Familial Influences on Poverty Among Young Children in Black Immigrant, U.S.-born Black, and Nonblack Immigrant Families

Abstract

This study examines how familial contexts affect poverty disparities between the children of immigrant and U.S.-born blacks, and among black and nonblack children of immigrants. Despite lower gross child poverty rates in immigrant than in U.S.-born black families, accounting for differences in family structure reveals that child poverty risks among blacks are highest in single-parent black immigrant families. In addition, within two-parent immigrant families, child poverty declines associated with increasing assimilation are greater than the respective declines in single-parent families. The heads of black immigrant households have more schooling than those of native-black households. However, increased schooling has a weaker negative association with child poverty among the former than among the latter. In terms of racial disparities among the children of immigrants, poverty rates are higher among black than nonblack children. This black disadvantage is, however, driven by the outcomes of first-generation children of African and Hispanic-black immigrants. The results also show that although children in refugee families face elevated poverty risks, these risks are higher among black than among nonblack children of refugees. In addition, the poverty-reducing impact associated with having an English-proficient household head is about three times lower among black children of immigrants than among non-Hispanic white children of immigrants.

Introduction

In the past two decades, significant progress was achieved in the study of the well-being of the children of immigrants. Increased scholarly focus on their demographic outcomes yielded new insights on their adjustment processes and critical information needed for policy development (e.g. Hernandez 1999; Hernandez and Charney 1998). This tradition is maintained by even more recent studies that investigate the outcomes of children in America’s new immigrant populations (Kulwicki and Rice 2003; Kurien 2005). Existing research on the children of immigrants, however, reveals that they experience greater socioeconomic disadvantages than their counterparts with U.S.-born parents. For example, poverty levels, overcrowded housing, and other indicators of economic hardship are higher in immigrant families than in families with only U.S.-born parents (Capps et al. 2005; Hernandez and Darke 1999; Oropesa and Landale 1997; Shields and Behrman 2004). More recent studies also underscore the persistence of the hardships faced by the children of immigrants across time. Van Hook et al. (2004), for example, argued that despite fluctuations in their poverty trends, children in immigrant families were more likely to live in poverty than the children of U.S. natives in the last four decades.

Recent transformations in the composition of the U.S. immigrant population provide even more opportunities for extending research on child poverty in immigrant families. One contributor to these changes is the recent surge in black immigration to the United States (Dixon 2006; Gelatt and Dixon 2006; Grieco 2004). Despite these transformations, however, only a handful of studies have examined the social and economic outcomes of children in black immigrant families (e.g., Massey et al. 2007; Rong and Brown 2001; Waters 2001). In terms of poverty research, our understanding of the outcomes of the children of black immigrants is generally based on useful insights provided by a few studies focusing on non-Hispanic blacks (e.g., Lichter et al. 2005; Oropesa and Landale 1997; Van Hook et al. 2004). Broader questions related to whether there are ethnic/national origin differences in child poverty among black immigrants still remain unanswered in the existing literature. Recent studies, however, have demonstrated that there are significant differences in incomes and schooling among black immigrants that may have significant implications for poverty research (Kent 2007; Logan and Deanne 2003). More generally, few studies have examined whether young children of black immigrants are fully able to overcome the disadvantages associated with race as they assimilate into the United States.

Concurrent with these gaps in the literature are two essential concerns about the role of family contexts in mediating poverty differences among immigrants’ children. First, although the impacts of familial contexts on poverty among U.S.-born blacks are now well established (e.g., Eggebeen and Lichter 1991; Jarrett and Burton 1999), little is known about the operation of similar influences in black immigrant families. Likewise, the question of whether familial influences affect child poverty disparities between U.S.-born blacks and their Hispanic and non-Hispanic black immigrant counterparts has not been systematically examined. A second concern relates to the question of how family contexts affect the socioeconomic assimilation of the children of immigrants. For example, there is limited research on how generational status is associated with child poverty in oneand two-parent families. Moreover, previous studies have not examined whether the impacts of assimilation on child poverty across immigrant family contexts are racially differentiated.

This study uses data from the 2000 U.S. census to systematically examine the dynamics of poverty among young children in immigrant families. It considers how child poverty varies within specific family contexts for immigrant families. In the process, attention is given to four specific objectives. First, the study examines the extent to which differences in family contexts account for poverty disparities between young black children in immigrant and U.S.-born families and between black and nonblack children of immigrants. Second, it assesses whether differences in the ethnic origins of black immigrant parents explain the disparities in child poverty between immigrant and U.S.-born blacks. Drawing on the segmented assimilation theory, its third objective is to determine whether child poverty varies by generational status and whether this link can be explained by race, ethnicity, and human capital. Finally, the study investigates whether the association between generational status and child poverty is conditional on race, ethnicity, and human capital as previous studies suggest (Zhou 1997).

I hypothesize that differences in economic assimilation between black and nonblack immigrant children result from disparities in parental human capital endowments and that these endowments will have weaker implications for poverty among black immigrants than among their nonblack counterparts.

Race and the Incorporation of Black Immigrants

Previous research on the significance of race for the incorporation of black immigrants provides a broader context for this analysis. Central to this discussion is the question of whether constructions of race within the United States affect how well black immigrants are integrated into the United States (Bashi and McDaniel 1997; Kalmijn 1995; Lee and Bean 2004; Model 1991; Waters 1994). In general, this question has been addressed from two main perspectives. First, in comparisons between black immigrants and U.S.-born blacks, the higher relative attainment of the former has been considered a reflection of the fact that immigrants’ unique cultural attributes allow them to circumvent the negative impacts of race (Ogbu 1983, 1990; Sowell 1981, 1983). On the contrary, the second perspective maintains that nativity-related socioeconomic differences among blacks do not preclude the macro-level disadvantages associated with racial minority status. Bashi and McDaniel (1997), for example, argued that black immigrants are assimilated into a U.S. racial hierarchy that significantly constrains their levels of attainment relative to nonblack immigrants. Thus, immigrants with phenotypical similarities to U.S.-born blacks will have nuanced patterns of incorporation into a U.S. society in which skin color can be a barrier to social mobility (Kasinitz and Vickerman 2001; Zhou 1997).

These perspectives have a number of generally untested implications for research on the welfare of black immigrants’ children. The first is that if black immigrants have cultural attributes that allow them to outperform U.S.-born blacks, their children should have a relative socioeconomic advantage that is robust across national origins and levels of parental human capital. Similarly, if the cultural superiority hypothesis is valid, the advantage of having immigrant rather than U.S.-born black parents should persist even within disadvantaged familial contexts (e.g., single-parent families). Moreover, if more favorable outcomes among immigrant blacks than among U.S.-born blacks reflect an immigrant’s ability to circumvent the negative impacts of race, their indicators of child well-being should be similar to those of nonblack immigrants. Likewise, the improvements to the welfare of immigrants’ children associated with higher levels of parental human capital or increasing assimilation should be broadly consistent across immigrant racial groups.

Theoretical Perspectives

Human Capital and Work Patterns

A useful framework for understanding the dynamics of poverty among black immigrants can be developed using insights from human capital theory. According to the theory, human capital characteristics, such as skill endowments, levels of schooling, and linguistic attributes, are important determinants of social and economic well-being (e.g., Becker 1962, 1995; Parcel and Dufur 2001). Within this framework, higher levels of human capital (e.g., more years of schooling) are hypothesized to be negatively associated with the likelihood of living in poverty. In fact, according to Becker (1995), the impact of human capital endowments was the principal determinant of the living standards of populations in the twentieth century.

Differences in parental human capital indicators, therefore, have important implications for child poverty research among black immigrant populations. In terms of language ability, for example, Rumbaut (1995) found low levels of English-language proficiency among many Hispanic-black immigrant groups, while Kossoudji (1988) reported a negative impact on Hispanic labor-force outcomes associated with their low levels of English proficiency. In combination, these two findings suggest that children in Hispanic-black immigrant families may face higher risks of poverty than their non-Hispanic black immigrant counterparts. Similarly, racial differences in the returns to English proficiency among immigrants may also affect poverty disparities between black and nonblack children of immigrants. As Stolzenberg and Tienda (1997) argued, holding other factors constant, individuals with minority characteristics receive lower returns to their language abilities than those with nonminority characteristics.

In terms of education, higher schooling levels among black immigrants relative to U.S.-born blacks (Kent 2007; Logan and Deanne 2003) imply that child poverty levels will be lower in black immigrant than in native black families. Nevertheless, because previous studies show that the educational endowments of immigrant and U.S.-born blacks are differentially rewarded in the labor market (Butcher 1993; Daneshvary and Schwer 1994; Dodoo 1991), the mediating role of parental educational attainment on child poverty disparities among blacks is not clear. Conceptually, because of these labor market disparities, the immigrant advantage associated with black immigrant schooling is likely to be offset by their lower returns to schooling relative to the respective returns of their native-born counterparts.

Within the U.S. immigrant population, documented differences in educational attainment that are conditional on race may also have implications for racial disparities in child poverty. Rong and Brown (2001), for example, reported that black immigrants generally have lower levels of schooling compared with nonblack immigrants. Rumbaut (1995) found an analogous schooling disadvantage among black immigrants relative to their white counterparts among immigrants from Latin America and the Caribbean. In addition to their relative schooling disadvantage, black immigrant parents may also be affected by a wage disadvantage relative to nonblack immigrant parents in the U.S. labor market (Butcher 1993). For example, Zavodny (2003) found lower wages among black relative to white Cuban immigrants that continue to persist with increasing duration of residence in the United States.

Parental work patterns also affect children’s poverty risks within immigrant families (Oropesa and Landale 2000; Van Hook et al. 2004). Nevertheless, these impacts are likely to be insignificant if parents are employed in low-paying jobs (Lichter and Eggebeen 1994; Lichter et al. 2005). Among black immigrants, however, average working hours are known to be higher among Africans than among Caribbean immigrants, who also work for more hours than U.S.-born blacks (Dodoo 1997). Fewer studies have systematically examined racial/ethnic differences in working hours within the U.S. immigrant population. One study by Logan et al. (2003), however, maintains that Asian immigrants, specifically Chinese and Koreans, work for more hours than some Hispanic-black immigrants (i.e., Dominicans) in the United States. Despite these differences, the implications of these differentials in parental working patterns for racial disparities in child poverty (e.g., between black and Asian immigrants) have not been extensively examined in previous studies.

Immigrant Assimilation

Assimilation theory also provides another useful framework for understanding child poverty outcomes in black immigrant families. In its conventional sense, the theory predicts that immigrants will become more like native-born populations, or experience improvements in their well-being, as they become more exposed to their host societies (e.g., Alba and Nee 1997). When exposure is operationalized to reflect sequential immigrant generations, conventional assimilation theory predicts a negative association between child poverty and increasing generational status. Alternative conceptualizations of assimilation theory, however, argue that assimilation patterns will be conditional on a range of structural factors that can lead to segmented patterns of immigrant incorporation (Portes and Zhou 1993; Zhou 1997). One determinant of segmented assimilation is the human capital endowment of immigrants (Zhou 1997). Accordingly, immigrant groups with higher levels of schooling are expected to have more positive outcomes as they assimilate compared with immigrants with lower schooling levels. In the same vein, because Hispanic black immigrants have lower levels of schooling than their black African counterparts (Kent 2007), increasing assimilation should hypothetically lead to slower declines in child poverty among the former than among the latter.

Besides human capital, immigrant racial and ethnic identities have an important influence on segmented assimilation patterns. Immigrants with visible minority characteristics (e.g., black immigrants) are expected to be more likely than immigrants without those visible characteristics (e.g., non-Hispanic whites) to experience downward assimilation patterns with increasing exposure to the United States (Portes and Zhou 1993; Rumbaut 1994; Zhou 1997). Undergirding these racial disparities are the impacts of factors associated with the social construction of race in the United States. As a result of these impacts, immigrants with visible minority characteristics may be more likely to face structural barriers to their social mobility than their counterparts with no visible minority racial characteristics.

Family Contexts

In addition to human capital endowments and working patterns, differences in family structure also have significant impacts on children’s poverty risks (Lichter 1997). Similarly, the dynamics of family structure are known to be important influences that drive racial differences in child poverty (Eggebeen and Lichter 1991; Lichter and Landale 1995). On the whole, the negative effects on children’s well-being associated with living in families with one or both parents absent are now well established (Lichter 1997; McLanahan and Schwartz 2002). Significantly, however, Cobb-Clark and Hildebrand (2006) found that the impact of family structure is strongly conditional on parental foreign-born status. In particular, they maintain that the socioeconomic disadvantage associated with single-parent families is stronger in families with an immigrant rather than a native-born household head.

Although previous research has examined how family structure affects child poverty among immigrants from a wide variety of racial/ethnic origins (e.g., Lerman 1996; Lichter et al. 2005; Oropesa and Landale 2000), studies on ethnic variations in these impacts in black immigrant families remain largely unavailable. In fact, what we know about the structure of black immigrant families generally comes from only a handful of studies. Tolnay (2004), for example, demonstrated that the children of black immigrants are more likely to live in two-parent families than their counterparts born to U.S.-born blacks. In addition, in his comparison of the living arrangements of black children of immigrants with those of their Asian, Hispanic, and European counterparts, Brandon (2002) found that black children are less likely than nonblack children to live in families with married parents. His study also points to significant racial differences in the change in children’s living arrangements with increasing assimilation. Accordingly, unlike many of their nonblack counterparts, black children of immigrants are increasingly more likely to live in single-parent families as their generational status increases.

Other studies extend research on the contextual impacts of immigrant families by pointing to the negative impacts of refugee status on the well-being of their children. Poverty rates in refugee families, for example, are generally higher than in nonrefugee families (Stepick and Portes 1986; Tang 2000). In addition, unlike nonrefugee immigrants in the United States, refugees do not always experience declines in their poverty status with increasing duration of residence (Fass 1986). Van Hook et al. (2004) have also suggested that the socioeconomic disadvantage of children in refugee families relative to the children of nonrefugees remained unchanged during the last four decades. Significantly, however, although the United States received large numbers of black (e.g., African), Asian (e.g., Hmong), and non-Hispanic white (e.g., Bosnian) refugees in recent decades, the question of whether the impact of refugee status on child poverty is differentiated by immigrant race has not been addressed in previous studies.

Data and Methods

Data used in this study are taken from a 5% sample of the 2000 U.S. census. Since these data are cross-sectional, their practical utility lies in the fact that they are able to provide initial descriptions of the relationships between race, immigration, and child poverty in U.S. families. Specifically, however, the empirical analysis focuses on children 10 years old and younger. This analytical focus is based on two factors. First, the timing of children’s exposure to poverty has significant implications for their long-term development. Guo (1998), for example, demonstrated that exposure to poverty from birth to preadolescence has a greater impact on later cognitive ability relative to exposure to poverty in early adolescence. Second, poverty status is also known to have more immediate implications for educational, behavioral, and other developmental outcomes in early childhood (Duncan et al. 1994; Ramey and Campbell 1991).

The 2000 census data contain information on basic demographic indicators (e.g., age and sex), economic characteristics (e.g., individual and household incomes, poverty status), family relationships (e.g., relationships with household heads), migration-related attributes (e.g., place of birth and duration of residence), and English proficiency. Moreover, since the census provides information on all individuals living within specific households, it is possible to link parental-level information (e.g., parent’s age, sex, schooling, and hours worked per week) with information for each child living with them using unique household identification numbers. These data can also be used to create measures of household structure. Accordingly, they are used in the study to identify single-parent families as households with a household head but without a spouse. By contrast, two-parent families have both household heads and spouses.

Data on countries of birth and the racial identity of household heads and their spouses are used to identify those who are black immigrants as members of the black foreign-born population of the United States. Consistent with previous studies (e.g., Hernandez and Charney 1998), the children in black immigrant families are identified in the analysis as children with either household heads or spouses of household heads who are black immigrants.¹ On the contrary, children in U.S.-born black families are identified as children with only black U.S.-born parents. Since the analysis of racial disparities is also an important part of the study, the study’s definition of children in black immigrant families is limited to the black children who live in such families.²

For all children of immigrants, information on whether they were born in the United States is used to distinguish between first- and second-generation children. First-generation children are those foreign-born children in immigrant families, and second-generation children are the U.S.-born children of these families. Data on the countries of birth of immigrant household heads and spouses are also used to distinguish between the children of black immigrants with different panethnic/national origins. For example, children in African immigrant families are identified as children in families in which either the household head or spouse is a black immigrant from Africa.³ Information on whether immigrant household heads or their spouses are of Hispanic origin is also used to identify Hispanic black immigrant families. In sum, four mutually exclusive types of black immigrant families are used in the study: Hispanic black, African, non-Hispanic Caribbean, and other black immigrant families. The last type mainly includes non-Hispanic black immigrant families from South America (mainly Brazil), the Middle East, and Europe.

In the absence of information on refugee status in the U.S. census, some previous studies have identified the children of refugees by using proxy methods that capture the extent to which parental place of birth corresponds to a list of 11 traditional refugee-sending countries (e.g., Van Hook et al. 2004). While this strategy is generally useful, data on refugee admissions from the statistical yearbooks of the Office of Immigration Statistics (U.S. Immigration and Naturalization Service 1997, 1998, 1999, 2000, 2001) reveal that the number of refugees entering the United States, as well as their countries of origin, vary across time. It is therefore possible that for some years, only a small proportion of all immigrants from these traditional sending countries will actually be refugees. To limit these potential biases, the current study uses a modified refugee-identification proxy method based on census information on country of birth and year of arrival, as well as year-specific information on refugee admissions provided by the U.S. Office of Immigration Statistics. Based on these data sources, I identify refugees as immigrants born in a country from which at least 50% of all immigrants arriving during the immigrant’s own year of arrival were refugees. For example, I classify immigrants from Sudan who arrived in 1998 as refugees because 1,252 officially documented refugees from Sudan arrived in the United States in 1998 (U.S. Department of Homeland Security 2006), and the census estimates that the sum of all Sudanese immigrant arrivals that year is about 2,300. Children in refugee immigrant families are then identified as children with at least one refugee immigrant parent.⁴

Analytical Strategy

Based on poverty thresholds in the U.S. census, an indicator of child poverty status is used as a dependent variable (living in poverty = 1; not living in poverty = 0) in logistic regression models with robust standard errors that adjust for clustering within households. The empirical analysis then proceeds in two stages. In the first stage, a pooled sample of children with immigrant and U.S.-born black parents is used to compare intraracial poverty disparities. These disparities are determined by comparing regression coefficients estimating the likelihood of child poverty in immigrant families relative to that found in U.S.-born black families (the reference group). Models with interaction terms are also used to examine whether various family-level influences (e.g., female-headed households) have differential impacts in immigrant and U.S.-born black families.

The second stage of the analysis uses census information for all children of immigrants. Dummy variables are then employed in logistic regression models to examine disparities in child poverty between black, Hispanic, Asian, and non-Hispanic white children of immigrants.⁵ In addition, interaction terms are used to examine whether the impacts of factors such as family structure, refugee status, and parental human capital vary across race. Finally, for each race, controls for child generational status are used to examine whether changes in child poverty risks with increasing generation are racially circumscribed.

Results

Descriptive Findings

Comparisons of the child and background characteristics of black immigrant families and U.S.-born black families are presented in Table 1. An obvious distinction between children in both groups is that they live in families with contrasting sociodemographic profiles. For example, the children of U.S.-born blacks are about twice as likely to live in single-parent families than their counterparts in each of the four black immigrant groups. In black immigrant families, however, the likelihood of living in single-parent families is differentiated by the ethnic origins of household heads and spouses. Thus, while the children of black Africans are the least likely to live in single-parent families, such family arrangements are most common among the children of non-Hispanic blacks from the Caribbean. A child’s generational status is, however, negatively associated with the likelihood of living in a single-parent family, except among Hispanic blacks.

Table 1.

A summary description of the characteristics of children in immigrant and U.S.-born black families

	Black Immigrants
	Caribbean	African	Hispanic	Other	All	U.S.-born blacks
First-Generation Immigrant	12	19.9	14.6	5.9	13.6	—
Second-Generation Immigrant	88	80.1	85.4	94.1	86.4	—
Single-Parent Family (%)	39.2***	24.6***	34.6***	35.4***	35.0***	62.8
First-generation immigrant in   single-parent family (%)	45.2	28.6	28.0	36.4	37.4	—
Second-generation immigrant in  single-parent family (%)	38.3	23.6	35.7	35.4	34.6	—
Family Size	4.7***	4.8***	4.8***	4.3*	4.7***	4.4
Young Children in Poverty (%)	21.0***	19.4***	30.5***	19.0***	21.3***	38.0
Household Head
Female	47.2***	26.8***	41.8***	41.2***	41.3***	64.1
College graduate	16.3***	49.2***	11.4***	26.3***	24.6***	9.4
English proficient	82.5	77.9	49.5	92.6	79.6	—
Hours worked per week (mean)	35.4***	37.9***	33.3***	38.4***	36.1***	31.6
Spouse
College graduate	49.9***	49.8***	42.7***	55.7***	49.7***	68.6
English proficient	49.2	56.4	30.9	58.9	50.2	—
Hours worked per week (mean)	29.8**	27.5**	24.1***	30.9	28.7***	30.3
N	17,910	7,312	2,693	2,575	30,490	276,775

^*p < .05;

^**p < .01;

^***p < .001

Table 1 also demonstrates that groups with a high prevalence of single-parent families (e.g., U.S.-born and non-Hispanic Caribbean black families) have higher percentages of female-headed families. Consistent with previous studies, the household heads of black immigrant families, especially those heading African immigrant households, have higher levels of educational attainment than their counterparts heading native-black households.

More importantly, Table 1 shows that young children in U.S.-born black families are more likely to live in poverty than their counterparts in immigrant families. While this finding is consistent with findings reported in previous studies, Table 1 clarifies the disparity further by demonstrating that the magnitude of the immigrant advantage is conditional on panethnic origins. In short, children with parents from the non-Hispanic black immigrant groups (i.e., African, non-Hispanic Caribbean, and other blacks) have poverty levels that are about 50% lower than those among the children of U.S.-born blacks. Children of Hispanic black immigrant families, however, have the highest child poverty levels among black immigrants. Not surprisingly, unlike their counterparts with non-Hispanic black immigrant parents, their poverty levels are only about 20% lower than those for U.S.-born black families.

Table 2 contains comparisons between black immigrants and their nonblack immigrant counterparts. Racial disparities shown in Table 2 are similar to those found in the United States in previous studies (Cameron and Heckman 2001; Eggebeen and Lichter 1991; Lichter et al. 2005; Zhou 1993). Specifically, a greater proportion of black immigrant children than immigrant children from Asian, Hispanic, and non-Hispanic white backgrounds live in single-parent families. In fact, black children of immigrants are collectively about three times more likely to live in single-parent families than their counterparts who are either Asian or non-Hispanic white. The higher prevalence of single-parent families among blacks does not appear to be explained by differences in their ethnic origins. As such, there is still a higher prevalence of single-parent families among black immigrants, relative to their nonblack counterparts, when estimates for the four immigrant groups in Table 1 are compared with those in Table 2.

Table 2.

A summary description of the characteristics of black, Asian, Hispanic white, and non-Hispanic white children in immigrant families

	Black	Asian	Hispanic White	Non-Hispanic White
First-Generation Immigrant	13.6	17.9a,e	15.8b,e	11.6d,e
Second-Generation Immigrant	86.4	82.1a,e	84.2b,e	88.4d,e
Single-Parent Family	35.0	12.0a,e	20.3a,e	10.2a,e
First-generation immigrant in single-parent   family (%)	37.4	12.4a,e	21.4a,e	9.8a,e
Second-generation immigrant in single-parent   family (%)	34.6	12.0a,e	20.2a,e	10.3a,e
Family Size	4.7	4.9b,e	5.3a,e	4.5a,e
Young Children in Poverty (%)	21.3	14.9a,e	30.2b,e	11.1a,e
Household Head
Female	41.3	15.8a,e	21.2a,e	17.3a,e
College graduate	24.6	45.3a,e	8.4a,e	40.4a,e
English proficient	79.6	53.2a,e	40.0a,e	82.0d,e
Hours worked per week (mean)	36.1	38.7c,e	37.1a,e	41.4a,e
Spouse
College graduate	49.7e	48.5a,e	26.8a,e	42.2b,e
English proficient	50.2	41.2a,e	26.5a,e	71.2a,e
Hours worked per week (mean)	28.7	23.9b,e	19.9a,e	22.3a,e
N	30,490	65,536	115,471	108,612

^aSignificantly different from each of the five black immigrant subgroups: p < .05

^bSignificantly different from African, non-Hispanic Caribbean, and other blacks: p < .05

^cSignificantly different from African, non-Hispanic Caribbean, and Hispanic blacks: p < .05

^dSignificantly different from African, Hispanic blacks, and other blacks: p < .05

^eSignificantly different from all black immigrants combined: p < .05

With regard to household head and spousal human capital, a lower percentage of black immigrants’ children live in families in which the household heads graduated from college in comparison with their nonblack counterparts. Nevertheless, accounting for differences in ethnic origin reveals that the heads in African immigrant households are more likely to have university credentials than their counterparts heading families with Asian, Hispanic white, and non-Hispanic white children (Tables 1 and 2). With regard to language ability, black children of immigrants have more English-proficient (i.e., speak only English or speak English very well) household heads or spouses of household heads than their Asian or Hispanic white counterparts. Non-Hispanic white children are especially more likely than other children to have English-proficient household heads or spouses of household heads. In addition, they are also more likely than other immigrant children to have household heads who work longer hours per week.

Table 2 further reveals significant racial disparities in poverty among the children of immigrants. The highest poverty rates are among white Hispanic and black children of immigrants. Thus, while the black children of immigrants are less likely to live in poverty than their U.S.-born counterparts (Table 1), they still have higher poverty levels than Asian or non-Hispanic white immigrant children (Table 2). Simultaneously, the prevalence of poverty among Hispanic white children is nonetheless higher than that of black children in African, non-Hispanic Caribbean, or other black immigrant households. There is no racial difference in the prevalence of poverty among Hispanic children of immigrants.

Multiple Regression Results

Family Contexts, Generational Status, and Poverty Disparities Among Blacks

The question of how family context explains child poverty disparities in immigrant and nonimmigrant black families is examined in the regression models shown in Table 3. In particular, the models assess whether differences in family structure mediate child poverty disparities that are conditional on the nativity status of heads and spouses. Simultaneously, they also examine variations in the impacts of assimilation on the poverty risks of black immigrants’ children across family structure.

Table 3.

Logistic regression coefficient estimates of the likelihood of living in poverty among the children of black immigrants and U.S.-born blacks

	Model 1	Model 2	Model 3
All Children in Black Immigrant Families			0.07
All Children in U.S.-born Black Families (ref.)			(0.00)
Single-Parent Families
First-generation children in immigrant families	1.52***	0.52***
Second-generation children in immigrant families	1.17***	0.32***
Child of U.S.-born blacks	0.71***	0.55***
Two-Parent Families
First-generation children in immigrant families	0.66***	0.37***
Second-generation children in immigrant families	−0.30***	−0.24***
Child of U.S.-born blacks (ref.)	(0.00)	(0.00)
Child’s Age	−0.02***	−0.03***	−0.03***
Male	−0.01	−0.01	−0.01
Other Family Characteristics
All single-parent families			1.42***
Children in refugee families		1.07***	1.62***
Family size		0.03***	0.05***
Parental Characteristics
Female head		0.59***	0.52***
Female head × children in immigrant families			−0.64***
Head is college graduate		−1.37***	−1.54***
Head college graduate × children in immigrant families			0.61***
Head is English proficient		−0.34***	−0.50***
Head’s working hours		−0.04***	−0.05***
Head work hours × children in immigrant families			0.00*
Spouse is college graduate		−0.84***
Spouse is English proficient		−0.26***
Spouse’s working hours		−0.04***
Constant	−1.16***	1.23***	0.06
N	307,264	307,264	307,264
Log Pseudo-Likelihood ***(’000)	−176.9	−146.7	−150.8

^*p < .05;

^**p < .01;

^***p < .001

Accounting for differences in family structure reveals important contextual effects that do not appear in the analysis of gross child poverty levels. Thus, although the children of black immigrants are less likely to live in poverty than their counterparts with U.S.-born parents (Table 1), when poverty risks are differentiated by family structure (Model 1), the results show that children in single-parent black immigrant families have the highest poverty risks among blacks. Chi-square tests of the differences between children with U.S.-born parents and their first- and second-generation immigrant counterparts in single-parent families indicate that the immigrant disadvantage relative to nonimmigrants is statistically significant (p < .001). Thus, among single-parent families, the likelihood of living in poverty among first-generation children of black immigrants is more than double that for the children of U.S.-born blacks. Model 1 further suggests that the lower average poverty level for immigrant relative to U.S.-born black families observed in Table 1 is mainly driven by the outcomes of second-generation children in two-parent immigrant families. Differences in family structure also appear to mediate the impacts of assimilation on child poverty risks in black immigrant families. Specifically, although Model 1 shows that there are higher poverty risks among first-generation children regardless of family structure, the intergenerational percentage reduction in child poverty is much smaller in single-parent families than in two-parent family contexts.

In Model 2, the results indicate that among black children, living in refugee families is associated with greater poverty risks than living in nonrefugee families.⁶ In addition, when other family characteristics (e.g., female household headship, household head human capital) are simultaneously controlled, living in single-parent families appears to have about the same effect on poverty on both first-generation children of immigrants and the children of native blacks. Accounting for these factors also eliminates the disadvantage of second-generation children in single-parent families relative to their counterparts in black single-parent families. Thus, a possible explanation for the reversal in the disadvantage of children in single-parent immigrant families (i.e., between Models 1 and 2) is that their socioeconomic outcomes are largely driven by background factors, such as the working patterns and sex of their household heads. However, despite the fact that the relative disadvantage of the children of immigrants within single-parent families disappears in Model 2, all children with only a single parent still have higher poverty risks compared with children with two U.S.-born black parents, even after household head, spousal, and other family characteristics are controlled.

Several interaction terms are used in Model 3 to investigate potential disparities in how other familial influences affect child poverty in immigrant and native black families. In the process, particular focus is given to the impacts of the characteristics of household heads. Controlling for differences in other familial characteristics, living in female-headed families has a weaker effect on child poverty in immigrant than in U.S.-born families. In other words, the socioeconomic disadvantages usually associated with living in female-headed households (Rocha 1997, Snyder and McLaughlin 2004) are less likely to impair children’s well-being in female-headed immigrant families than in native-born black families. In terms of nativity differences in parental human capital, Model 3 indicates that having a college-educated household head is associated with smaller child poverty reductions in black immigrant than in U.S.-born black families. Given the high relative levels of schooling of black immigrant household heads (Table 1), the weaker effect of household heads’ schooling suggested by these results is particularly surprising. The main implication of this finding is that educated black immigrant household heads face more impediments in the use of their human capital endowments for improving children’s well-being than their native-born counterparts.

Model 3 also suggests that the working hours of household heads have no effect on child poverty differences in immigrant and native-born black families. Black immigrant household heads, however, work for longer hours than their U.S.-born counterparts (Table 1), which should conceptually result in a stronger negative association between parental work hours and child poverty in immigrant than in U.S.-born black households. The absence of poverty differentials associated with the working hours of household heads in Model 3 is, however, consistent with observations of a wage disadvantage among black immigrants relative to their U.S.-born counterparts (Butcher 1993; Daneshvary and Schwer 1994). In other words, the conceptual advantages associated with longer working hours seem to be largely offset by impacts of disparities in wages. Therefore, Model 3 suggests that the heads of black immigrant families work for longer hours than their counterparts in native black families to achieve about the same improvements on their children’s well-being.

Table 4 aims to determine whether the overall immigrant advantage across ethnic origins observed in Table 1 remains even after factors, such as generational status and familial characteristics, are controlled. In the process, three insights on the mediating effects of generational status within ethnic origin on poverty disparities among blacks are generated. First, despite higher child poverty levels in native black families than in African and Hispanic black families, first-generation children in these two immigrant groups face higher poverty risks than their counterparts with native black parents (Model 1). In fact, when both ethnicity and generational status are considered, child poverty risks among blacks are highest for the first-generation children of Hispanic black immigrants.⁷ Second, Table 4 suggests that the disadvantage of first-generation children of Hispanic black immigrants may be explained by the fact that they live in less favorable familial environments. Accounting for the human capital of household heads and spouses as well as other background characteristics in Model 2, therefore, has a greater impact on reducing the baseline poverty risks (from Model 1) of the children of Hispanic black immigrants than those of other black children. The third finding presented in Table 4 is that first-generation children of non-Hispanic Caribbean immigrants may live in more favorable economic contexts than their other first-generation counterparts. As such, they are the only group of first-generation children with lower poverty risks than the children of U.S.-born blacks (Model 1).

Table 4.

Logistic regression coefficient estimates examining poverty disparities between children of Black immigrants with various ethnic origins and the children of US-born Blacks

	Model 1	Model 2
African Families
First-generation immigrant	0.53***	0.55***
Second-generation immigrant	−0.59***	−0.19**
Black Hispanic Families
First-generation immigrant	0.68***	0.15
Second-generation immigrant	0.08	−0.28**
Non-Hispanic Caribbean Families
First-generation immigrant	−0.17*	−0.33***
Second-generation immigrant	−0.56***	−0.54***
Other Black Immigrant Families
First-generation immigrant	0.11	0.16
Second-generation immigrant	−0.64***	−0.30**
U.S.-born Black Families (ref.)	(0.00)	(0.00)
Child’s Age	−0.02***	−0.03***
Male	−0.01	−0.01
Family Characteristics
Single-parent families	1.83***	0.65***
Children in refugee families		0.73**
Family size		0.04***
Parental Characteristics
Female head		0.60***
Head is college graduate		−1.39***
Head is English proficient		−0.29***
Head’s working hours		−0.04***
Spouse is college graduate		−0.82***
Spouse is English proficient		−0.44***
Spouse’s working hours		−0.04***
Constant	−1.63***	1.38***
N	307,264	307,264
Log Pseudo-Likelihood ***(’000)	−176.9	−146.6

^*p < .05;

^**p < .01;

^***p < .001

Family Contexts, Assimilation, and Racial Disparities Among Young Children in Immigrant Families

Comparisons of the familial influences on poverty among black and nonblack children of immigrants are shown in Table 5. In particular, the question of whether the impacts of specific poverty risk factors are mediated by race is examined by interacting race with selected familial characteristics (e.g., living in refugee families). Among the main findings shown in Table 5 is the fact that the negative impacts of several risk factors are more deleterious among black than nonblack children of immigrants. In Model 1, for example, the socioeconomic disadvantage associated with refugee families is clearly differentiated by race. In short, having a refugee parent significantly elevates the relative poverty risks of black children relative to those of their Asian and non-Hispanic white counterparts. The disparate effect of refugee families across race is, however, driven by the effects of other background factors. Consequently, when the human capital of household heads and spouses, gender of household heads, and other factors are controlled, the relative disadvantage of black children in refugee families ceases to be significant (Model 5). In terms of racial disparities in the impacts of single-parent families, Model 1 also shows that Asian children are comparatively less likely than their black or non-Hispanic white counterparts to live in poverty. In such contexts, however, the relative likelihood of living in poverty is highest among non-Hispanic white children.

Table 5.

Coefficient estimates showing differences in familial influences on poverty among black children of immigrants and their Asian, Non-Hispanic white, and Hispanic white counterparts

	Model 1	Model 2	Model 3	Model 4	Model 5
Black	−0.90***	−0.82***	−0.14**	−0.43***	−0.24**
Asian	−0.91***	−0.86***	−0.42***	−0.70***	−0.50***
Non-Hispanic White	−1.29***	−1.20***	−0.24***	−1.19***	−0.48***
Hispanic White	(0.00)	(0.00)	(0.00)	(0.00)	(0.00)
Age					−0.02***
Male					−0.01
Second-Generation Immigrant					−0.48***
Single-Parent Family	1.00***	0.96***	1.09***	0.93***	0.80***
Black × single-parent family	0.41***				0.41***
Asian × single-parent family	0.27***				0.07**
Non-Hispanic white × single-parent family	0.50***				0.52***
Refugee Family	0.42				0.14
Black × refugee family	1.69*				1.28
Asian × refugee	0.75				0.30
Non-Hispanic white × refugee family	1.53				0.69
Family Size					0.01
Female Head		0.33***			−0.10***
Black × female head		0.17**			0.11
Asian × female head		0.13**			−0.01
Non-Hispanic white × female head		0.12**			−0.28***
Head is College Graduate			−1.05***		−0.99***
Black × head is graduate			−0.09		0.06
Asian × head is college graduate			0.04		0.06
Non-Hispanic white × head is graduate			0.12		0.16*
Head is English Proficient			−0.57***		−0.50***
Black × head is proficient			−0.21***		−0.15*
Asian × head is proficient			−0.18**		−0.06
Non-Hispanic white × head is proficient			−0.63***		−0.40***
Head’s Working Hours				−0.04***	−0.04***
Black × head’s working hours				−0.01***	−0.01***
Asian × head’s working hours				−0.01***	0.00*
Non-Hispanic white × head’s working hours				0.01***	0.00
Constant	−1.07***	−1.14***	−0.85***	0.39***	1.04***
N	320,108	320,108	320,108	320,108	320,108
Log Pseudo-Likelihood ***(’000)	−145.2	−145.5	−139.1	−132.5	126.8

^*p < .05;

^**p < .01;

^***p < .001

Models 2 to 4 specifically focus on whether the gender of household heads and other factors have disparate impacts on children that are conditional on race. As in Table 5, these models focus only on the mediating effects of the characteristics of household heads.⁸ In Model 2, interaction terms for female-headed households and child race suggest that the negative impact on poverty associated with female-headed households is slightly higher among black than nonblack children. Chi-square tests of the differences between their respective coefficients, however, indicate that the differences between blacks and their non-Hispanic white and Asian counterparts are not statistically significant. Non-Hispanic white children in female-headed households are nonetheless different from other children in such households in one important regard: their relative poverty disadvantage may be explained by factors such as the human capital and working hours of their household heads. Thus, when these factors are controlled (Model 5), non-Hispanic white children in female-headed households have comparatively lower poverty risks than other children in female-headed immigrant households.

Model 3 tests whether the poverty-reducing impacts of household heads’ education and linguistic abilities varies across race. Having a household head who graduated from college is associated with a reduction in children’s poverty risks that is greater among black than among nonblack children. However, these differences are not statistically significant. Notwithstanding this nonsignificant finding, fundamental racial differences in the association between child poverty and household heads’ English-language proficiency remain. For example, Model 3 shows that the relative poverty reduction associated with having an English-proficient household head is about three times lower among black than among non-Hispanic white children. Nevertheless, English proficiency seems to have a stronger poverty-reducing impact among black, Asian, and non-Hispanic white children compared with their Hispanic white counterparts. In Model 4, black children experience slightly more reductions in their poverty risks than non-Hispanic and Hispanic white children as the working hours of household heads increase. However, because the absolute effects of working hours are smaller in relative terms compared with the effects of factors such as language proficiency, the contribution of the work patterns of household heads to racial disparities in real terms are likely to be trivial.

Previous studies on the children of immigrants have broadly described intergenerational child poverty disparities that are conditional on race. Consequently, the analysis of these declines in Table 6 focuses on comparisons between black children, differentiated by ethnic origins, and their nonblack counterparts. In doing so, Table 6 allows examination of whether the predicted disadvantage of black relative to white immigrants, based on segmented assimilation theory, is robust to differences in black ethnic origins. Similarly, because the theory suggests that human capital endowments affect immigrant outcomes during assimilation, the analysis examines whether the high schooling levels of the household heads of black children in African families confer more socioeconomic advantages to them compared with other children as generational status increases.

Table 6.

Logistic regression coefficient estimates of the impacts of generational status and black ethnic origin on poverty disparities among children in immigrant families

	Model 1	Model 2
Black
First-generation immigrant
African	0.25**	0.50***
Black Hispanic	0.38**	0.35***
Non-Hispanic Caribbean	−0.03	−0.11
Other	0.05	0.27
Second-generation immigrant
African	−0.77***	−0.17**
Black Hispanic	0.03	−0.04
Non-Hispanic Caribbean	−0.47***	−0.34***
Other	−0.59***	−0.14
Asian
First-generation immigrant	−0.38***	−0.11**
Second-generation immigrant	−0.96***	−0.52***
Non-Hispanic White
First-generation immigrant	−0.41***	−0.08*
Second-generation immigrant	−1.31***	−0.61***
Hispanic White
First-generation immigrant	0.43***	0.37***
Second-generation immigrant (ref.)	(0.00)	(0.00)
Age		0.01***
Male		−0.01
Family Characteristics
Single-parent family		0.35***
Refugee family		0.70***
Family size		0.01*
Parental Characteristics
Female head		0.31***
Head is college graduate		−0.80***
Head is English proficient		−0.43***
Head’s working hours		−0.04***
Spouse is college graduate		−0.56***
Spouse is English proficient		−0.30***
Spouse’s working hours		−0.04***
Constant	−0.91***	1.45***
N	320,108	320,108
Log Pseudo-Likelihood ***(’000)	−150.5	−117.9

^*p < .05;

^**p < .01;

^***p < .001

As Table 6 illustrates, black children in immigrant families generally face higher poverty risks than their non-Hispanic white counterparts, regardless of ethnic origins or children’s generational status. With the exception of the children of Africans, black children also experience smaller intergenerational poverty reductions than non-Hispanic white children (Model 1). In addition, intragenerational comparisons further confirm that black children are consistently more likely than their non-Hispanic white and Asian counterparts to live in poverty in each immigrant generation. The higher poverty risks of black relative to Asian and non-Hispanic white children are generally unexplained by factors such as refugee status, family size, or parental human-capital characteristics. As such, even after these factors are controlled (Model 2), Asian and non-Hispanic whites still have lower poverty outcomes than their counterparts in the four black immigrant family types. Children in all four black immigrant groups are, however, less likely to live in poverty than their Hispanic white counterparts, regardless of generational status. Nevertheless, poverty disparities between black and white Hispanic children remain imperceptible within each immigrant generation (Model 1) even after other factors are controlled (Model 2).

In terms of disparities in the association between generational status and child poverty, racial differences in the magnitude of the decline between the first two generations generally reflect patterns of disparities in the schooling of household heads. Thus, while children in African families have higher levels of poverty than non-Hispanic white and Asian children, their intergenerational decline in child poverty is larger than that of all other children of immigrants (Model 1). Conceptually, this finding is consistent with the notion that children’s economic assimilation processes can be facilitated by high levels of parental human capital. At the same time, the persistent disadvantage of the children of Africans, relative to their non-Hispanic white and Asian counterparts, is also very instructive. What this suggests is that, despite its effect on intergenerational improvements in children’s well-being, the human capital endowments of household heads have a limited effect on the elimination of racial/ethnic socioeconomic disparities among the children of immigrants.

Conclusions

Recent increases in black immigration to the United States present new opportunities for extending research on the children of immigrants. This study expands our knowledge on black immigrants’ adaptation processes by examining the economic well-being of their children. Yet, as the results suggest, a number of familial risk factors affect poverty among black children of immigrants differently than among their nonblack counterparts. More importantly, the findings show that black immigrant families face greater structural impediments to improving the welfare of their children than do the families of nonblack immigrants. In this regard, these findings are consistent with Bashi and McDaniel’s (1997) notion that black immigrants are incorporated into a U.S. racial system in which barriers to social mobility are generally conditional on race. The analysis also demonstrates that familial contexts also mediate disparities in child poverty between immigrant and U.S.-born blacks. Therefore, despite lower overall child poverty levels for immigrant families relative to black U.S.-born families, it seems that black immigrant parents work for more hours than native black parents to achieve about the same improvements to their children’s welfare. These findings suggest that black immigrants are unlikely to have cultural attributes that give them a robust socioeconomic advantage in improving their children’s welfare. Rather, the practical imperatives associated with parental work conditions can significantly circumscribe the implications of parental economic activity on child well-being among blacks.

With regard to child poverty among blacks, this study provides three additional clarifications on the critical differences between immigrant and U.S.-born families. First, the findings show that while overall levels of child poverty are lower for immigrant families than for U.S.-born black families, it seems that poverty comparisons at a higher level of aggregation conceal the mediating role of family contexts in understanding disparities in poverty. Accordingly, accounting for differences in family structure reveals that the highest poverty risks faced by black children are found among children in single-parent black immigrant families. Second, the results further suggest that improvements in child well-being associated with assimilation are weaker in single-parent family contexts than in two-parent households. Thus, even though child poverty declines between the first and second generation in black immigrant families, second-generation children in single-parent families still have worse outcomes than children in either single- or two-parent native-born black families. Relative levels of socioeconomic assimilation are, however, greater among children in two-parent black immigrant families than among their counterparts with single parents. In fact, as these results show, the overall child poverty advantage of black immigrants is primarily driven by the lower poverty risks faced by second-generation children in two-parent families.

A third important finding that emerged from this study is that disparities in child poverty among black immigrants vary by ethnicity. Unlike previous studies, this study shows that children in Hispanic black immigrant families, especially those in the first generation, face higher poverty risks than other black children. However, the disadvantage of the children of Hispanic black immigrants appears to reflect the low levels of human capital among their household heads. Thus, when these background characteristics are controlled, their comparative disadvantage relative to their non-Hispanic black counterparts (i.e., the children of African, non-Hispanic Caribbean, and other black immigrants) generally disappears. Differences in ethnic origins also mediate the suggested impact of assimilation on reductions in child poverty. Accordingly, intergenerational poverty declines are greater among ethnic groups with higher levels of human capital (i.e., African immigrants) relative to those with less human capital (e.g., Hispanic blacks).

Analysis among immigrant groups suggests that child poverty is higher among black than nonblack children of immigrants, which is consistent with the findings of previous studies (Lichter et al. 2005; Oropesa and Landale 1997). Critical to the understanding of these differences, however, are important insights on the disparate influences of familial contexts that are generally conditional on race. First, disadvantaged family contexts appear to have more adverse effects on the well-being of black immigrant children relative to their nonblack counterparts. For example, the decline in children’s welfare associated with living in refugee families is higher among blacks than among nonblacks. In addition, conventional pathways associated with improvements in children’s well-being, such as improved linguistic ability among household heads, appear to be less consequential for black children than for non-Hispanic white children. Furthermore, the implications of parental human capital for the socioeconomic assimilation of the children of immigrants may be tempered by disadvantages associated with race for black immigrants. Accordingly, child poverty risks are still much higher in black African families than in Asian and non-Hispanic white families, despite the fact that their household heads had the most schooling. Finally, the analyses indicate that poverty levels among Hispanic children of immigrants are not necessarily differentiated by race. However, despite these similarities, child poverty risks fall between the first and second generation slightly faster among white Hispanic children than among black Hispanic children in immigrant families (Table 6, Model 1).

Seemingly, interactions between familial contexts and race are critical to understanding the dynamics of child poverty in immigrant families. Thus, this study’s findings are partly consistent with segmented assimilation theory that predicts, among other things, that the outcomes of immigrants during the assimilation process will be conditional on racial and ethnic identity. Yet, as the results suggest, interactions between familial contexts and race may contribute to even more segmented patterns of assimilation among the children of immigrants. For the black children of immigrants, these results reinforce the view that socioeconomic disadvantages associated with their minority status may be transferred between generations as they assimilate. Beyond these racial influences, however, are the impacts of familial contexts on the creation of further disparities in their assimilation patterns. Constricted socioeconomic assimilation patterns within single-parent families, and the higher prevalence of such families among black immigrants, may also have implications for understanding other disparities among immigrants’ children. Further research on how race and family interactions affect child well-being among immigrants will enhance our understanding of the effects of early childhood environments on immigrant disparities later in the life course.