benefits of these households have been identified for children's development and well-being. For example, children in single-parent households have been found to take greater responsibility for household tasks than children in two-parent households. Along with increased responsibilities within the home, children in one-parent households also apparently develop higher levels of personal autonomy and independence (see Richards and Schmiege 1993).

Nonmaternal Childcare. With more American women working for pay than ever before, more preschool-age children are receiving care from their fathers, other family members, or from nonrelatives during their mothers' hours of employment. As of 1991, 9.9 million children age five or younger required care during the hours their mothers were employed. Of these children, the majority were cared for in a home environment: 36 percent were cared for in their own homes (usually by the father or another relative); 31 percent were cared for in another home (usually by a nonrelative in the care provider's home). A further 9 percent were cared for by their mothers while they worked—this was generally in a home environment as well, as most of these mothers had home-based paid work. Approximately one-quarter of the children who required care while their mothers worked were enrolled in an organized day care facility (U.S. Bureau of the Census 1995).

With the increasing use of childcare provided by individuals other than the mother, concerns have been raised regarding the impact of nonmaternal care on children's physical, cognitive, and psychological development. The majority of studies on this topic have concluded that nonmaternal childcare, and nonparental childcare more generally, is not in itself harmful to children. In fact, day care settings apparently aid the development of children's social skills. However, the quality of the care—whether in a family or nonfamily setting—is extremely important for children's development and well-being. A comprehensive report published by the National Research Council found that children whose families were undergoing psychological stress or economic deprivation were more likely to receive care in lower quality settings. Hence, these children are vulnerable not only to poverty and psychological stress within the home, but they are also likely to suffer from the effects of poor-quality care outside the home (Hayes, Palmer, and Zaslow 1990).

Eldercare. Most older Americans live independently in the community and are in relatively good health. Most also provide economic and other support to their adult children, such as providing babysitting or other services. It is only when the elder's health or economic status is severely compromised that the balance of exchange tips in the direction of receiving a greater amount of support than elders provide to their children.

The bulk of care provided to older Americans who are frail or ill is provided by family members, usually women, who are typically spouses or children of the elder (Stone, Cafferata, and Sangl 1987; Wolf, Freedman, and Soldo 1997). If a spouse or child is not available, other relatives or a friend may provide care (though assistance from these latter sources is typically less intensive and for briefer periods of time). Much research attests to the extensive and prolonged care provided by family members to elders within a home environment (see Dwyer and Coward 1992). In many if not most cases, it is only after families have exhausted their physical, economic, and emotional resources that a decision is made to place an elder in a longterm care facility. Indeed, only 5 percent of all Americans age sixty-five and older reside in a nursing home or other long-term care facility at a given point in time. The chances of nursing home placement rise with age: Nearly 25 percent of those age eighty-five or older live in a nursing home. However, even among this "oldest-old" age group, the majority of individuals reside in the community rather than a nursing home (Morgan and Kunkel 1998).

Recognizing that women typically are the primary care providers to older family members, researchers and policymakers have raised concerns that such care will be curtailed in the future due to trends in women's employment and an increasing older population. The prevalence of elder care among the employed population is difficult to estimate due to variations in sampling across studies and how caregiving is defined. However, an averaged estimate from studies on this topic is that about one in five employees has some elder care responsibilities (Gorey, Rice, and Brice 1992).

Research has shown that caregiving can be emotionally gratifying for individuals who are providing care to a loved one (Lechner 1992). However, intensive, long-term caregiving can produce serious physical, economic, and emotional strains, especially when caregiving interferes with the provider's work or household responsibilities (Gerstel and Gallagher 1993). A number of studies have found that employment status does not alter the type or amount of care provided to the elderly by their families, but some family members reduce or terminate their employment in order to provide elder care. A national survey found that 9 percent of employees in the study had quit their jobs and a further 20 percent had altered their work schedules to accommodate their caregiving responsibilities. Women were more likely than men to rearrange their work schedules or to reduce or terminate their employment in order to provide family care (Stone, Cafferata, and Sangl 1987).

The strains of intensive caregiving can compromise the provider's ability to continue in the caregiving role. A number of researchers have concluded that greater assistance from the government as well as from the workplace is needed to assist family members and others who provide care to older adults in the community (e.g. Lechner 1992). Such assistance not only benefits the health and well-being of care providers and recipients, but also allows elders to continue living in the community setting for a longer period of time, as is preferred by most elders and their families.

CROSS-NATIONAL COMPARISONS

The family trends described here are not unique to the United States. Other western societies also have witnessed a rise in the age at first marriage, an increased divorce rate, increased numbers of children born to unmarried women, and an increase in the numbers of women who participate in paid labor. Although the trajectories of these trends are similar for most western societies, differing proportions of individuals and families are represented in these trends across societies. For example, although the average age at marriage has been rising in the United States and most other western societies, the average age at first marriage is lower in the United States than in most of the European countries. And although a rising divorce rate has been a feature of the majority of western societies, the rate of divorce is highest in the United States. Out-of-wedlock births also have increased in western societies over the course of the century. However, in contrast to Scandinavian countries including Sweden and Denmark, more children in the United States live in a single-parent household than with cohabiting parents.

A striking gap between the United States and a number of other countries is found in its relatively high rate of infant mortality. Infant mortality is linked with poverty and the lack of adequate nutrition and health care. The wealth enjoyed by the United States as a nation belies the economic deprivation experienced by subgroups within the society. In 1988, the U.S. infant mortality rate was 10 babies per 1,000 live births. This rate was higher not only as compared with other western societies, but also compared with many nonwestern countries. In a ranking of infant mortality rates worldwide, which were ordered from the lowest to the highest rates, the United States ranked eighteenth from the bottom. This rate was higher than that found in countries including Singapore, Spain, and Ireland, which are poorer relative to the United States, but which have lower levels of economic inequality within the society (Aulette 1994). A statement from the Children's Defense Fund illustrates the high risks for infant mortality faced by racial-ethnic minority groups within the United States: "A black child born in the inner-city of Boston has less chance of surviving the first year of life than a child born in Panama, North or South Korea or Uruguay" (Children's Defense Fund 1990, p. 6, cited in Aulette 1994, p. 405). (The famine experienced within North Korea during the 1990s likely would remove that country from this listing.)

The United States is unique among western societies in its lack of a national health care program, which helps to explain its higher infant mortality rate. In 1997, 43.4 million Americans, or slightly over 16 percent of the U.S. population, had no health insurance coverage for the entirety of that calendar year. The lack of health insurance is especially acute among the poor. Although the Medicaid program is intended to provide health coverage to the poor, 11.2 million poor Americans, or one-third of all poor people in the United States, had no health coverage in 1997 (Bennefield 1998).

In the absence of concerted social policy measures, infant mortality and other risks faced by poor Americans are unlikely to diminish in the near future. Using the Gini index (a measure of income concentration), the U.S. Bureau of the Census reported that income inequalities within the country increased by 16 percent between 1968 and 1992. Even greater inequalities have developed since 1992: Between 1968 and 1994 the rate of increase in U.S. income inequality was over 22 percent (Weinberg 1996). At the same time, the percentage of Americans with incomes below the poverty line also increased—from 11.7 percent in 1979 to 13.3 percent in 1997 (Dalaker and Naifeh 1997).

AMERICAN FAMILIES AND THE FUTURE

Traditional distinctions between "family" and nonfamily" are increasingly challenged. Though still a relatively small proportion of all households (about 4 percent; U.S. Bureau of the Census 1998), the number of cohabiting heterosexual couples has increased greatly in the past several decades. Marriage between homosexuals is not legally recognized, but some have elected to adopt each other legally, and a growing number are raising children. In addition to the "traditional" nuclear family form of two parents with children, other family types can be expected to continue in the future. These include single parents, blended families resulting from remarriage, and households in which other relatives such as grandparents reside.

Increased longevity has brought about some of the most important changes in American family life over the past century. Children are more likely than ever before to interact with their grandparents. Further, many persons are becoming grandparents while their own parents are still alive (Uhlenberg 1986). Research has documented the prevalence and importance of social interaction, emotional support, financial help, and other assistance between the generations. For all types of American families, indications are that high levels of interaction and assistance between the generations will continue in the future. It remains to be seen how U.S. social policy will respond to the needs of family care-providers, and to the needs of poor Americans and their families, in the years to come.

references

Amato, Paul R., and Alan Booth 1991 "Consequences of Parental Divorce and Marital Unhappiness for Adult Well-Being." Social Forces 69:895–914.

Astone, Nan Marie, and Sara S. McLanahan 1991 "Family Structure, Parental Practices and High School Completion." American Sociological Review 56:309–320.

Aulette, Judy Root 1994 Changing Families. Belmont, Calif.: Wadsworth.

Bennefield, Robert L. 1998 Health Insurance Coverage:1997. U.S. Bureau of the Census, Current Population Reports, Series P60–202. Washington, D.C.: U.S. Government Printing Office.

Boris, Eileen, and Peter Bardaglio 1987 "Gender, Race, and Class: The Impact of the State on the Family and the Economy, 1790–1945." In Naomi Gerstel and Harried Engel Gross, eds., Families and Work. Philadelphia: Temple University Press.

Bumpass, Larry, James A. Sweet, and Teresa Castro-Martin 1990 "Changing Patterns of Remarriage." Journal of Marriage and the Family 52:747–756.

Casper, Lynn M., and Ken Bryson 1998 Household andFamily Characteristics: March 1998 (Update). U.S. Bureau of the Census, Current Population Reports, Series P20–515. Washington, D.C.: U.S. Government Printing Office.

Cherlin, Andrew J. 1996 Public and Private Families. New York: McGraw-Hill.

——1992 Marriage, Divorce, Remarriage, rev. ed. Cambridge, Mass.: Harvard University Press.

——, Frank F. Furstenberg, Jr., P. Lindsay Chase-Lansdale, Kathleen E. Kiernan, Philip K. Robins, Donna Ruane Morrison, and Julien O. Teitler 1991 "Longitudinal Studies of Effects of Divorce on Children in Great Britain and the United States." Science 252:1386–1389.

Collins, Randall 1988 Sociology of Marriage and the Family: Gender, Love, and Property, 2d ed. Chicago: Nelson-Hall.

Dalaker, Joseph, and Mary Naifeh 1998 Poverty in theUnited States: 1997. U.S. Bureau of the Census, Current Population Reports, Series P60–201. Washington, D.C.: U.S. Government Printing Office.

Demos, John 1970 A Little Commonwealth. New York: Oxford University Press.

Driver, Harold E. 1969 Indians of North America, 2d ed. Chicago: University of Chicago Press.

Dwyer, Jeffrey W., and Raymond T. Coward 1992 "Gender, Family, and Long-Term Care of the Elderly." In Jeffrey W. Dwyer and Raymond T. Coward, eds., Gender, Families, and Elder Care. Newbury Park, Calif.: Sage.

Furstenberg, Frank F., Jr., and Andrew J. Cherlin 1991 Divided Families: What Happens to Children When Parents Part? Cambridge, Mass.: Harvard University Press.

Genovese, Eugene D. 1986 "The Myth of the Absent Family." In Robert Staples, ed., The Black Family:Essays and Studies, 3d ed. Belmont, Calif.: Wadsworth.

Gerstel, Naomi, and Sally K. Gallagher 1993 "Kinkeeping and Distress: Gender, Recipients of Care, and Work-Family Conflict." Journal of Marriage and the Family 55:598–607.

Gorey, Kevin M., Robert W. Rice, and Gary C. Brice 1992 "The Prevalence of Elder Care Responsibilities Among the Work Force Population." Research onAging 14(3):399–418.

Grambs, Jean Dresden 1989 Women over Forty: Visionsand Realities. New York: Springer.

Gutman, Herbert 1976 The Black Family in Slavery andFreedom, 1750–1925. New York: Pantheon.

Hayes, Cheryl D., John L. Palmer, and Martha J. Zaslow (eds.) 1990 Who Cares for America's Children? ChildCare Policy for the 1990s. Washington, D.C.: National Academy Press.

Howe, Louise Knapp 1972 The Future of the Family. New York: Simon and Schuster.

Jacobson, Paul H. 1959 American Marriage and Divorce. New York: Rinehart.

Kain, Edward 1990 The Myth of Family Decline: Understanding Families in a World of Rapid Social Change. Lexington, Mass.: Heath.

Lasch, Christopher 1977 Haven in a Heartless World: TheFamily Besieged. New York: Basic Books.

Laslett, Barbara 1973 "The Family as a Public and Private Institution: An Historical Perspective." Journal of Marriage and the Family 35:480–494.

Laslett, Peter, and Richard A. Wall (eds.) 1972 Householdand Family in Past Time. New York: Cambridge University Press.

Lechner, Viola M. 1992 "Predicting Future Commitment to Care for Frail Parents Among Employed Caregivers." Journal of Gerontological Social Work 18:69–84.

Lee, Sharon M., and Keiko Yamanaka 1990 "Patterns of Asian American Intermarriage and Marital Assimilation." Journal of Comparative Family Studies 21:287–305.

Leslie, Gerald R., and Sheila K. Korman 1989 The Familyin Social Context, 7th ed. New York: Oxford University Press.

Levitan, Sar A., Richard S. Belous, and Frank Gallo 1988 What's Happening to the American Family? Tensions,Hopes, Realities. Baltimore, Md.: Johns Hopkins University Press.

Lugaila, Terry A. 1998 Marital Status and Living Arrangements: March 1998 (Update). U.S. Bureau of the Census, Current Population Reports, Series P20–514. Washington, D.C.: U.S. Government Printing Office.

Lurie, Nancy Oestreich 1985 "The American Indian: Historical Background." In N.R. Yetman, ed., Majority and Minority, 4th ed. Boston: Allyn and Bacon.

Marden, Charles F., and Gladys Meyer 1973 Minorities inAmerican Society, 4th ed. New York: D. Van Nostrand.

Marger, Martin N. 1991 Race and Ethnic Relations: American and Global Perspectives. Belmont, Calif.: Wadsworth.

McLanahan, Sara, and Lynne Casper 1998 "Growing Diversity and Inequality in the American Family." In Andrew J. Cherlin, ed., Public and Private Families. Boston: McGraw-Hill.

Morgan, Leslie, and Suzanne Kunkel 1998 Aging: TheSocial Context. Thousand Oaks, Calif.: Pine Forge Press.

Moynihan, Daniel P. 1965 The Negro Family: The Case forNational Action. Washington, D.C.: Office of Policy Planning and Research, U.S. Department of Labor.

Nock, Steven L. 1987 Sociology of the Family. Englewood Cliffs, N.J.: Prentice Hall.

Ollenburger, Jane C., and Helen A. Moore 1998 ASociology of Women: the Intersection of Patriarchy, Capitalism, and Colonization, 2d ed. Upper Saddle River, N.J.: Prentice Hall.

Olson, David H., and John DeFrain 1994 Marriage andthe Family: Diversity and Strengths. Mountain View, Calif.: Mayfield.

Peterson, Richard R. 1996 "A Re-Evaluation of the Economic Consequences of Divorce." American Sociological Review 61:528–536.

Richards, Leslie N., and Cynthia J. Schmiege 1993 "Problems and Strengths of Single-Parent Families." Family Relations 42:277–285.

Saluter, Arlene F. 1996 Marital Status and Living Arrangements: March 1994. Current Population Reports, Series P20–484. Washington, D.C.: U.S. Government Printing Office.

Staples, Robert (ed.) 1986 The Black Family: Essays andStudies, 3d ed. Belmont, Calif.: Wadsworth.

Stone, Robyn, Gail Lee Cafferata, and Judith Sangl 1987 "Caregivers of the Frail Elderly: A National Profile." The Gerontologist 27:616–626.

Sweet, James A., and Larry L. Bumpass 1987 AmericanFamilies and Households. New York: Russell Sage.

Uhlenberg, Andrew 1986 "Death and the Family." In A.S. Skolnick and J.H. Skolnick, eds., Family in Transition. Boston: Little, Brown.

U.S. Bureau of the Census 1999 Interracial MarriedCouples: 1960 to Present. http:www.census.gov/population/socdemo/ms-la/tabms-3.txt

U.S. Bureau of the Census 1998 Statistical Abstract of theUnited States, 1998. Washington, D.C.: U.S. Government Printing Office.

U.S. Bureau of the Census 1996 65+ in the United States. Current Population Reports, Special Studies, Series P–23, No. 190. Washington, D.C.: U.S. Government Printing Office.

U.S. Bureau of the Census 1995 Population Profile of theUnited States: 1995. Current Population Reports, Series P23–189. Washington, D.C.: U.S. Government Printing Office.

U.S. Bureau of Labor Statistics 1991 Employment andEarnings. Washington, D.C.: U.S. Government Printing Office.

Weinberg, Daniel H. 1996 A Brief Look at U.S. IncomeInequality. U.S. Bureau of the Census, Current Population Reports, Series P60–191. Washington, D.C.: U.S. Government Printing Office.

Wolf, Douglas A., Vicki Freedman, and Beth J. Soldo 1997 "The Division of Family Labor: Care for Elderly Parents." Journals of Gerontology 52B:102–109.

Zinn, Maxine Baca, and D. Stanley Eitzen 1987 Diversityin American Families. New York: Harper and Row.

LAURIE RUSSELL HATCH