The last decades of the 20th century were a period of significant change in family life in the United States. Among the well-documented changes are a rising age at marriage, an increase in cohabitation, and a dramatic shift in the proportion of children born outside marriage (Bramlett and Mosher, 2002; Casper and Bianchi, 2002; Wu and Wolfe, 2001). Coupled with a high divorce rate, these trends have led to high rates of female family headship and a growing share of children with restricted access to their fathers’ resources.

These changes in family patterns have taken place alongside rapid growth in immigration and concomitant changes in the racial and ethnic composition of the U.S. population. The average annual inflow of immigrants more than doubled between the 1970s and 1990s, and the share of immigrants from Latin America increased at the same time (Martin and Midgley, 2003). Thus, the Hispanic population grew from 5 percent of the total U.S. population in 1970 to 13 percent in 2000. Furthermore, population projections suggest that Hispanics will comprise 20 percent of the U.S. population in 2030 (National Research Council, 1997).

This chapter addresses the intersection of these two domains of rapidly changing demographic behavior. Specifically, we analyze the family patterns of Hispanics, focusing on several key issues. First, to place the present in a larger context, we document trends in several indicators of family change. Comparisons between Hispanic subgroups, non-Hispanic whites, and non-Hispanic blacks provide information on the extent to which Hispanics have shared in the general shifts in family configurations that took place during the past several decades. This issue is fundamental to understanding the nature of family life among Hispanics as well as links between changing family processes and family members’ access to social and economic resources. As noted by Vega (1995, p. 6), “Changing family structures, including marital disruption and cohabitation, could represent the most important issue for Latino family theory and research in the decade ahead.”

A second issue addressed in the chapter is generational variation in family patterns within Hispanic subgroups. Our descriptive analyses demonstrate that Hispanics—like other racial/ethnic groups—exhibit many behaviors that are consistent with what some scholars call “family decline” (Popenoe, 1993). At the same time, Hispanics (especially Mexican Americans) are typically described as oriented toward family well-being, rather than individual well-being (Sabogal, Marin, Otero-Sabogal, VanOss Marin, and Perez-Stable, 1987; Valenzuela and Dornbusch, 1994; Vega, 1995). To the extent that such “familism” remains alive among U.S. Hispanics, one would expect it to reduce the erosion of traditional family patterns or to contribute to new family forms in which family support remains high. However, it is possible that the process of assimilation reduces familism and encourages the individualism that some have argued is at the heart of recent changes in family behavior. After describing racial/ethnic differences in the characteristics of family households and the living arrangements of individuals of various ages, we focus on differences within Hispanic groups by generational status. Our comparisons of the family patterns of the first generation (foreign-born), the second generation (native-born of foreign parentage), and the third or higher generations (native-born of native parentage) will shed light on the dynamics of assimilation with respect to family patterns.

A third topic considered in the chapter is racial/ethnic mixing in sexual partnerships of various types, including marriage, cohabitation, and parenthood. Intermarriage is a long-standing theme in the study of assimilation. It has been considered both an indicator of assimilation and a means by which assimilation is achieved (Gordon, 1964; Lieberson and Waters, 1988). According to the classic assimilation theory, intermarriage between an immigrant group and the dominant population reduces social boundaries and eventually leads to a reduction in the salience of an ethnic identity. Because the offspring of intermarried couples may opt out of defining themselves as members of an ethnic group, intermarriage may affect the future size and shape of an ethnic population. Among Hispanics, intermarriage with non-Hispanic whites or non-Hispanic blacks may ultimately lead to a blurring of racial/ethnic boundaries. At the same time, intermarriage between members of different Hispanic subgroups may strengthen pan-ethnicity, or the adoption of a “Hispanic” identity instead of an identity as a member of a specific national-origin group. While recognizing the importance of intermarriage, we contend that in the current era of what is called the “retreat from marriage,” the study of racial/ethnic mixing in sexual partnerships must be expanded to include unions other than traditional marriages. Thus, we examine ethnic endogamy and exogamy among Hispanics in both marriage and cohabitation. Given the growing separation of marriage and childbearing, we also examine racial/ethnic mixing in both marital and nonmarital childbearing.

It is now widely recognized that Hispanic national-origin groups differ markedly with respect to their histories of immigration, settlement patterns, socioeconomic position, and other circumstances (Bean and Tienda, 1987; Oropesa and Landale, 1997; Portes and Rumbaut, 2001). There is a general consensus among experts on the Hispanic population that, to the extent possible, research should disaggregate the generic category “Hispanic” into specific national-origin groups. Thus, all of our analyses present information separately for Mexicans, Cubans, Puerto Ricans, Central/South Americans, and other Hispanics.1 In addition to addressing differences between Hispanics and non-Hispanics, we examine the diversity of family patterns among the specific Hispanic groups.

Several broad conclusions are supported by our analyses. First, Hispanics exhibit high levels of familism relative to non-Hispanics on a variety of structural/demographic indicators. However, they are also participating in the general changes in family life that are under way in the United States. Second, analyses conducted separately by national origin suggest declining familism across generations (with some exceptions). Third, all Hispanic subgroups exhibit substantial declines in ethnic endogamy across generations. This pattern suggests that assimilation is occurring and that racial/ethnic boundaries for Hispanics are not sharp. Nonetheless, the Mexican-origin population stands out for its high levels of ethnic endogamy in marriage, cohabitation, and parenthood.


One of the most significant changes in family behavior that occurred during the past several decades is the retreat from marriage. Although most individuals marry eventually, a declining percentage of men and women are entering marriage in their teens and early 20s (Ventura and Bachrach, 2000). At the same time, most young people begin having sex in their mid-to late teens (Alan Guttmacher Institute, 1999), and cohabitation has become so widespread that it has largely offset the decline in marriage (Bumpass and Lu, 2000). Thus, the process of union formation has changed substantially. In addition, divorce rates remain high, although they have declined slightly since their peak around 1980 (Casper and Bianchi, 2002). The growing proportion of women who are unmarried (but sexually active and often cohabiting), increasing birth rates among unmarried women, and decreasing birth rates among married women have all contributed to a striking increase in the proportion of births occurring outside marriage (Wu et al., 2001).

summarizes information on trends in several family-related behaviors from 1980 to 2000. The top panel shows the percentage married among females ages 20 to 24. At each time point, Mexican-origin females were the most likely to be married and non-Hispanic black females were the least likely to be married. For example, in 1980 roughly half of Mexican females ages 20 to 24 were married compared with one-fourth of their non-Hispanic black counterparts. The figures for non-Hispanic whites (45 percent), Cubans (40 percent), and Puerto Ricans (38 percent) are intermediate between those for Mexicans and non-Hispanic blacks. Between 1980 and 2000, there was a marked decline in early marriage for each of the racial/ethnic groups shown. However, the percentage change in the percentage married was weaker for Mexican women (−20 percent) than for Cubans (−31 percent), Puerto Ricans (−37 percent), non-Hispanic whites (−39 percent), and non-Hispanic blacks (−44 percent).2 Thus, while all groups have shared in the retreat from early marriage, young Mexican women are more likely to enter marriage by their early 20s than the other Hispanic and non-Hispanic groups.


Selected Indicators of Family Change, by Race and Ethnicity .

The second through fourth panels of the table focus on various aspects of fertility. The total fertility rates (TFRs) presented in the second panel describe the number of children the typical woman in a particular racial/ethnic group would have if her fertility throughout her reproductive period reflected the prevailing age-specific fertility rates for the racial/ethnic group at a given point in time. In 1980, the TFR for each Hispanic subgroup except Cubans was higher than that for non-Hispanic whites (1.7), but only Mexicans exhibited substantially higher fertility (TFR = 2.9). The TFRs for Puerto Ricans and other Hispanics (both 2.1) were slightly higher than the non-Hispanic white rate (1.7), but slightly lower than the non-Hispanic black rate (2.4).

Despite the long-term trend toward lower fertility, the TFR increased between 1980 and 2000 for all groups except non-Hispanic blacks. The TFR rose by 11 percent for non-Hispanic whites (from 1.7 to 1.9), 13 percent for Mexicans (from 2.9 to 3.3), 26 percent for Puerto Ricans (from 2.1 to 2.6), and 44 percent for Cubans (from 1.3 to 1.9) and other Hispanics (from 2.1 to 3.0). The generally greater increase in fertility among Hispanics compared with non-Hispanic whites resulted in more diversity in fertility in 2000 than in 1980. Currently, the average Mexican, Puerto Rican, and other Hispanic woman can expect to have about one more child than the average non-Hispanic white woman.3 Cubans are an exception, with a TFR that is nearly identical to that of whites. The TFRs for all Hispanic groups except Cubans also exceed that for non-Hispanic blacks.

The third panel presents figures on nonmarital childbearing. In 1980, the percentage of births to unmarried women was more than twice as high for each Hispanic subgroup (except Cubans) as it was for non-Hispanic whites (10 percent). The figures range from 20 percent for Mexicans to 46 percent for Puerto Ricans. Over the subsequent 20 years, all groups experienced a substantial increase in nonmarital childbearing. The percentage of births to unmarried women more than doubled for non-Hispanic whites (percentage change of 134 percent), Mexicans (101 percent), and Cubans (173 percent), and increased by more than half for Central/South Americans (64 percent) and other Hispanics (97 percent). The two groups that showed less growth over the 20-year period (Puerto Ricans and non-Hispanic blacks) had relatively high shares of nonmarital births at the first point in time (46 and 57 percent, respectively). Overall, these figures indicate that each Hispanic subgroup has experienced the trend toward nonmarital childbearing that has been documented for the general U.S. population. Nonetheless, there remain substantial racial/ethnic differences in the percentage of births to unmarried mothers in 2000. Non-Hispanic whites (22 percent) and non-Hispanic blacks (69 percent) fall at the two extremes of the distribution. While Cubans are closer to non-Hispanic whites (27 percent) and Puerto Ricans are closer to non-Hispanic blacks (59 percent), Mexicans (41 percent) and Central/South Americans (44 percent) are equidistant from the extremes.4

The fourth panel sheds light on differences and similarities in the timing of entry into motherhood across the groups. In 1980, less than 5 percent of births to non-Hispanic whites, Cubans, and Central/South Americans were to women under 18 years of age. The figures were slightly higher for other Hispanics (7 percent) and Mexicans (8 percent), and substantially higher for Puerto Ricans (10 percent) and non-Hispanic blacks (13 percent). Consistent with the well-established decline in teenage childbearing in the United States, the trend from 1980 to 2000 shows a substantial decrease in the percentage of births to young teen mothers for almost all groups. However, the decline has not been as great for most Hispanic subgroups as it has been for non-Hispanic whites and non-Hispanic blacks. In 2000, Mexican (6 percent), Puerto Rican (7 percent), and other Hispanic (7 percent) infants were more likely than Cuban (3 percent) and Central/South American (3 percent) infants to have a teenage mother. The figures for the former groups are more similar to that for non-Hispanic blacks (7 percent), while those for the latter are similar to that for non-Hispanic whites (2 percent).

The last panel of the table focuses on the structure of family households. Available data for 1980 show that whites (12 percent) and Mexicans (15 percent) had relatively low levels of female family headship, but Puerto Ricans (38 percent) and non-Hispanic blacks (40 percent) had substantially higher levels. An increase in the percentage of female householders is evident for three of the four groups for which complete data are shown (non-Hispanic whites, non-Hispanic blacks, and Mexicans). Puerto Ricans are the exception, showing a slight decline in the percentatge of family households with a female head over the two-decade period. In 2000, the various Hispanic subgroups fall between the extremes occupied by non-Hispanic whites and blacks with respect to family structure. About 14 percent of white families had a female householder, compared with about 20 percent of Mexican and Cuban families, 25 percent of Central and South American families, 36 percent of Puerto Rican families, and 45 percent of non-Hispanic black families.

In summary, shows that trends for each dimension of family life are generally similar for Hispanic subgroups and the non-Hispanic majority. However, consistent with differences in their histories and social locations (see Chapter 2), there are substantial differences across Hispanic subgroups—and between Hispanic subgroups and non-Hispanics—in specific aspects of family behavior. Moreover, there are a few instances of divergence (i.e., widening of group differences) over time between Hispanic and non-Hispanic groups. For example, the 1980–2000 increase in fertility (as measured by the TFR) was somewhat greater for Hispanic groups than for non-Hispanic whites. In addition, there was a weaker decline in teenage childbearing among Hispanics compared with non-Hispanics. The growing divergence between Hispanic and non-Hispanic fertility patterns is undoubtedly linked to the relatively rapid growth of the immigrant population (Suro and Passel, 2003). Since Latin American immigrants have higher fertility and tend to bear their children earlier than native-born Hispanics, a shift in the generational composition of the Hispanic population would contribute to such a pattern. Also noteworthy is the considerably greater increase in female family headship among Mexican Americans compared with non-Hispanic whites and blacks.


Recent scholarship on current family patterns among Hispanics emphasizes several distinct themes, which can be broadly classified as stressing either the structural conditions in which Hispanics live or the role of culture in shaping values and behavior. We discuss each in turn.

The Role of Structural Conditions

One recurrent theme in the study of Hispanic families is the impact of socioeconomic disadvantage on family life (Baca Zinn and Wells, 2000; Massey, Zambrana, and Bell, 1995; Oropesa and Landale, 2004; Vega, 1995). Due to a complex set of factors, including the hardships of immigration, low levels of human capital, racial discrimination, and settlement patterns, Hispanic poverty rates remain high. In 2002, about 22 percent of Hispanics were poor, a figure roughly comparable to that for blacks (24 percent) and almost three times that for non-Hispanic whites (8 percent) (Proktor and Dallaker, 2003).5 A constellation of behaviors and conditions that are associated with poverty, especially low skill levels, job instability, and inadequate earnings for males, play a central role in recent explanations of the retreat from marriage, nonmarital childbearing, and female family headship (Oppenheimer, 2000; Sweeney, 2002; Wilson, 1987). Contemporary scholarship on Hispanic families is highly critical of a “culture of poverty” interpretation of the link between poverty and family patterns. Rather, it emphasizes a “social adaptation” paradigm, in which individuals and families adapt to the situations they face as a result of their social and economic position in U.S. society (Baca Zinn and Wells, 2000; Vega, 1995).

An issue that has received attention is whether links between poverty and family processes among Hispanics can be understood using frameworks developed to study the experience of other disadvantaged groups (i.e., blacks). Massey et al. (1995) argue that the Hispanic experience is fundamentally different from that of blacks in five important ways. First, consistent with Bean and Tienda’s seminal work (1987), they contend that Hispanics cannot be understood as a single group; analyses must be conducted separately for each Hispanic subgroup because of differences in their histories and current situations. Second, Hispanics are heterogeneous with respect to race, while blacks are relatively homogeneous. Furthermore, foreign-born Hispanics experience a marked disjuncture between the way race is viewed in Latin America and the racial dynamics they encounter in the United States. Third, related to their diverse racial features, Hispanics experience more varied levels of segregation (and consequently, more varied opportunities) than do non-Hispanic blacks, but this is changing. Fourth, the Hispanic experience remains bound up with immigration. Massey et al. (1995) argue that the dynamics of immigration must be explicitly considered in studies of Hispanic family patterns. This requires attention to the complexities of international migration (e.g., selective migration) as well as consideration of issues related to the assimilation process. Finally, Hispanics differ from blacks in that their experience is influenced by their use of the Spanish language. Given these differences, Massey and colleagues argue that studies of Hispanic families cannot simply adopt theories developed to explain the experience of other disadvantaged groups. Although socioeconomic disadvantage is central to the Hispanic experience, its effects on family patterns must be understood in the context of more complex frameworks that simultaneously consider the aforementioned issues.

The Role of Culture

Another theme that is widespread in studies of Hispanic families is the idea that Hispanics are characterized by familism or a strong commitment to family life that is qualitatively distinct from that of non-Hispanic whites (Vega, 1995). The concept of familism can be found in the sociological literature as early as the mid-1940s (Burgess and Locke, 1945; Ch’Eng-K’Un, 1944). Although it has been used in somewhat varied ways since that time, there is general agreement that familism entails the subordination of individual interests to those of the family group. Some authors have stressed the attitudinal foundations of familism (Bean, Curtis, and Marcum, 1977; Burgess and Locke, 1945; Gaines et al., 1997; Lesthaeghe and Meekers, 1986; Rodriguez, Kosloski, and Kosloski, 1998; Oropesa and Gorman, 2000), while others have emphasized behavioral manifestations (Tienda, 1980; Winch, Greer, and Blumberg, 1967). Recent scholarship puts forth the view that familism is a multidimensional concept encompassing at least three features: a structural/demographic dimension,6 a behavioral dimension, and an attitudinal dimension (Valenzuela and Dornbusch, 1994). The structural dimension is evident in such family configurations as family size, family structure (including the presence or absence of nuclear and extended kin), and fertility patterns. The behavioral dimension includes behaviors that indicate the fulfillment of family role obligations, such as the sharing of economic resources, mutual assistance and social support, and frequent contact among family members. The attitudinal (or normative) dimension entails values that emphasize the importance of the family and prescribe loyalty, reciprocity, and solidarity among family members (Sabogal et al., 1987; Steidel, Contreras, and Contreras, 2003).

Early scholarship often regarded familism as an impediment to socioeconomic advancement in urban industrial societies because such societies emphasize individualism, competition, and geographic mobility. For example, some studies argued that familism hindered the socioeconomic success of Mexican Americans (Valenzuela and Dornbusch, 1994). More recently, however, this view has been turned on its head and familism is generally viewed as a protective factor. Studies of a variety of outcomes (e.g., physical and mental health, education) among Hispanics propose that extended family networks, family cohesion, and high levels of social support reduce the adverse consequences of poverty (Guendelman, 1995; Landale and Oropesa, 2001; Rumbaut and Weeks, 1996; Sabogal et al., 1987; Zambrana, Scrimshaw, Collins, and Dunkel-Schetter, 1997). Thus, recent scholarship regards familism as a positive attribute of Hispanic families that may decline with acculturation to U.S. family norms and adaptation to life in the United States.

Although a comprehensive assessment of the three dimensions of familism is beyond the scope of this chapter, we focus on the structural dimension in Tables through . Based on weighted data from the 1998–2002 March Current Population Surveys (pooled across years), we provide descriptive information on the characteristics of Hispanic families and the living arrangements of individuals in different age groups. Comparisons are made across racial/ethnic groups and within Hispanic subgroups by generational status.7


Percentage Family Households by Race/Ethnicity and Generational Status of Householder.


Living Arrangements by Generation, Mexican Children, and Elderly Persons .

Characteristics of Family Households

addresses a fundamental question: What percentage of all households are family households? The U.S. Census Bureau defines a family household as a household maintained by a householder who is in a family; a family is a group of two or more people (one of whom is the householder) who are related by birth, marriage, or adoption and reside together (U.S. Census Bureau, 2000).8 It is important to note that the Census Bureau does not regard cohabitation as a family status. Given the growing role of cohabitation in U.S. family life (Bramlett and Mosher, 2002; Bumpass and Lu, 2000) and its prominence among some Hispanic subgroups, we believe it is important to recognize cohabiting unions. Thus, we depart from the Census Bureau’s definition of a family household by treating cohabitation as a family status. Households in which the householder is cohabiting with a partner are therefore included as family households in Tables and .9


Characteristics of Family Households by Race/Ethnicity and Generation of Householder.

The top panel of presents unadjusted percentages for all households and for households broken down by the generational status of the householder. Because the propensity to live in family versus nonfamily households varies by age, we also present comparable information standardized for the age of the householder. The age-standardized percentages are especially important for comparisons between Hispanic subgroups and non-Hispanic whites, since the former are relatively young populations.

Both the unstandardized and age-standardized percentages for all households (i.e., not disaggregated by generation) show that all Hispanic subgroups are more likely to reside in family households than are non-Hispanic whites and non-Hispanic blacks. The age-standardized percentages for Hispanic groups range from 72 percent (Puerto Ricans) to 82 percent (Mexicans), while those for non-Hispanic whites and non-Hispanic blacks are 69 and 66 percent, respectively.10 This is consistent with the thesis of relatively high levels of familism among Hispanics, especially Mexican Americans. Focusing on within-group differences by generation, the age-standardized pattern is similar for Mexicans, Puerto Ricans, and Central/South Americans: households in which the householder is foreign-born are more likely to be family households than those in which the householder is native-born (of native or foreign parentage). For example, 84 percent of households headed by a first-generation Mexican are family households, compared with 81 percent of households headed by a second-generation Mexican and 78 percent of households headed by a Mexican in the third (or higher) generation. Although the pattern for Cubans is not linear, households in which the householder is third (or higher) generation are the least likely to be family households.

provides information on various structural characteristics of family households. We distinguish between married-couple households, cohabiting-couple households, and households with a female householder who does not live with a partner.11 The figures for all family households (i.e., not disaggregated by generation) show considerable variation across Hispanic subgroups in household type. Cuban and Mexican households are the most likely to be headed by a married couple (75 and 69 percent, respectively, compared with 79 percent for non-Hispanic whites) and the least likely to be headed by a female with no spouse or partner present (16 and 18 percent, respectively, compared with 11 percent for non-Hispanic whites). Puerto Ricans represent the other extreme: 53 percent of Puerto Rican family households are headed by a married couple and 34 percent are headed by a female with no spouse or partner present. Cohabitation is the least common arrangement shown, but it is significant for all groups. About 6 to 7 percent of Hispanic family householders in all subgroups except Cubans (4 percent) live with a cohabiting partner. These percentages are slightly higher than that for non-Hispanic whites (5 percent) and roughly comparable to that for non-Hispanic blacks (6 percent).

Other noteworthy group differences for all family households are the slightly larger household size and the greater prevalence of extended families12 among Hispanics, relative to non-Hispanic whites. With respect to the latter, about 6 to 10 percent of family households in each Hispanic subgroup are extended, compared with 3 percent of non-Hispanic white family households. The figure for non-Hispanic blacks (7 percent) is comparable to those presented for the Hispanic groups.13

As noted earlier, there are two major explanations for differences in family patterns between Hispanic subgroups and the comparison groups (non-Hispanic whites and non-Hispanic blacks). One explanation points to differences in the structural positions of the groups, especially the disadvantaged socioeconomic status of some Hispanic subgroups (and non-Hispanic blacks) relative to non-Hispanic whites. The other emphasizes cultural orientations and values vis-à-vis the family. Evaluation of these perspectives is complex and beyond the scope of the present study; however, to provide some information on the role of structural characteristics, we standardized the educational distributions of the groups being compared. Specifically, using direct standardization, we calculated what the family characteristics of each group would be if the educational distribution of its householders was the same as that of non-Hispanic white householders.14 With education controlled, similar patterns were evident, although differences were attenuated (results not shown). For example, the percentage of family households with a female householder was 15 percent for Cubans, 17 percent for Mexicans, and 29 percent for Puerto Ricans in the standardized analysis, compared with 11 percent for non-Hispanic whites. In the unstandardized analysis, it was 16 percent for Cubans, 18 percent for Mexicans, and 34 percent for Puerto Ricans.

also shows differences in family household characteristics by the generational status of the householder. Although there are some inconsistencies across national-origin groups, the pattern for several Hispanic subgroups suggests declining familism across generations. For example, among Mexicans, foreign-born householders are more likely to be married and less likely to cohabit or to be female family heads than their native-born counterparts. Among the foreign-born, 72 percent are married, 5 percent are cohabiting, and 15 percent are single female householders; the comparable figures for the native-born of native parentage are 65 percent married, 7 percent cohabiting, and 22 percent single female householders. In addition, the mean household size and the percentage of extended family households are higher among foreign-born Mexicans than native-born Mexicans. For example, among the foreign-born, 10 percent of households are extended, compared with 7 percent among the native-born of native parentage. Similar generational patterns are found among Puerto Ricans and Central/South Americans, except that family size does not vary by generation for Puerto Ricans. However, there are irregular or opposite patterns for Cubans and other Hispanics. When the educational distribution of household heads is standardized (each generation of each Hispanic subgroup given the educational distribution of the total non-Hispanic white population), the generational patterns remain unchanged (results not shown).

Living Arrangements

The structure and composition of households are experienced by individuals in different ways as they move through the life course. Thus we summarize in the living arrangements of individuals in four broad age groups (0–17; 18–24; 25–64; 65+). Some of the largest differences in living arrangements by race and ethnicity are found for children. Among Hispanics, the percentage living with both parents ranges from 42 percent for Puerto Ricans to 69 percent for Cubans (with the figures for Mexicans and Central/South Americans about 67 percent). Again, the figures for Hispanics fall between the extremes represented by the experience of non-Hispanic whites (77 percent) and non-Hispanic blacks (37 percent), although Hispanics are generally closer to whites. As one would expect, Puerto Rican (46 percent) and non-Hispanic black children (49 percent) are the most likely to live in a mother-only family. Both groups are more than twice as likely to live in such a family arrangement as non-Hispanic white, Mexican, Cuban, and Central/South American children.15


Living Arrangements by Age and Ethnicity .

There is less racial and ethnic variation in living arrangements in early adulthood (18 to 24) and the middle adult years (25 to 64). However, several group differences are noteworthy. In early adulthood, Cubans stand out for their comparatively low rates of household headship and high propensity to remain in the parental home. Fully 62 percent of Cubans ages 18 to 24 live in their parent’s household, compared with less than 50 percent for all other Hispanic groups. This living arrangement may facilitate the relatively high levels of education attained by Cubans in young adulthood. Also noteworthy are the considerably greater shares of Hispanic and black young adults living with “other relatives,” compared with white young adults. This pattern carries over to middle adulthood (ages 25 to 64), and in fact is one of the major ways in which living arrangements vary by race and ethnicity during the middle adult years. For example, while only 2 percent of non-Hispanic whites ages 25 to 64 live with other relatives, fully 10 percent of Mexicans and 12 percent of Central/South Americans do so. Doubling up with relatives may be an economic strategy that is employed under conditions of economic disadvantage.

Among the elderly (ages 65+), the most striking differences in living arrangements are between Hispanics and non-Hispanic whites, rather than among Hispanic subgroups. In particular, Hispanics are considerably more likely to live with other relatives and less likely to live alone than are non-Hispanic whites. For example, only 5 percent of non-Hispanic whites live with other relatives, compared with 19 percent of Mexicans and Cubans, 15 percent of Puerto Ricans, and 33 percent of Central/South Americans. These differences undoubtedly reflect both differences in economic resources and cultural preferences regarding the care of the elderly.

Information on living arrangements by race/ethnicity and generational status for each age group cannot be presented, given space constraints. However, in we provide data for Mexican Americans on generational differences in living arrangements among children and the elderly. The top panel shows a striking difference between children with foreign-born parents (first- and second-generation children) and children with native-born parents. Children in the former groups are much more likely to live with both parents (72–73 percent) than children in the latter group (56 percent). About 17 percent of first-generation children live with only one parent (14 percent with mother and 3 percent with father), compared with 24 percent of second-generation children and 37 percent of native-born children with native-born parents. Thus, children of the foreign-born experience greater parental union stability than children of the native-born.

The situation of Mexican American elderly persons also varies by generation. First, foreign-born elderly persons are less likely to be the householder or the spouse or partner of the householder (54 percent) than the native-born of foreign parentage (69 percent) or the native-born of native parentage (63 percent). They are also less likely to live alone (15 percent, compared with about 20–21 percent for the native-born groups). Instead, the foreign-born are considerably more likely to live with other relatives (30 percent), such as their children, than the native-born of foreign percentage (9 percent) and native parentage (14 percent).

Overall, Hispanics exhibit higher levels of familism than non-Hispanics on most of the structural indicators examined. A notable exception is female family headship, which is considerably more prevalent in all Hispanic subgroups than among non-Hispanic whites. At the same time, there is considerable diversity in the family characteristics of Hispanics by both national origin and generation. Although the findings are not entirely consistent across Hispanic groups, within-group generational differences generally suggest declining familism across generations. This is especially the case for Mexican Americans, a group that exhibits lower levels of family-oriented behavior on every indicator among the native-born compared with the foreign-born.


As is common practice in social demographic research, our analysis to this point has assumed that racial/ethnic categories are fixed and reflect unambiguous distinctions among individuals. However, the social construction of race and ethnicity—and the complexities involved in racial and ethnic identities—are increasingly emphasized by contemporary social scientists. The dominant view is that racial and ethnic categories reflect shared social meanings, rather than biological differences between groups, and that social interpretations of the categories are tied to long-standing power differentials (Waters, 2002). In addition, the fluidity of racial and ethnic identities across situations, over time, and across generations is stressed.

One important factor in the fluidity of racial/ethnic boundaries is intermarriage, which has long been considered an indicator of the social distance between groups (Rosenfeld, 2002). The prevalence of intermarriage is strongly influenced by two factors: the strength of preferences for endogamy and demographic factors that govern opportunities for in-group and out-group marriage (e.g., the relative size of groups, the sex ratio, residential segregation) (Stevens and Tyler, 2002). Some studies of intermarriage have taken as their primary question the extent to which social boundaries exist between groups (i.e., there is a preference for in-group versus out-group marriage) and thus have attempted to control for opportunities and constraints imposed by demographic factors when examining patterns of intermarriage. In this chapter, our aim is descriptive and thus does not require controlling for demographic factors. Our goal is to describe patterns of ethnic mixing in marriage, cohabitation, and parenthood. Regardless of whether preferences or demographic factors underlie patterns of interethnic mating, the long-term consequences for racial/ethnic identities are likely to be the same. Higher rates of ethnic mixing between Hispanic subgroups and other groups will potentially reduce racial/ethnic boundaries. One important mechanism through which this potentially occurs is fertility. For instance, offspring with one Hispanic parent and one non-Hispanic white parent are likely to identify more weakly with a specific Hispanic subgroup or with the pan-ethnic Hispanic or Latino labels than offspring with two Hispanic parents, especially coethnic parents (Duncan and Trejo, 2004; Hirschman, 2002).

In , we present summary information on ethnic endogamy16 versus exogamy in marriages and cohabiting unions.17 The data are broken down by the female partner’s ethnicity and generational status. For marriages, there are differences in levels of ethnic endogamy across Hispanic groups, with Mexican Americans exhibiting a higher level of endogamy than all other groups. Among married Mexican women, 84 percent have a Mexican husband; the corresponding figures are 74 percent for Cubans, 65 percent for Central Americans and South Americans, 62 percent for Puerto Ricans, and 55 percent for other Hispanics. The higher level of in-group marriage among Mexican Americans is undoubtedly influenced by the size of the U.S. Mexican population, which allows for relatively high levels of contact with other Mexican Americans. The generational pattern with respect to ethnic endogamy in marriage is very similar across Hispanic groups. In each Hispanic subgroup, there is a marked decline in ethnic endogamy from the first generation to the second. Among Mexicans and Puerto Ricans, a decline is also evident between the second generation and the native-born with native parents; however, among Central Americans and South Americans and other Hispanics, roughly comparable percentages of second- and third (or higher)-generation women are married to partners with similar national origins.


Ethnic Endogamy Versus Exogamy in Coresidential Unions, by Female Partner’s Ethnicity and Generation .

The other side of endogamy is exogamy, and the data for each Hispanic subgroup indicate that married Hispanic women who do not have a co-ethnic husband are relatively likely to be married to a non-Hispanic white.18 For example, 12 percent of married Mexican American women have a non-Hispanic white husband, while only 2 percent are married to a non-Mexican Hispanic and less than 1 percent are married to a non-Hispanic black. Exogamous marriages represent 16 percent (100 – 84) of all marriages among Mexican American women; in such marriages, 78 percent (12.3/15.7) of husbands are non-Hispanic white. The generational pattern with respect to marriages between Hispanics and non-Hispanic whites is also important. In each Hispanic subgroup, the percentage of women with a non-Hispanic white husband rises dramatically across generations.

The second most common type of exogamous marriage involves Hispanic spouses from dissimilar national origins. While such marriages are not very common among Mexicans (2 percent of all marriages), they constitute between 9 percent (Cubans) and 13 percent (Central/South Americans) of all marriages among Hispanic women in other groups. Marriages with Hispanic (but not coethnic) husbands constitute 15 percent (2.4/15.7) of all exogamous marriages among Mexican Americans, compared with 26 percent (11.8/44.7) for other Hispanics, 31 percent (11.9/38.0) for Puerto Ricans, 34 percent (8.8/25.6) for Cubans, and 38 percent (13.2/34.7) for Central/South Americans.

also presents information on cohabiting unions. With few exceptions, the overall level of ethnic endogamy is lower for cohabiting unions than for formal marriages. Among Mexican Americans, for example, 74 percent of all cohabiting unions are endogamous, compared with 84 percent of marriages. In addition, using exogamous unions as the base, the distribution of unions by the race/ethnicity of the partner differs somewhat from that for marriages. In particular, exogamous cohabiting unions are generally less likely to involve a non-Hispanic white partner and more likely to involve a Hispanic partner or a black partner than are exogamous marriages. The figures for black partners are especially striking. Among Mexican American women, for example, about 4 percent (.7/15.7) of exogamous marriages involve a black spouse, while 9 percent (2.4/26.2) of exogamous cohabiting unions involve a black partner. Similarly, among Puerto Ricans, 11 percent (4.0/38.0) of exogamous marriages involve a black partner, compared with 19 percent (8.1/41.8) of exogamous cohabiting unions.

Due to sample size limitations, the full array of generational differences in endogamy in cohabiting unions can be presented only for Mexican Americans. Among Mexican Americans, the generational pattern of endogamy is similar to, albeit stronger than, that observed for marriages—declining percentages in endogamous unions across generations. In addition, exogamous unions involving Mexican American women and non-Hispanic white partners become more common in each successive generation. This is also the case for unions with non-Hispanic black partners, but the overall percentage of unions with non-Hispanic blacks is small.

Interethnic unions are of interest in their own right, but their consequences for ethnic boundaries are greatest when they produce children. Children of mixed unions face complex identity issues, one of which is whether to retain a mixed identity or to adopt one parent’s racial/ethnic identity or the other’s. We have seen that mixed unions among Hispanic women most commonly involve a non-Hispanic white partner. Because such unions both signal and facilitate assimilation into mainstream white society, their offspring are likely to identify less strongly with their Hispanic national origins than children with two coethnic parents. Although numerous factors affect the size and composition of Hispanic groups (e.g., rates of immigration and return migration, socioeconomic mobility), ethnic mixing undoubtedly will contribute to greater fluidity in ethnic identities and therefore play an important role (Hirschman, 2002; Waters, 2002).

In , we expand our analysis by examining interethnic mating among parents of children born in 2000, using data from the 2000 Detail Natality File.19 We first present information on all births and then disaggregate the data into births to married and unmarried mothers. As was the case in the previous table on union patterns, we organize the data by the mother’s ethnicity and generation. However, due to the limited information collected on the birth certificate, we are able to distinguish only between foreign-born mothers and native-born mothers. For mothers in each Hispanic subgroup, the percentages of births in which the father is coethnic, from a different Hispanic group, non-Hispanic white, and non-Hispanic black are shown. These percentages are based on cases in which the father’s race and ethnicity are known; however, since missing information on fathers is problematic in birth certificate data, we also show the percentage of cases in each group with missing information on the father’s ethnicity.


Ethnic Endogamy Versus Exogamy in Parenthood, by Mother’s Ethnicity and Generation .

Focusing first on all births, there are substantial differences in intermating patterns by Hispanic ethnicity and generation. As was the case in our analysis of marital and cohabiting unions, the level of ethnic endogamy is higher among Mexican Americans than for other Hispanic groups. Moreover, for all groups except Mexican Americans, coethnicity of parents is considerably lower than coethnicity of married or cohabiting partners. For example, among Puerto Ricans, 62 percent of married partners and 58 percent of cohabiting partners have similar Hispanic origins; however, only 52 percent of births can be attributed to coethnic parents. The most striking pattern shown in the table, however, is that for generation: infants of foreign-born mothers are substantially more likely to have coethnic parents than infants of native-born mothers. The percentages of children born to coethnic parents for foreign-born and native-born mothers, respectively, are 93 and 74 for Mexicans, 61 and 47 for Puerto Ricans, 70 and 38 for Cubans, 68 and 34 for Central American and South American mothers, and 68 and 46 for other Hispanic mothers. Exogamous unions producing children are highly likely to be with Hispanic fathers (from other national-origin groups) or with non-Hispanic white fathers, with one exception. Mexican-origin women are considerably more likely to bear a child with a non-Hispanic white partner than with a non-Mexican Hispanic partner.

When births are broken down by the marital status of the mother, several important differences in ethnic mixing are evident. First, considerably fewer births to unmarried Hispanic mothers involve partnerships with non-Hispanic white males than is the case for births to married Hispanic mothers. Second, births outside marriage are more likely to involve a non-Hispanic black father than births within marriage. For example, about 8 percent of infants of unmarried Puerto Rican mothers had non-Hispanic white fathers, compared with 24 percent of infants of married Puerto Rican mothers. Children born to unmarried Puerto Rican women were much more likely to have a black father (15 percent) than children born to married Puerto Rican women (8 percent). This pattern is similar across all Hispanic groups. Given the relatively high propensity of non-Hispanic whites to bear children within marriage and the relatively high propensity of non-Hispanic blacks to bear children outside marriage, these patterns appear to reflect the preferences and circumstances of fathers.

In summary, several broad conclusions can be drawn from our analyses of ethnic mixing. First, there are substantial differences across Hispanic groups in the level of ethnic endogamy in marriages, cohabiting unions, and parenthood. The most significant differences are those between Mexican Americans and all other groups: Mexican Americans are substantially more likely to be paired with a coethnic partner in marriage, cohabitation, and parenthood than are Puerto Ricans, Cubans, Central/South Americans, or other Hispanics. Second, in all Hispanic groups, there are marked declines in ethnic endogamy in marriage, cohabitation, and parenthood across generations. This is consistent with a large body of research that shows that intermarriage is a sensitive indicator of assimilation. Finally, the most provocative findings emerge from a comparison of results for marriage, cohabitation, and parenthood. In marriage, there is a higher level of ethnic endogamy than in cohabitation and parenthood. Moreover, among exogamous unions, matches with non-Hispanic white partners are more common in marriage than in cohabitation or parenthood. Unions among partners from different Hispanic origins or between Hispanics and non-Hispanic blacks are considerably more evident in cohabitation and parenthood than they are in marriage. In particular, unions between Hispanics and non-Hispanic blacks are prominent in parenthood, especially nonmarital births.


Portrayals of U.S. Hispanics consistently emphasize their relatively high level of familism and links between familism and traditional family patterns in Latin American–and Caribbean-origin countries. Familism is typically regarded as a multidimensional concept that reflects both values and behaviors that emphasize the needs of the family over the needs of individuals (Vega, 1995). Key questions for understanding family life among Hispanics are (1) whether familistic values and behaviors are more prominent among Hispanics than among other racial and ethnic groups and (2) whether familism wanes with exposure to the U.S. social context (i.e., duration of U.S. residence for the foreign-born or generational status for all members of a Hispanic group). Evaluations of Hispanic familism, however, are complicated by the fact that family behavior is not shaped solely by normative orientations and values; it is also strongly influenced by socioeconomic position and the structure of economic opportunities in the broader society. Thus, contemporary scholars generally argue that Hispanic family patterns can best be understood within a social adaptation framework, which stresses the interplay between familistic values and the circumstances experienced by Hispanics in their everyday lives.

Because the data presented in this chapter are descriptive, we cannot evaluate the relative importance of the aforementioned factors in shaping family behavior among Hispanics. Instead, we identify structural characteristics of families that suggest variation in familism by race/ethnicity and generational status. Several patterns are consistent with the idea that Hispanics are family oriented, relative to non-Hispanics. First, with the exception of Cubans, Hispanics have higher fertility than non-Hispanics. Childbearing also begins earlier in Hispanic women’s lives than it does for non-Hispanic white women. Second, Hispanics are more likely to live in family households than are non-Hispanic whites and blacks. Third, the family households of Hispanics are slightly larger and much more likely to be extended than those of non-Hispanic whites. At the same time, the figures for family structure and children’s living arrangements show that traditional two-parent families are not more common among Hispanics than non-Hispanic whites. In fact, female family headship and one-parent living arrangements for children are considerably more prevalent among Hispanics than non-Hispanic whites, although less prevalent than among non-Hispanic blacks.

A related issue is whether familism declines as Hispanic groups spend more time in the United States. Although comparisons across generations using cross-sectional data must be used cautiously to address this question,20 our analysis of structural measures of familism shows some support for the declining familism thesis. The support is strongest for the Mexican-origin population. On every indicator, the second and third (or higher) generations exhibit less traditional family behavior than the first generation. For instance, in 15 percent of households headed by a first-generation Mexican, the householder is a female with no partner present, compared with 23 percent of households headed by a second- or third (or higher)-generation Mexican. The implications of these differences are particularly striking for children: about 14 percent of first-generation Mexican children live in a mother-only family, compared with 20 percent of second-generation children and 31 percent of third (or higher)-generation children. A similar but somewhat weaker pattern of declining familism across generations is shown for Puerto Ricans, but the evidence is considerably more mixed for the other Hispanic subgroups.

A limitation of this study is that we have only examined the structural dimension of familism. This is due, in part, to the absence of national-level databases that include both information on other dimensions of familism and sufficient numbers of the various Hispanic subgroups to allow for analysis. Future research on attitudinal and behavioral aspects of familism is needed, given the unevenness of conclusions that can be drawn from the existing literature and data. For example, perhaps the best general-purpose survey for describing the attitudinal and behavioral dimensions of familism is the National Survey of Families and Households (NSFH). This survey includes numerous questions that tap normative beliefs about the obligations of parents to support their adult children and the obligations of adult children to support aging parents. It suggests that members of Hispanic groups are more likely than non-Hispanic whites to recognize both parental and filial obligations (results available upon request), although the difference may be due in part to nativity differences between groups and the tendency of the foreign-born to value parental and filial duties. Indeed, Hispanics are more likely than non-Hispanic whites to say they would rely on their children or their parents for emergency help, for a loan, or advice (Kim and McKenry, 1998). These findings are consistent with research based on other data sets, which show that Hispanic adolescents, irrespective of nativity, more strongly respect their parents and feel more obligated to provide their parents with support in the future than non-Hispanic whites (Fuligni, Tseng, and Lam, 1999).

Such findings on the attitudinal dimension of familism stand in sharp contrast to a more complicated set of findings from NSFH-based studies that focus on the behavioral dimension of familism, in particular social participation and both instrumental (money/help) and noninstrumental (advice/support) transfers within families. A concise summary of this literature is complicated by the fact that there is little consistency across studies in research methodology. For example, only some studies disaggregate Hispanics by national origin and generational status, and many studies are restricted to particular stages of the life course (e.g., old age). In addition, there are inconsistencies in the types of support examined as well as whether information is provided on the direction of exchanges (i.e., the providers and recipients of support are identified) (Hogan, Eggebeen, and Clogg, 1993; Lee and Aytec, 1998; Spreizer, Schoeni, and Rao, 1996). Nonetheless, whether one focuses on Hispanics as a generic category or specific subgroups such as Mexican Americans, there is some indication that Hispanics tend to socialize more frequently with relatives than others (Kim and McKenry, 1998). As for giving and receiving support within families, the NSFH suggests that ethnic differences are either trivial or various Hispanic groups tend to participate in fewer exchanges than others. This may be due, in part, to the role of migration in separating family members (Hogan et al., Clogg, 1993) or to the relative lack of resources to give (Lee and Aytac, 1998). More systematic attention to differences in family relations and exchanges by national origin and generation is needed before firm conclusions about these issues can be drawn.21

Another topic considered in this chapter is ethnic mixing in family formation. The future size and composition of the Hispanic population will be shaped by the processes that constitute the well-known demographic balancing equation: population change = births − deaths + net migration. High rates of immigration and relatively high fertility will continue to fuel the rapid growth of the Hispanic population. While these factors are fundamental, there are additional complications in the situation of Hispanics that are not taken into account in population projections based on the balancing equation. Specifically, the equation assumes that there is no intermarriage and that the racial and ethnic identities of children are identical to those of their mothers (National Research Council, 1997). As we have seen, the assumption of full ethnic endogamy is untenable, as is the premise of fixed identities across generations.

Recent changes in family formation behavior and the complexities of ethnic mixing will play significant roles in the future size and composition of Hispanic subgroups. Hispanics have shared in the trend toward cohabitation and nonmarital childbearing that has characterized the general U.S. population. Currently, more than 40 percent of births to Hispanic mothers take place outside marriage (National Center for Health Statistics, 2003), and roughly half of those births are to cohabiting couples (Bumpass and Lu, 2000). Our analysis shows that ethnic exogamy is common in marriage and in marital births among Hispanics—but exogamy is even more prominent in cohabiting unions and in nonmarital childbearing. Thus, recent shifts in the union context of childbearing are linked to growth in the population of children with mixed ethnic backgrounds and to a blurring of boundaries between specific Hispanic subgroups and both other Hispanic subgroups and non-Hispanics.

Importantly, there are differences between Hispanic subgroups and within Hispanic subgroups by generational status in the extent of ethnic mixing. The most consequential differences are those between the Mexican-origin population and all other Hispanic groups. Relative to the other Hispanic subgroups, the Mexican-origin population exhibits much higher levels of ethnic endogamy in marriage, cohabitation, and parenthood. Moreover, while ethnic endogamy in parenthood is lower for native-born mothers than for foreign-born mothers in each Hispanic group, the level of endogamy among native-born Mexican mothers exceeds that for foreign-born mothers in the other groups. Thus, the Mexican-origin population is unique among Hispanics in its high level of ethnic endogamy in marriage, cohabitation, and parenthood. This suggests that there will be fewer exits from the Mexican American population due to mixed racial/ethnic backgrounds of offspring (and consequent identity shifts) than is the case for other groups.

A question that remains unanswered is: What are the implications of these interethnic mating patterns for the future of racial and ethnic boundaries in the United States? Some scholars argue that race and ethnicity are in the process of being reconfigured in U.S. society. Due to the large-scale immigration of groups that are not readily classified as whites or blacks—and to the growth of the mixed-race population—the old black–white dualism is being transformed into a black–nonblack dualism (Gans, 1999). According to Gans (1999), Hispanics and Asians are “in reserve” as a residual category that will be sorted into the principal categories over time by the dominant white society. This sorting process is likely to depend on the socioeconomic position and phenotypic characteristics of Hispanic- and Asian-origin individuals.

Several features of ethnic mixing among Hispanics are consistent with the idea that Hispanics will be classified with whites into the nonblack category of the new racial dualism. First, with the exception of Mexican Americans, the level of exogamy among Hispanics is high and sizeable proportions of exogamous unions are with non-Hispanic whites. Second, very low proportions of exogamous unions are with non-Hispanic blacks. And third, the level of intermixing with non-Hispanic whites increases markedly across generations. In all Hispanic groups except Mexican Americans, more than half of the unions of native-born women are exogamous,22 and such unions frequently involve non-Hispanic white partners. At the same time, there are features of ethnic mixing that are not consistent with the idea of a growing black–nonblack dichotomy in which Hispanics are blending into an undifferentiated nonblack group. One such feature is the relatively high level of ethnic endogamy among Mexican Americans, which will undoubtedly contribute to the persistence of a Mexican ethnic identity and culture. Given the size of the Mexican-origin population and continued high rates of immigration from Mexico, this pattern suggests that “Mexican” or “Hispanic” may continue to be quasi-racial categories for many years to come. Another important factor is the shift in ethnic mixing that has accompanied the trends toward cohabitation and nonmarital childbearing. Cohabitation and nonmarital childbearing among Hispanics are more likely to entail partnerships with non-Hispanic blacks than are marriage and marital childbearing. This is especially the case for some Hispanic subgroups, including Puerto Ricans, Central/South Americans, and Cubans.

In sum, the overall pattern of ethnic mixing among Hispanics does not have unambiguous implications for the future of racial and ethnic boundaries in the United States. Mexican Americans are likely to maintain a distinct ethnic identity, although some blurring of boundaries will occur due to unions with non-Hispanic whites. Other Hispanic subgroups are less likely to sustain distinct identities over time. Furthermore, their higher levels of ethnic mixing with other Hispanic groups and non-Hispanic blacks suggest somewhat greater ambiguity with regard to their placement in a black–nonblack racial system. In short, while current patterns of immigration and ethnic mixing are contributing to a softening of some racial/ethnic boundaries, both race and ethnicity are likely to remain salient and to intersect in complex ways.

APPENDIX TABLE A5-1. Percentage Family Households by Race/Ethnicity and Generational Status of Householder

APPENDIX TABLE A5-2. Characteristics of Family Households by Race/Ethnicity and Generation of Householder


  1. Alan Guttmacher Institute. Teen sex and pregnancy. Facts in brief. 1999. [accessed March 2004]. Available:http://www​
  2. Angel JL, Angel RJ, Markides KS. Later-life immigration: Changes in living arrangements, and headship status among older Mexican-origin individuals. Social Science Quarterly. 2002;81(1):389–403.

  3. Angel RJ, Angel JL, Lee G-Y, Markides KS. Age at migration and family dependency among older Mexican immigrants: Recent evidence from the Mexican American EPESE. The Gerontologist. 1999;39:59–65. [PubMed: 10028771]
  4. Angel RJ, Angel JL, McClellan JL, Markides KS. Nativity, declining health, and preferences in living arrangements among elderly Mexican Americans. The Gerontologist. 1996;36:464–473. [PubMed: 8771974]
  5. Baca Zinn M, Wells B. Diversity within Latino families: New lessons for family social science. In: Demo DH, Allen KR, Fine MA, editors. Handbook of family diversity. New York: Oxford University Press; 2000. pp. 252–273.

  6. Bean FD, Tienda M. The Hispanic population of the United States. New York: Russell Sage Foundation; 1987.

  7. Bean FD, Curtis RL, Marcum JP. Familism and marital satisfaction among Mexican Americans: The effects of family size, wife’s labor force participation, and conjugal power. Journal of Marriage and the Family. 1977;39(4):759–767.

  8. Bramlett MD, Mosher WD. Cohabitation, marriage, divorce, and remarriage in the United States. National Center for Health Statistics. Vital Health Statistics. 2002;23(22) [PubMed: 12183886]
  9. Bumpass L, Lu H. Trends in cohabitation and implications for children’s family contexts in the United States. Population Studies. 2000;54:29–41. [PubMed: 28489526]
  10. Burgess EW, Locke HS. The family: From institution to companionship. New York: American Book; 1945.

  11. Casper LM, Bianchi SM. Continuity and change in the American family. Thousand Oaks, CA: Sage; 2002.

  12. Ch’Eng-K’Un C. Familism the foundation of Chinese social organization. Social Forces. 1944;23(1):50–59.

  13. Duncan B, Trejo SJ. Ethnic choices and the intergenerational progress of Mexican-Americans. Austin: University of Texas Population Research Center; Apr, 2004. (Working Paper #04-05-02.)

  14. Fields J. Current Population Reports. Washington, DC: U.S. Census Bureau; 2003. Children’s living arrangements and characteristics; pp. 20–547.

  15. Fuligni AJ, Tseng V, Lam M. Attitudes toward family obligations among American adolescents with Asian, Latin American, and European backgrounds. Child Development. 1999;70:1030–1044.

  16. Gaines SO Jr, Barajas L, Hicks D, Lyde M, Takahashi Y, Yum N, Garcia BF, Marelich WD, Bledsoe K, Steers WN, Henderson MC, Granrose CS, Rios D, Farris KR, Page MS. Links between race/ethnicity and cultural values as mediated by racial/ethnic identity and moderated by gender. Journal of Personality and Social Psychology. 1997;72(6):1460–1476. [PubMed: 9177025]
  17. Gans HJ. The possibility of a new racial hierarchy in the twenty-first century United States. In: Lamont M, editor. The cultural territories of race: Black and white boundaries. Chicago, IL: University of Chicago Press; 1999. pp. 371–390.

  18. Gordon M. Assimilation in American life: The role of race, religion, and national origins. New York: Oxford University Press; 1964.

  19. Guendelman S. Immigrants may hold clues to protecting health during pregnancy: Exploring a paradox. Oakland: University of California; 1995. (Wellness Lecture Series 5.)

  20. Hirschman C. The meaning of race and ethnic population projections. In: Denton Stewart NA, Tolnay E, editors. American diversity: A demographic challenge for the twenty-first century. Albany: State University of New York Press; 2002. pp. 51–72.

  21. Hogan DP, Eggebeen DJ, Clogg CC. The structure of intergenerational exchanges in American families. American Journal of Sociology. 1993;98:1428–1458.

  22. Kalmijn M. Intermarriage and homogamy: Causes, patterns, trends. Annual Review of Sociology. 1998;24:395–421. [PubMed: 12321971]
  23. Kim HK, McKenry PC. Social networks and support: A comparison of African Americans, Asian Americans, Caucasians, and Hispanics. Journal of Comparative Family Studies. 1998;29:313–334.

  24. Landale NS, Oropesa RS. Migration, social support and perinatal health: An origin-destination analysis of Puerto Rican women. Journal of Health and Social Behavior. 2001;42:166–183. [PubMed: 11467251]
  25. Lee Y-J, Aytac IA. Intergenerational financial support among whites, African Americans, and Latinos. Journal of Marriage and the Family. 1998;60:426–441.

  26. Lesthaeghe R, Meekers D. Value changes and the dimensions of familism in the European community. European Journal of Population. 1986;2:225–268.

  27. Lieberson S, Waters MC. From many strands: Ethnic and racial groups in contemporary America. New York: Russell Sage Foundation; 1988.

  28. Martin JA, Hamilton BE, Ventura SJ, Menaker F, Park MM. Births: Final data for 2000. Hyattsville, MD: National Center for Health Statistics; 2002. (National Vital Statistics Reports 50(5))

  29. Martin P, Midgley E. Immigration: Shaping and reshaping America. Population Bulletin. 2003;58(2):5–42.

  30. Massey DD, Zambrana RE, Bell SA. Contemporary issues in Latin families. In: Zambrana R, editor. Understanding Latino families: Scholarship, policy and practice. Thousand Oaks, CA: Sage; 1995. pp. 190–203.

  31. National Center for Health Statistics. Health: United States, 2003. Hyattsville, MD: U.S. Department of Health and Human Services; 2003.

  32. National Research Council. The new Americans: Economic, demographic, and fiscal effects of immigration. In: Smith JP, Edmonston B, editors. Panel on the Demographic and Economic Impacts of Immigration Committee on Population and Committee on National Statistics. Washington, DC: National Academy Press; 1997.

  33. Oppenheimer VK. The continuing importance of men’s economic position in marriage formation. In: Waite LJ, editor. The ties that bind: Perspectives on marriage and cohabitation. New York: Aldine de Gruyter; 2000. pp. 283–301.

  34. Oropesa RS, Gorman BK. Ethnicity, immigration, and beliefs about marriage as a “tie that binds” In: Waite LJ, editor. The ties that bind: Perspectives on marriage and cohabitation. New York: Aldine de Gruyter; 2000. pp. 188–226.

  35. Oropesa RS, Landale NS. Immigrant legacies: Ethnicity, generation, and children’s familial and economic lives. Social Science Quarterly. 1997;78(2):399–416.

  36. Oropesa RS, Landale NS. The future of marriage and Hispanics. Journal of Marriage and Family. 2004;66:901–920.

  37. Popenoe D. American family decline, 1960–1990: A review and appraisal. Journal of Marriage and Family. 1993;55:527–555.

  38. Portes A, Rumbaut RG. Legacies: The story of the immigrant second generation. New York: Russell Sage Foundation; 2001.

  39. Proktor BD, Dalaker J. Poverty in the United States, 2000. Washington, DC: U.S. Census Bureau; 2003. (Current Population Reports, P60-222)

  40. Ramirez RR, de la Cruz P. The Hispanic population in the United States. Washington, DC: U.S. Census Bureau; 2002. (Current Population Reports, P20-545.)

  41. Rodriguez J, Kosloski M, Kosloski K. The impact of acculturation on attitudinal familism in a community of Puerto Rican Americans. Hispanic Journal of Behavioral Sciences. 1998;20(3):375–390.

  42. Rosenfeld MJ. Measures of assimilation in the marriage market: Mexican Americans, 1970–1990. Journal of Marriage and Family. 2002;64:152–162.

  43. Rumbaut RG, Weeks JR. Unraveling a public health enigma: Why do immigrants experience superior perinatal health outcomes? Research in the Sociology of Health Care. 1996;13B:337–391.

  44. Sabogal F, Marin G, Otero-Sabogal R, VanOss Marin B, Perez-Stable EJ. Hispanic familism and acculturation: What changes and what doesn’t? Hispanic Journal of Behavioral Sciences. 1987;9(4):397–412.

  45. Spreitzer E, Schoeni R, Rao KV. Tracing intergenerational relations through reports of transfers of time and money: A comparative study of African Americans, Hispanics, and whites. International Journal of Sociology and Social Policy. 1996;16:11–34.

  46. Steidel AG, Contreras L, Contreras JM. A new familism scale for use with Latino populations. Hispanic Journal of the Behavioral Sciences. 2003;25(3):312–330.

  47. Stevens G, Tyler MK. Ethnic and racial intermarriage in the United States: Old and new regimes. In: Denton NA, Tolnay SE, editors. American diversity: A demographic challenge for the twenty-first century. Albany: State University of New York Press; 2002. pp. 221–243.

  48. Suro R, Passel J. The rise of the second generation: Changing patterns in Hispanic population growth. Washington, DC: Pew Hispanic Center; 2003.

  49. Sweeney MM. Two decades of family change: The shifting economic foundations of marriage. American Sociological Review. 2002;67:132–147.

  50. Tienda M. Familism and structural assimilation of Mexican Americans in the United States. International Migration Review. 1980;14(3):383–408. [PubMed: 12279974]
  51. Valenzuela A, Dornbusch SM. Familism and social capital in the academic achievement of Mexican origin and Anglo adolescents. Social Science Quarterly. 1994;75:18–36.

  52. Vega WA. The study of Latino families. In: Zambrana R, editor. Understanding Latino families: Scholarship, policy, and practice. Thousand Oaks, CA: Sage; 1995. pp. 3–17.

  53. Ventura SJ. Vital and Health Statistics. Hyattsville, MD: U.S. Department of Health and Human Services; May, 1980. Trends and differentials in births to unmarried women: United States, 1970–1976. Series 21, No 36.

  54. Ventura SJ, Bachrach CA. Nonmarital childbearing in the United States, 1940–1999. Hyattsville, MD: National Center for Health Statistics; 2000. National Vital Statistics Reports, vol. 48, No. 16.

  55. Waters MC. The social construction of race and ethnicity: Some examples from demography. In: Denton NA, Tolnay SE, editors. American diversity: A demographic challenge for the twenty-first century. Albany: State University of New York Press; 2002. pp. 25–49.

  56. Wilson WJ. The truly disadvantaged. Chicago, IL: University of Chicago Press; 1987.

  57. Winch RF, Greer S, Blumberg RL. Ethnicity and extended familism in an upper middle-class suburb. American Sociological Review. 1967;32(2):265–272. [PubMed: 6041102]
  58. Wu LL, Wolfe B, editors. Out of wedlock: Causes and consequences of non-marital fertility. New York: Russell Sage Foundation; 2001.

  59. Wu LL, Bumpass LL, Musick K. Historical and life course trajectories of nonmarital childbearing. In: Wu LL, Wolfe B, editors. Out of wedlock: causes and consequences of nonmarital fertility. New York: Russell Sage Foundation; 2001. pp. 3–48.

  60. Zambrana RE, Scrimshaw SCM, Collins N, Dunkel-Schetter C. Prenatal health behaviors and psychosocial risk factors in pregnant women of Mexican origin: The role of acculturation. American Journal of Public Health. 1997;87:1022–1026. [PMC free article: PMC1380943] [PubMed: 9224189]


In some cases, additional information on the subgroups that comprise the Census Bureau’s Central/South American category is provided. These subgroups include Dominicans, Guatemalans/El Salvadorans, other Central Americans, Colombians, Ecuadorians/Peruvians, and other South Americans.


The figures for the percentage change between 1980 and 2000 were calculated from more precise information (i.e., rounded to hundredths rather than tenths) than that presented in the table. Thus, in some cases, they differ slightly from calculations based on the numbers in the first and third columns of the table.


Because the TFR is based on age-specific fertility rates, it essentially “controls” for the age distribution of groups. Consequently, the youthful age structures of the Hispanic subgroups, relative to non-Hispanic whites, do not explain their relatively high fertility.


In 2000, roughly 50 percent of nonmarital births to both Hispanic and non-Hispanic white women occurred within cohabitation, compared with 22 percent for non-Hispanic blacks (Bumpass and Lu, 2000). The role of cohabitation in nonmarital childbearing also varies across Hispanic subgroups; however, comparable information is not available for specific Hispanic groups.


There is considerable variation among Hispanic subgroups in poverty. In 2001, about 23 percent of Mexicans, 26 percent of Puerto Ricans, 16 percent of Cubans, 15 percent of Central/South Americans, and 18 percent of other Hispanics were poor (Ramirez and de la Cruz, 2002).


For ease of presentation, we refer to the structural/demographic dimension as the structural dimension in the remainder of the text. A similar shorthand is used when discussing the variables used to measure this dimension of familism.


To simplify the presentation of results, the text reports numbers that have been rounded to the nearest whole number. The rounding is based on more precise data than the information that appears in the tables (i.e., rounded to hundredths rather than tenths).


A householder is the first adult household member listed on the census form. The instructions indicate that this should be the person (or one of the people) in whose name the home is owned or rented.


To allow for comparisons with prior studies, we also provide tables in the appendix that are based on the Census Bureau’s definition of a family household. is comparable to and is comparable to . Using the Census Bureau definition, households in which the householder is cohabiting are defined as family households only if there are other relatives of the householder living in the dwelling unit. Thus, a householder living with a cohabiting partner and her children would not be defined as a family household.


Additional analyses (not shown) examined first- and second-generation Dominicans, Guatemalans/El Salvadorans, other Central Americans, Colombians, Ecuadorans/Peruvians, and other South Americans. Because of the recency of immigration from these countries, the third generation was not of sufficient size for inclusion in the analysis. Each of these groups exhibited considerably higher age-standardized percentages of family households than non-Hispanic whites (ranging from 74 percent for other South Americans to 78 percent for Colombians, compared with 69 percent for non-Hispanic whites).


Because of space limitations, we do not present information on the percentage of family households headed by a male householder who does not live with a partner.


We define extended family households as households that are extended vertically or laterally to include relatives who are not part of the nuclear family.


Additional analyses (not shown) that disaggregated Central/South Americans into Dominicans, Guatemalans/El Salvadorans, other Central Americans, Colombians, Ecuadorans/Peruvians, and other South Americans showed that Dominican families were more likely than all other Hispanic families (including Mexicans, Puerto Ricans, and Cubans) to be headed by a female householder with no partner present (42 percent) and to be extended (12 percent). Because of the recency of immigration from Central and South America, the additional analyses were restricted to the first and second generations.


The educational categories used in the standardization were: less than high school; high school graduate; some college; college graduate.


It should be noted that the Current Population Survey variables on children’s living arrangements do not consider the parent’s cohabitation status. Estimates suggest that 12 percent of Hispanic children in mother-only families are living with a single cohabiting mother. The comparable figures for non-Hispanic whites and non-Hispanic blacks are 14 percent and 6 percent, respectively (Fields, 2003).


A marriage (or partnership) is defined as ethnically endogamous if the partners are members of the same Hispanic-origin group (e.g., a Mexican woman is married to a Mexican man).


Intermarriage can be examined using prevalence measures (based on the stock of marriages at a given point in time) or incidence measures (based on marriages that occur during a given period of time) (Kalmijn, 1998). Because our analysis is based on cross-sectional data that describe the characteristics of the general population, we examine intermarriage with prevalence measures.


provides information on exogamous unions with Hispanics, non-Hispanic whites, and non-Hispanic blacks. The rows do not sum to 100 percent because information on exogamous unions with partners from other ethnic groups (e.g., Asians, Native Americans) is omitted from the table.


The natality public-use data files include information on all births occurring in the United States. Data are obtained from birth certificates for births occurring in each state and the District of Columbia. The data are compiled and released in electronic format by the National Center for Health Statistics.


Generational differences found in cross-sectional data can be influenced by in the characteristics of immigrants arriving in the United States at different points in time, as well as differences in the context of reception at the time of arrival.


This is the case for all coresidential unions combined (marriages and cohabiting unions) and partnerships producing children born in 2000.