Skip over navigation

Journal Issue: Immigrant Children Volume 21 Number 1 Spring 2011

The Living Arrangements of Children of Immigrants
Nancy S. Landale Kevin J. A. Thomas Jennifer Van Hook

Living Arrangements of Children of Immigrants

We combine five years of data (2005–09) from the Current Population Survey (CPS), a large nationally representative data set, to document the living arrangements of children under age eighteen. We emphasize two aspects of living arrangements: parental marital and residential status (married co-resident parents, cohabiting co-resident parents, single parent, no resident parents) and the presence of other adults in the household (grandparent, other relative, nonrelative). We focus first on differences in children's living arrangements by parental nativity (whether parents are native- or foreign-born) for four racial and ethnic groups: Hispanics, Asians (including Pacific Islanders), non-Hispanic whites, and non-Hispanic blacks (table 1).

Despite differences across the broad groups, one pattern consistently emerges. Children of immigrants are considerably more likely to live with married parents than are children of natives (52 percent versus 44 percent for Hispanics; 65 percent versus 50 percent for Asians; 63 percent versus 58 percent for non-Hispanic whites; 44 percent versus 24 percent for non-Hispanic blacks). As illustrated in figure 1, the greater likelihood that children of natives will live with a single parent explains most of this difference, although such children are also slightly more likely to live in a household with no resident parents.

Differences by parental nativity in extended-family living arrangements are less consistent across racial and ethnic groups. For example, among Hispanics and blacks, children of immigrants are less likely to live with grandparents than are children of natives (8 percent versus 14 percent for Hispanics; 10 percent versus 15 percent for blacks). Among Asians and non-Hispanics, the share living with grandparents differs little by parental nativity. In contrast, as figure 2 shows, children of immigrants are much more likely to live with extended kin other than grandparents in all groups except non- Hispanic whites. Among Hispanics, for example, 23 percent of children of immigrants have other extended kin living in their households, compared with 14 percent of children of natives. On balance then, children of immigrants are more, but only slightly more, likely to live with either grandparents or other extended-family members than children of natives, except among blacks (31 percent versus 29 percent among Hispanics, 19 percent versus 18 percent among non-Hispanic whites, and 30 percent versus 22 percent among Asians). Among blacks the division is equal at 30 percent.

Finer distinctions among immigrants' children reveal somewhat different living arrangements by the child's generational status (table 2), which is based on the nativity of the child as well as of his or her parents.12 Table 2 separates children with immigrant parents into three groups: the first generation, the second generation, and the 2.5 generation. First-generation children were born outside the United States and had at least one foreign-born parent. Second-generation children were born in the United States and had two foreign-born parents. U.S.-born children with one foreign-born and one U.S.-born parent are the 2.5 generation. Finally, third- or higher-generation children were born in the United States and had two U.S.-born parents.

In general, the share of children living with married parents declines with each generation in the United States, but first-generation children are slightly less likely to live with married parents than are second-generation children. Living with a married parent who has an absent spouse (not shown) is also particularly prevalent among first-generation children. In addition, such children are distinctive in being more likely than other children of immigrants to live in households with no resident parent. These various arrangements suggest that newly arrived immigrant families may encounter complications that reduce children's chances of living with both parents and lead some children to live in households that provide no parental supervision. However, with the exception of first-generation Asian children (who make up 22.5 percent of Asian children of immigrants), first-generation children account for no more than 20 percent of children of immigrants in the major racial and ethnic groups.

Important distinctions also exist by country (or region) of origin within each broad racial or ethnic group. Children in the three specific groups that we highlight (Mexicans, Southeast Asians, and Caribbean blacks) share common disadvantages that stem from their parents' relatively low education and income. Because of the histories and broader contexts of immigration from their countries of origin, however, the groups differ in parental legal status (for example, unauthorized versus authorized), parental work patterns, the types of communities in which they live, and their reception by the native-born population. We thus discuss the family situations and living arrangements of each group separately.

Children in Mexican Immigrant Families
Given the volume of immigration from Mexico, the predominance of children of immigrants among Mexican-origin children, and the comparative youth of the Mexican-origin population, children of Mexican immigrants are of special importance in shaping the future of the U.S. population. According to Census Bureau projections, the Hispanic population will account for nearly one-quarter of the nation's total population by 2040. Jennifer Glick and Jennifer Van Hook estimate that the Mexican-origin population alone will account for 15–17 percent of the U.S. population by then.13 It is therefore important to understand the circumstances that may influence the future outcomes of today's Mexican-origin children.

The major challenge facing Mexican immigrants and their children is their limited opportunity for economic integration, owing in large part to their low education, skills, and financial resources. On average, foreign-born Mexicans have completed eight and a half years of education, compared with about twelve years for native-born Mexicans and more than thirteen years for native-born whites.14 Together with their limited English proficiency and frequently unauthorized legal status, the low education of Mexican immigrant parents severely limits their opportunities for stable, well-paid employment.15 With the premium for education and skills especially high in today's high-tech economy, it is no surprise that about 34 percent of Mexican children of immigrants are poor, compared with 24 percent of Mexican children of natives.16 For these reasons, many scholars and policy analysts are concerned that the Mexican-origin population may remain socially marginalized and economically disadvantaged well into the future.

For children, living in poverty increases the risk of negative outcomes, including health and developmental problems, poor academic performance, low completed education, and low earnings in adulthood. Because poverty and family structure are linked, poor children often face not only resource deficits but also other risk factors associated with single parenthood, such as high family stress, inadequate supervision, multiple family transitions, and frequent residential moves. For Mexican-origin children, however, poverty and family structure vary in a less straightforward manner. Although Mexican children of immigrants have a higher poverty rate than Mexican children of natives, they are more likely to live in two-parent families. As shown in the top panel of table 3, 56 percent of Mexican children of immigrants live with two married parents, compared with 45 percent of Mexican children of natives. When cohabitating parents are included, fully 75 percent of Mexican children of immigrants live in a two-parent family, compared with 63 percent of Mexican children of natives. The favorable family structures of children with foreign-born parents may reduce some of the risk factors typically associated with poverty.

Despite that initial advantage, however, Mexican-origin children increasingly face challenges related to their household and family structure as their families become more settled. In particular, the favorable two-parent family structure becomes much less common among native-born children in both the 2.5 and third generations (see table 4). That pattern suggests that as Mexican families spend more time in the United States (as indexed by generation), the strong parental bonds that protect Mexican-origin children erode. Over time, Mexican families may be more and more subject to the same forces that are increasing single parenthood among American families generally.

Even though Mexican children of natives are more likely to live in single-parent families than Mexican children of immigrants, they have a lower rate of poverty. That finding, however, does not mean that family structure is inconsequential. Poverty rates would be even lower for children of natives if not for their disadvantaged family structure. As illustrated in figure 3, children of natives are less likely to live in poverty regardless of family type. For example, in married-couple families, 28 percent of children of immigrants are poor, compared with 11 percent of children of natives. The explanation for this difference, in large part, is that foreign-born Mexican parents have lower human capital and earnings than do their native-born counterparts. In addition, although employment rates of foreign- and native-born Mexican men are roughly comparable, the employment rate of foreign-born Mexican women is substantially lower (56 percent) than those of their native-born (76 percent) and white counterparts (80 percent).17 Similarly, in single-parent Mexican families, children of immigrants have higher poverty rates (51 percent) than children of natives (40 percent). Still, children of natives are four times more likely to be poor if they live in a single-parent than in a married-couple family. Thus the higher prevalence of single-parent families among Mexican children of natives suggests that economic progress is being eroded by shifts in family structure.

Mexican immigrant families also face special challenges associated with migration itself. Mexican immigrants are predominantly labor migrants, sojourners who come to the United States temporarily to work during their early adult years (as early as late adolescence through their mid-thirties). At least initially, they maintain strong ties with their households and families in Mexico, sending remittances and visiting or even eventually returning home. Others remain permanently in the United States even though many are unauthorized. Not surprisingly, these migration patterns shape children's living arrangements. For example, the circular nature of Mexican labor migration appears to contribute to the formation of highly unstable households made up of both extended kin and non-kin. In addressing the question of why Mexican immigrants are more likely than U.S.-born Mexicans to live in extended families, Van Hook and Glick recently contrasted an explanation focused on cultural norms brought from Mexico with an explanation that stresses the use of extended-family co-residence as a survival strategy.18 They showed that recent immigrants live in household structures very different from those in Mexico, with considerably more lateral extension (for example, living with adult siblings) and co-residence with nonrelatives. As Mexican immigrants live longer in the United States, they are less likely to live in either of those arrangements and less likely to live with extended kin altogether. The study also found that extended-family arrangements are highly unstable, with considerable turnover of household members. Although Van Hook and Glick's research was not based on families with children, it suggests that living in an extended-family may temporarily benefit Mexican children of recent immigrants by helping their parents cope with the many challenges they face when they first arrive in the United States. But such an arrangement is unlikely to be stable or to contribute to children's long-term well-being.

One particularly troubling difficulty posed by migration is that it can separate children from their parents, either because one family member migrates first and later brings over other family members (stage migration) or because a parent is deported or deterred from the dangerous border crossing. Ethnographic accounts detail "transnational family" patterns among labor migrants from Central America and Mexico,19 whereby one or more family members (often a parent) will migrate for work, leaving other family members behind. Children born in the country of origin may be left there in the care of a single parent or relative even as new U.S.-born siblings are raised in the United States, so children in both countries are living apart from one or both parents and siblings. Little is known about how many children live in these types of families, but the number may be substantial. In the United States, although most Mexican children of immigrants live with two parents, 21 percent live with only one parent. Of these "single" parents, 17 percent are married but live apart from their spouse. Although the whereabouts of these parents is unknown, they may be living in Mexico. In a study of an immigrant-sending community in central Mexico, Joanna Dreby found that 28 percent of children had one or both parents living in the United States.20 Clearly further study is warranted to learn more about how long children of immigrants remain separated from their immediate family members and how that separation affects their well-being and future integration into U.S. society. Because of the limits of cross-sectional data for studying family separations and instability, it will be necessary to build binational data sets that follow children and parents over time to advance research in this area.

Migration also separates children from their parents when foreign-born adolescents travel to the United States alone in search of work. Among foreign-born Mexicans aged twelve to seventeen, fully 12 percent live in U.S. households with no resident parent.21 These youths are highly likely to live with relatives other than parents or grandparents (62 percent), such as siblings, cousins, or aunts and uncles, or in households that do not include family members (27 percent). Because they are rarely enrolled in school and are subject to limited supervision, youth living apart from their parents are at high risk of negative short-term outcomes (such as unmet health care needs or drug and alcohol abuse) and negative long-term outcomes (such as limited education and skill development).22 Yet few studies provide information on the circumstances of Mexican children who live apart from their parents. Researchers know little about the stability of their living arrangements or about whether living with extended kin is protective for these vulnerable youth.

Yet another risk factor for Mexican children of immigrants is the mixed legal status of their family members. Almost half of Mexican children of immigrants live in families where the children are citizens and the parents are not. (The comparable share for children of Asian immigrants is 13 percent; for children of European immigrants, 14 percent).23 Beyond lacking U.S. citizenship, many of the parents in Mexican mixed-status families are unauthorized, especially those who have immigrated relatively recently. Using data from 2004, Jeffrey Passel showed that most Mexico-born U.S. residents who entered the country after 1990 were unauthorized, with figures ranging from 70 percent during 1990–94 to 85 percent during 2000–04.24 Although the citizen children of unauthorized parents are on an equal legal footing with all citizen children, their parents' unauthorized status affects them adversely in many ways. Unauthorized parents typically work in unstable, low-wage jobs that do not carry health benefits. Thus Mexican children of unauthorized parents are more likely to be poor than other Mexican children of immigrants. In addition, as Passel notes in his article in this volume, unauthorized parents often fail to take advantage of public benefit programs for which their children qualify, because they fear deportation. These hardships may be intensified by unstable living arrangements and periods of separation from one or both parents. Researchers as yet know little about the family situations of children with unauthorized parents and should make that topic a high priority for future work.

The lives of children in mixed-status families would become especially difficult if U.S. citizenship laws were to change. One particularly disturbing recent development has been the mounting criticisms of birthright citizenship, which grants citizenship to all persons born in the United States as stated in the Fourteenth Amendment to the U.S. Constitution. For the United States to deny citizenship to the U.S.-born children of unauthorized immigrants, as some have advocated, could jeopardize child well-being and Mexicans' prospects for social integration. Jennifer Van Hook and Michael Fix projected the size of the unauthorized population if all children with an unauthorized mother were denied legal status.25 Their mid-level estimates suggest that within four decades, the unauthorized population would be 72 percent higher than the number under current law, and 15 percent of the unauthorized would be third- or higher-generation Americans. Because infants would be the first to lose U.S. citizenship, children would be disproportionately affected. By 2050 the share of all U.S. children who would be unauthorized would more than double, mostly likely exceeding 5 percent. In all likelihood, most of these children would be of Mexican or other Hispanic origin.

Children in Southeast Asian Immigrant and Refugee Families
Refugees come to the United States under very different circumstances than do Mexican labor migrants. Many flee their countries of origin from stressful and sometimes dangerous situations with little or no planning and may be ill prepared for life in the United States. Unlike labor migrants, however, they and their children are legally resident in the United States and receive settlement assistance from the federal government.

Refugees admitted to the United States in recent decades have increasingly come from diverse countries of origin.26 Yet much of what scholars know about the living arrangements of children in refugee families comes from studies of the children of immigrants from Southeast Asia and Indochina—largely because Southeast Asian refugees have been in residence in the United States longer than most of their contemporary counterparts. The three major Southeast Asian refugee groups in the United States are the Vietnamese (whose arrival between 1970 and 2000 resulted in a tenfold increase in the Asian foreign-born population), Cambodians, and Laotians.27 Refugees from Laos include former Hmong guerrillas, a group that fought on behalf of the U.S. government during the Vietnam War, as well as their descendants. In addition to their larger numbers, Vietnamese refugees differ from their counterparts from Cambodia and Laos in other important ways. For example, although children were overrepresented in refugee movements from all three countries, Cambodian and Laotian refugee groups brought with them more children than did the earlier Vietnamese groups.28 Similarly, because Vietnamese refugee families fled to the United States earlier, their families have a greater share of second-generation children.29

Many studies of Southeast Asian immigrants base their discussion of living arrangements on the circumstances of the refugees' flight from conflicts and the ensuing implications for both their mode of entry and their situation after arrival. The refugee experience poses a range of challenges for Southeast Asian children of immigrants through its influence on household characteristics. Parental social and economic attributes, for example, differ for children in refugee and nonrefugee families, with other Asian immigrants tending to be more highly skilled and better educated than Southeast Asian refugees.30 According to Ruben Rumbaut's study of children in San Diego, children in Southeast Asian refugee groups are less likely than other children to have parents who graduated from college. They are also the least likely to live in families that owned their own home.31 Human capital also varies among the refugee groups, with families from Cambodia and Laos more disadvantaged than those from Vietnam. Rumbaut finds that parental schooling and home ownership rates are much lower among Cambodians and Laotians than among the Vietnamese. Other studies find parental human capital lowest among the Laotian Hmong refugees, many of whom were poor rural farmers before migrating to the United States.32

Other characteristics of Southeast Asian refugee parents depend on when they arrived in the United States. Earlier refugee cohorts from Vietnam, for example, had more schooling when they arrived than did more recent arrivals.33 That disparity has important implications for child well-being among the Vietnamese because recent Southeast Asian refugee cohorts arrive with a greater number of children than earlier cohorts.34

The unique context of Southeast Asian refugee immigration also has implications for the family characteristics of the children, some of which pose significant challenges. First, a relatively high share of Southeast Asian children of immigrants lives in nontraditional family structures because of the death of family members, either in war or from hardships of life in refugee camps.35 Cambodian refugee households suffered the worst effects. As observed by Nga Nguy, many Cambodian immigrants had spouses in their native country who were killed or simply taken away by Khmer Rouge guerrillas before they arrived in the United States.36 More than a third of all Cambodian refugees are estimated to have lost either a family member or a close friend.37

Family deaths naturally diminished the likelihood that children would live in two-parent families. Rumbaut, for example, finds that Cambodian youths are less likely than other immigrant youths to live with two parents. Studies of Hmong and other Laotian youths report similar findings. According to the Youth Development Study in Minnesota, most Hmong youths live in households missing either one or two parents who died in conflict or in refugee camps; about two-thirds live in families without a biological father.38 A significant number of Laotian immigrants to the United States also arrived as single parents, having lost their partners to conflict.39

Consequently Southeast Asian children of immigrants are more likely to live in single-parent families than their Asian counterparts overall (16 percent versus 12 percent for all Asians). They are, however, with the exception of Cambodian-origin children, less likely to live with a single parent than either Mexican or Caribbean black children of immigrants. As table 3 shows, about 16 percent of Southeast Asian children of immigrants live in single-parent families (24 percent for Cambodians), compared with 21 percent of Mexican and 43 percent of Caribbean black children of immigrants. Thus, although the significance of premigration parental mortality for family structure may have declined, the high prevalence of single-parent families among Southeast Asian immigrants suggests that their family structure is also a product of other social determinants.

Family structure among Southeast Asian youths is determined by the absence not only of Southeast Asian fathers but also of American fathers of children born outside the United States. For example, Jeremy Hein maintains that a significant number of first-generation Vietnamese and Cambodian children who arrived in the United States with only their mothers and siblings had fathers who were American soldiers.40 Because many of these fathers also died during the wars in their respective countries, only a few of their children were reunited with their fathers after arriving in the United States.

As another consequence of their war experiences, Cambodian immigrants created complex networks of extended-family relationships that foster family cohesion across fragmented households. Hein finds that these networks involve attaching isolated individuals and fragmented families to other families through friendship, fictive kinship, or marriage. It is not unusual for these households to contain multiple generations, as well as married siblings or friends who are unrelated to other household members but nonetheless considered part of the family. Among Vietnamese refugees, interstate mobility after arrival in the United States also complicates household structures. According to Nazli Kibria, many Vietnamese refugees migrate from one U.S. state to another to live with friends and other kin, thus creating new households that allow them to pool resources to combat poverty.41 Hmong household relationships too are often highly complex. Estimates from the 2000 census indicate that the Hmong are more likely than the rest of the U.S. population to live in households that include grandchildren, parents, siblings, and other kin members.42

Table 3 shows that the likelihood of living with a grandparent varies considerably by country of origin among Southeast Asian children of immigrants, from a high of 23 percent for Cambodians to a low of 7 percent for Laotians. All Southeast Asian children, however, are highly likely to live in households with relatives other than grandparents. Combining grandparents and other relatives, fully 37 percent live in households with relatives other than their parents. Almost half of Cambodian children live in complex family households. Southeast Asian refugees are, therefore, highly likely to have siblings, other relatives, and nonrelatives within their households who help provide child care and with whom they share resources.43 Nonetheless, the share of other relatives in their households is roughly comparable to that of Caribbean black children of immigrants and only somewhat higher than that of Mexican children of immigrants.

Southeast Asian children also have larger families than do other immigrant groups. Among Southeast Asians, Hmong families are the largest and also the youngest.44 Southeast Asian families are large for several reasons.45 The first is their fertility rate, which exceeds that of all other immigrants except Mexicans. The second is their desire to retain the characteristics of traditional Southeast Asian families after arriving in the United States. For example, Hmong immigrant families, like families in their country of origin, are formed early in the life course because of early marriage among females and the importance of childbearing.46 As many as half of Hmong girls in California are estimated to marry before age seventeen.47 Zha Blong Xiong and Arunya Tuicomepee report that the Hmong have higher teen birth rates than blacks, Latinos, and other Asians.48 Not surprisingly, their analysis also shows that families consisting of married couples with children are more prevalent among the Hmong than among the U.S. population overall, again reflecting the importance of early marriage and childbearing among Hmong adolescents.

Studies about the possible effects of Southeast Asian childbearing patterns on socioeconomic outcomes report mixed findings. For example, the high fertility rate of Southeast Asians has been linked with an increased risk of welfare dependency.49 And findings show that large family size is associated with low labor force participation among females. Yet, according to many studies, the link between early childbearing and low educational attainment is weaker among Hmong teenage mothers than among their non-Hmong counterparts.50

Southeast Asian youths who immigrate to the United States by themselves are especially vulnerable, particularly when they live in households with no parents present. As table 3 shows, 5.2 percent of Southeast Asian children (and 6.1 percent of Vietnamese children) live without parents. Some of these unaccompanied youths arrive in the United States either as orphans or having been sent by parents to establish initial ties to facilitate future immigration through family reunification preferences.51 Many of these children must make significant life-course transitions, such as their first employment experience, without their parents.52 Despite such known vulnerabilities, however, only a few studies have systematically examined the living arrangements of unaccompanied refugee youths from Southeast Asia. A 1988 study found that many resettled within new U.S. families after arriving in the United States.53 Similar patterns have been found in more recent refugee groups.54

Research on the implications of these living arrangements for children's outcomes focuses on unaccompanied Southeast Asian refugee youths in American foster families—finding, for example, that they have lower grades than their counterparts in ethnic foster families.55 Unaccompanied siblings within the same foster family face other difficulties. For example, Mary Ann Bromley found that youths whose oldest sibling was their "household head" before immigrating to the United States have trouble adjusting when their sibling is replaced as household head by their foster father.56 She also reports that unaccompanied refugee youths in American families are likely to feel isolated and have symptoms of post-traumatic stress disorder, although these feelings generally disappear as their stay in the United States lengthens. When they transition to independent living, older unaccompanied youths in foster care face many practical difficulties, such as taking care of themselves and finding employment to meet their expenses.57

Children in Black Caribbean Families
As U.S. immigration flows have become more diverse, the black foreign-born population has grown larger. Now one of the nation's fastest-growing immigrant groups, black immigrants, particularly those from the Caribbean, have drawn attention from scholars and social commentators for their economic success despite their disadvantaged racial origins.58 Nevertheless, research and policy attention to the living arrangements of their children is generally limited—and at odds with the "success story" often told about black immigrants. Recent studies suggest that the children of black immigrants are more likely than other children to face several types of familial vulnerabilities that have significant implications for their well-being. For example, among all children of immigrants, the children of black immigrants are the least likely to live with two married parents; they are highly likely to live in single-parent families or with grandparents rather than parents.59 In addition, they live in less favorable familial circumstances as they assimilate. As their generational status increases, they are more likely than children in other immigrant groups to continue to live in socioeconomically vulnerable household contexts, such as in single-mother households.

Most studies on the living arrangements of children of black immigrants focus on the largest such group—Caribbean immigrants— in part because they arrived earlier than black immigrants from other regions.60 Recent estimates indicate that more than half of all black children of immigrants have Caribbean-origin parents.61 We thus confine our review of the research to the children of black Caribbean immigrants.

Household living arrangements among black Caribbean immigrants are influenced by gender disparities in Caribbean immigration to the United States. Specifically, there are more female than male Caribbean immigrants, and this has influenced the sex composition of adults in immigrant families.62 Caribbean-origin children of immigrants, especially those from the English-speaking Caribbean, are more likely to live in female-headed families than are children in many other immigrant groups.63 Some scholars suggest that the high prevalence of single-parent families among Caribbean immigrants also reflects the influence of pre-migration familial norms unique to the Caribbean region. The higher prevalence of female-headed households among Caribbean than non-Caribbean immigrants in South Florida, for example, reflects the higher prevalence of such families in Caribbean countries of origin.64 At the same time, female-headed households among Caribbean immigrants sometimes result from shifts in who is designated as household head. Such shifts may arise from the post-immigration economic influence of women in families accompanied by husbands or fathers during their initial migration to the United States.65

Table 3 compares the family structures of black children of immigrants from the Caribbean and from Africa. Caribbean-origin youth are considerably less likely to live with married parents (33 percent) than their counterparts whose parents migrated from Africa (55 percent). They are more likely than any other group shown in table 3 except black children of native-born Americans to live in a single-parent family (43 percent compared with 55 percent).

The prevalence of single-parent families among Caribbean immigrants varies by group and by state of residence. Sherri Grasmuck and Ramon Grosfoguel maintain, for example, that in New York, Dominican immigrants have more female-headed households than do Jamaicans or Haitians.66 But in both California and Florida, Rumbaut finds that the children of Jamaican and Haitian immigrants are the most likely to live in father-absent families.67 Regardless of place of residence, however, Caribbean children in single-parent families fare worse than their counterparts in two-parent families. Among Caribbean immigrants in Southern Florida, for example, children in single-parent families were found to have lower grade point averages, as well as lower math and reading scores, than those in two-parent families.68 In addition, Mary Waters' work among Caribbean youths in New York indicates that children in female-headed single-parent families generally have working mothers whose ability to supervise them is constrained by their limited access to networks of extended-family members and friends.69

Among Caribbean immigrants, single-parent households are sometimes temporary family arrangements associated with sequential patterns of family migration in which females initially migrate with their children to be followed by their spouses.70 Indeed, our analysis of the CPS data shows that among all black children of immigrants living in single-parent households, roughly one in five has a married parent living elsewhere. Stage-migration patterns may thus separate members of black immigrant families, much as they do Mexican immigrant families. Even when the "married-but-apart" group is added to the "married" category, however, the resulting share is substantially lower than that among other children of immigrants. Moreover, evidence suggests that a large share of the U.S.-born children of Caribbean immigrants lives in female-headed single-parent families. Waters, for example, notes a high prevalence of female single-parent households among second-generation black Caribbean children of immigrants,71 suggesting that the high prevalence of single-parent living arrangements among Caribbean families cannot be explained simply by sequential migration patterns and traditional or home country familial norms. The persistence of single-parent families across generations suggests post-immigration influences that are yet to be examined systematically.

Extended-family members who remain in the Caribbean generally play a crucial role in the residential patterns of these children. Parents sometimes send children back to their country of origin to keep them from being socialized negatively by their peers or to influence their developmental trajectories. Once back in the Caribbean, children usually live in nonparent households headed by extended-family members.72 Likewise, when limited resources prevent the entire family from immigrating, siblings left behind live with extended-family members.73 These children, who are generally very young, are raised in nonparent households until their early teenage years, when they are reunited with their parents in the United States.74 Post-migration changes in households also have social implications for the integration of newly arriving Caribbean teenagers into the family. Extended separation between parents and their children may be especially stressful for children whose parents divorce or remarry, or both, in their absence, especially when children have to live with new stepparents.75 Consequently the reunification of Caribbean children and their immigrant parents in the United States after long separation is often associated with elevated parent-child conflict.76