Abstract
Is immigrant groups' assimilation to host society at odds with their engagement with the country of ancestral origin? This study divides the concept of assimilation into socioeconomic resources and attachment to host society, and argues that assimilation and transnational perspectives are coexisting paradigms. Analyses using the nationally representative samples of Latino and Asian Americans indicate that 1) higher-order generations reduce the odds of home country engagement, i.e. frequent return visits, 2) attachment to American society does not discourage return visits, 3) socioeconomic resources increase frequent visits, and 4) the country of origin is a significant predictor of home country visits.
Introduction
Sociological studies on immigration have traditionally focused on the assimilation of immigrants to the host society. Past studies emphasize the processes of immigrant adaptation to American society through language acquisition, socioeconomic mobility, and marriage with native-born Americans (e.g., Gordon, 1964). In general, most empirical research shows increased levels of assimilation as immigrants increase their exposure to American society across generations (e.g., Alba and Nee, 2003). A logical corollary of assimilation research is that contacts with one's country of origin would decline over generations as immigrants become settled and adapted to American society. This view is challenged, however, by the emerging literature on transnationalism, which posits that continued involvement with the country of origin is a common pattern among immigrants (Glick Schiller et al., 1995). In a recent study, Portes, Haller and Guarnizo (2002) found that transnational engagement among Latino immigrants is associated with higher human capital resources, such as higher education, higher occupational status, and longer length of stay in the US. This presents an interesting puzzle: is engagement with the country of origin really at odds with assimilation to American society?
This paper conceptualizes and analyzes one aspect of transnationalism, i.e. frequent contact with a country of origin, and argues that previous studies subsumed two different dimensions --the socioeconomic resources and attachment to host society -- under one unified label of assimilation. By specifying both of these dimensions and their impact on transnational engagement, I examine whether assimilation to the host country and transnational ties to the country of origin are competitive, complementary, or independent concepts. Specifically, my model separates the attachments that motivate immigrants and their children to maintain transnational home engagement from the socioeconomic resources that enable them to do so. More exposure to the host society, measured by generational status, would produce socioeconomic resources such as income and education that allow immigrant groups to pursue cross-national activities. The exposure to American society is, at the same time, expected to decrease the level of home engagement as suggested by classical assimilation research, because a greater level of attachment to the US reduces immigrants and their children's motivations to remain involved in the country of ancestral origin. This framework allows for the possibility that the assimilation and transnational engagement are independent of each other. Immigrants may retain their attachment to their country of origin over generations, independent of their levels of attachment to the US (Glick Schiller et al., 1995).
I examine these perspectives using a nationally representative sample of Latino and Asian Americans in the US. Previous studies of transnationalism have focused primarily on first and second generation Latino Americans. Despite the socioeconomic, political, and linguistic diversity of immigrants from Asian countries, a general pattern of transnational engagement for Asian immigrant groups has not been given a central place in the transnational literature (Portes et al., 2002). By testing a model with a diverse group of Asian and Latino immigrants and their descendents, this paper aims to explore the nature of transnational ties for Asian and Latino individuals in the Unites States. Examining both Asian and Latino samples would also expand our scope condition and allow us to test the robustness of findings from the Latino samples.
Literature
Assimilation Perspective
Assimilation is defined in the sociological literature as the erosion of differences between groups, between the majority population and minorities, and between immigrants and the native born individuals (Alba and Nee, 2003; Gordon, 1964; Massey, 1981). Classical assimilation theory argues that immigrant groups become more integrated into mainstream America by entering primary-group associations with the native born (Gordon, 1964). Gordon's influential book claimed that participation in social cliques, neighborhoods, friendships, and intermarriages with the native population would eventually lead to erosion of distinctive ethnic groups. Adaptation to host society is, according to Gordon, assumed to take place at the cost of immigrants groups' ethnic characteristics, including ethnic identity, language, and cultural values brought from abroad (Gordon, 1964:81). In other words, a classical assimilation perspective assumes that immigrants' interactions with native-born Americans would gradually replace those from their original society.
Recent studies on assimilation provide more nuanced processes of immigrant adaptation (Brubaker, 2001; Alba and Nee, 2003). Segmented assimilation theory, for example, highlights the roles of community and institutional factors in various pathways toward incorporation. In addition to individual characteristics associated with the degree of exposure to American society, political relations between sending and receiving countries, the nature of co-ethnic communities, prejudice in receiving society, and parental SES are shown to shape the opportunities and constraints especially among the first and second generations (Zhou, 1997; Portes and Rumbaut, 2006). Other studies elaborate the changing nature of ethnic boundaries, instead of assuming the complete disappearance of ethnic groups (Gans, 1997). Studies emphasize the blurred racial/ethnic boundaries after the generations of intermarriages (Alba and Nee, 1997), arguing that individuals have options for shifting their identities depending on the structure of racial discrimination (Waters, 1994), and sometimes without any commitment to the identity of the ancestral origin (Alba and Nee, 2003). Modified by recent studies, assimilation theory now provides more nuanced explanations for the different rates and patterns of adaptation by groups and individuals (Levitt and Jaworsky, 2007).
The assimilation hypothesis, informed by the assimilation literature, would predict a negative association between assimilation and immigrants' ties to their countries of origin. Alba and Nee (2003) argue that transnational ties would decline significantly after the second generation because of language barriers between native-born children of immigrants and their relatives in the sending countries. The assimilation perspective would therefore suggest the erosion of distinctive ethnic ties and an eventual decline in engagement with home countries with the passage of time (Alba and Nee, 2003; Guarnizo et al., 2003; Sana, 2005). Specifically, immigrant groups' attachments to and contact with the country of origin are expected to become less salient over generations as their descendents establish their lives and become rooted in the US. The negative relationship between assimilation and transnationalism is presented as line 1 in Figure 1.
Figure 1. Conceptual Relationship between Assimilation and Transnational Ties.
Transnational Perspective
Transnationalism is an emerging research priority within the field of international migration studies. In the early 1990s, the phenomenon of increased cross-national activities among immigrants was highlighted by anthropologists (Glick Schiller, 1999; Glick Schiller et al., 1992; Levitt et al., 2003). The studies have shown that immigrants maintain their ties with their countries of origin through various means, including travels to home, remittances, voting, and religious activities. The volume and intensity of their economic, political, and social engagement with the country of origin lead transnationalists to argue that immigrants' lives and their identities may be developed in relation to more than one nation. By acknowledging the multiple identities and loyalties of immigrants, the transnational perspective challenged the perspective of classical assimilation, which assumed that immigrants' lives are bounded by nation-states and that assimilation and transnational involvements are incompatible. This line of argument claims that establishing a new life in a destination country does not necessarily detract from immigrants' economic, political, and social commitments to their country of origin (Foner, 2000; Glick Schiller et al.; 1995, Levitt, 2001). The immigrant transnational hypothesis, therefore, would expect a constant level of transnational involvement among immigrants and possibly among the subsequent generations, despite their assimilation into host society. Line 2 in Figure 1 represents the immigrant transnational perspective.
Another line of the transnational argument narrows the definition of transnationalism and suggests a complementary relationship between assimilation and transnationalism (Portes et al., 1999). These studies define transnational migrants as “a new class of immigrants, economic entrepreneurs or political activists who conduct cross-border activities on a regular basis” (Guarnizo et al., 2003:1213), emphasizing the behavioral aspects of immigrants' engagement with their countries of origin. Transnationalism, according to these authors, is available for a limited number of immigrants who are able to maintain active and regular involvement with a country of origin. Their studies on the first generation Latino Americans have consistently found a positive association between transnationalism and adaptation to American society. Latino immigrants who have spent more years in the US, with higher educational levels and higher occupational statuses are more likely to engage in transnational activities. Higher human capital and presence of social networks are considered important for transmigrants to sustain complex cross-national activities (Guarnizo et al., 2003; Portes et al., 2007). This third hypothesis, informed by behavioral transnational studies, therefore, predicts a positive relationship between the level of adaptation to the US and the level of transnational engagement. Line 3 represents this third perspective.
This study considers problematic that assimilation and transnational perspectives do not engage much with each other. While the three lines of research above attempt to define the relationship between assimilation and transnational ties, less attention has been paid to reconcile the seemingly contradictory hypotheses and their conflicting findings. In an attempt to provide an integrative framework, the next section suggests the ways in which three perspectives contribute to explaining immigrant assimilation and transnationalism.
Conceptual Model and Hypotheses
Return Visits
This study focuses on Latino and Asian American's return visits to a country of ancestral origin (Waldinger, 2007; Kasinitz et al., 2002; Rumbaut, 2002). Return visits are a grass-roots practice for immigrants and their descendents to maintain connections and identity with their family and way of life in the country of origin while settling in a new society. By physically moving between the home and a new destination country, immigrants and their children link two distinct localities and their social practices and foster a transnational field (Duval, 2004; O'Flaherty et al., 2007). Studies of return visits, therefore, provide a way to examine how transnational networks are actually maintained and practiced. Studies also demonstrate that return visits to the home country are essential for maintaining transnational ties to original places especially for children of immigrants. Visiting their original country and having face-to-face interactions with relatives often affirms values that their family brought from abroad and creates an emotional tie that helps them identify with their ethnic origin. One study of children of Chinese migrants, for example, describes the experience of return visits as the bridging of identity between being “Chinese” and being “Chinese Americans” (Louie, 2002).
Conceptual Model - Resources and Motivation
I propose that both assimilation and transnational engagement are processes that comprise common aspects of the immigrant adaptation. Figure 2 shows the conceptual model which divides assimilation into two components: socioeconomic resources and attachment to American society. Each component of assimilation produces different mechanisms to link exposure in American society to home country engagement. A framework that distinguishes the two dimensions of assimilation helps us understand how assimilation to host society and transnational ties can be complementary, competitive, or independent concepts.
Figure 2. Conceptual Model.
The first component of assimilation, socioeconomic resources, consists of the income and educational attainment that influences the capability of immigrants and their descendents to engage in cross-national activities. Having a sufficient income and a stable job, for example, would facilitate immigrants' return visits. Trips to Asian countries from the US are relatively expensive compared to trips between the US and Latin American countries. Having money, therefore, would allow Asian Americans to purchase airplane tickets and take vacation time from work to visit their countries of origin. Socioeconomic resources, therefore, are expected to increase transnational engagement. Portes and colleagues' (2002) studies, which found positive relationships between human capital and transnationalism among Latino immigrants, partially highlighted this path.
The second component of assimilation, attachment to the US, indicates immigrants' and their descendents' affinity for cultural practices in American society, which influences their motivation for transnational engagement. The assimilation and transnational perspectives disagree over the nature of the relationship between attachment to one's destination society and commitment to one's sending society. Assimilation theory assumes a zero-sum relationship where exposure to American society increases attachment to the US while at the same time decreasing the attachment to the original culture. By interacting with the native-born Americans through schooling, employment, and other institutional settings, immigrants and their children learn English, establish their new life, and consequently become a citizen of destination society (Gordon 1964). This theory would argue that the increasing level of attachment to the US would replace immigrants' and their children's old attachment to the country of origin, and reduce their motivation to engage in transnational activities. If this hypothesis holds, connections with the country of origin should decline as immigrant groups become exposed to American society and increase their attachment to it.
Immigrant transnationalism literature, however, points out that attachment to the original country is independent from attachment to the destination society. Transnational scholars would claim that immigrants and their descendents are capable of maintaining multiple attachments: the level of attachment to their countries of origin will independently affect their transnational home engagement and will not be influenced by the level of attachment to the United States (Glick Schiller, Basch and Blanc-Szanton, 1995).
Resources and attachment to the destination society may be further influenced by the degree of exposure to the US. According to classical assimilation theory, generational changes would increase socioeconomic status, attachment to host society, but decrease the attachment to countries of ancestral origin.
In sum, several research hypotheses can be drawn to link the level of exposure to the level of home engagement. It should be noted, however, that the causal relationship between the variables of assimilation and the level of home engagement is ambiguous.
H1: Socioeconomic resources increase the level of transnational engagement.
H2: Attachment to the US decreases the level of transnational engagement.
H3: Attachment to the original country increases the level of transnational engagement.
H4: Exposure to the US increases the resources and attachment to the US, but reduces attachment to the country of ancestral origin.
Typology of Transnationalism
Combining various levels of resources and attachment allows us to see four different aspects of transnational home engagement previously discussed separately in the literature (Figure 3). These can be, in turn, influenced by institutional and demographic factors. The first category is identity-based transnational home engagement where attachment to the country of origin outweighs socioeconomic resources. Letter-writing among Polish immigrants at the turn of the 20th century (Thomas, Znaniecki, and Zaretsky, 1984) or Cuban Americans' phone calling at the turn of the current century (Waldinger 2007) may exemplify this type of “imagined” transnationalism based on identity and loyalties. Institutional factors are most likely to affect this type of home engagement. Recent studies point out the influence of governmental arrangements on cross-national networks and activities (World Bank, 2006; Portes et al., 2007). For instance, travel restrictions between the US and Cuba, or between the US and Vietnam until the mid-1990's may constrain the opportunity to visit the country of origin, and may force immigrants to engage more in identity-based home engagement.
Figure 3. Typology of Transnational Home Engagement.
The second category, resource-based transnationalism, is the situation where resources and attachment are complementary. Individuals with surplus resources and commitment to engage in the affairs of a country of origin are most likely to undertake long term cross-border activities. Active involvement with local politics in the country of origin, regular participation in a cultural group, and ethnic entrepreneurship that sustain their livelihood (Portes et al., 1999) are considered prototypes of this transnationalism. A lack of institutional constraints and presence of family members in the original country may promote long-term commitment to the affairs of countries of origin.
The third category, symbolic transnationalism or classical assimilation entails a lack of cross-national activities and attachment to the country of origin. As Alba and Nee (2003) explained in terms of ethnic identity in American society, descendents of immigrants may occasionally claim their connections to their countries of ancestral origin by emphasizing their ethnic heritage. This type of claim, however, is temporary and contingent on opportunities. Presence of immediate family members in the US may root immigrant individuals' and their children's lives in the destination society.
Marginalization could be an alternative route to the classical assimilation. As segmented assimilation theory proposes, children of immigrants could lose their ties to their country of origin without gaining economic progress in the US (Zhou, 1997). Being a minority race and experiencing discrimination in the US may force them to experience downward assimilation in the destination society, and attenuate the loyalty to the country of origin.
The following empirical section examines one of the key questions in the literature, i.e., whether generational status affects the transnational engagement. This shows the general trend of transnational activities across generations and race/ethnicity, and demonstrates whether the differences across generations and between Latino and Asian Americans are a function of socioeconomic adaptation and attachment to American society at the individual level.
Methods
Data
Data for this study come from the National Latino and Asian American Survey (NLAAS). NLAAS collected information from nationally representative samples of adults from four Latino groups (Cuban, Mexican, Puerto Rican, and other Latino descent), four Asian groups (Chinese, Filipino, Vietnamese, and other Asian descent), and a control group of non-Hispanic, non-Asian, white respondents. Household population aged 18 years or older, who met self-identified ethnic group criteria, and who lived in one of the 50 states and Washington D.C., was asked to participate in the interview survey. Computer-assisted face-to-face interviews were conducted from May 2002 to November 2003 in English, Spanish, Mandarin, Cantonese, Tagalong or Vietnamese (Heeringa et al., 2004).
The households were sampled using two sampling methods: core sampling based on multistage stratified area probability sampling designs; and high-density supplemental sampling to oversample contiguous groupings of Census blocks with 5% or greater density of Chinese, Filipino, Vietnamese, and Puerto Rican. An eligible respondent in Latino and Asian households was then asked to participate in interviews2. The overall sample consists of 2,095 Asians, 2,554 Latinos, and 215 whites. The final response rate was 73%.
Analyses conducted in this study adjust for the hierarchical nature of the multistage cluster survey data. Analytic weights were developed for each sample respondent to take into account the three factors: 1) differences in individual selection probabilities; 2) adjustment for non-response; and 3) post stratification of the sample to 2000 Census population totals for geographic region, age, gender and ethnicity groupings of the survey population. Coefficients and standard errors were estimated using the Taylor series linearization method implemented in the SUrvey DAta ANalysis (SUDAAN) software system.
Measurement
Dependent variables used in this study are the overall frequency of return visits to the country of origin (since immigration for the first generation respondents), and the number of visits in the12 months prior to the interview. To assess the overall frequency of visits, the questionnaire asked “How often have you returned to your country of origin (the country of origin of your parents, if respondent was born in the US) – often, sometimes, rarely, or never?” The response was collapsed into 0=rarely/never, or 1=sometimes/often for multivariate analyses. For the second variable, the number of return visits in the previous one year, respondents were asked “How many times have you returned to your country of origin (the country of origin of your parents, if respondent was born in the US) in the last year?” The variable was recoded into a binary variable where 0 indicates no visits and 1 indicates once or more visits in the previous year.
One of the key independent variables, exposure to the US, is measured as a generational status. The foreign-born individuals who arrived in the US after age 12 are considered as the first generation; the native-born of foreign parent(s) and those who came to the US before age 13 (1.5 generation) are defined as the second generation; and the native-born of native parents are defined as the third generation and above.
Country of origin
Ethnic origin is a proxy for the country of ancestral origin. Respondents were asked to choose their main ethnic origin besides being American. The response was categorized into one of four Latino groups (Cuban, Puerto Rican, Mexican, and Other Latino3) and four Asian groups (Vietnamese, Filipino, Chinese, and Other Asian4). Mexican and Chinese are treated as reference groups in the multivariate analyses.
Resources and Attachment
Socioeconomic resources are analyzed as the level of educational attainment and household income, controlling for the household size. Educational attainment is measured as the number of years of schooling. Dummy variables were created to represent: less than a high school education (less than 12 years); high school graduate (12 years); some college education (13-15 years); and college graduate and above (16+ years). Annual household income was divided by the size of the household. The multivariate analyses use logged values. The degree of attachment to the US is measured as the English language proficiency and citizenship (Gordon, 1964; Yang, 1994). Attachment to the country of origin is measured as Spanish or Asian language proficiency. The language proficiency scales were created using three question items asking the respondent to rank his or her ability to speak, read and write in English, Spanish, and Asian languages. The language scale construction originated from the Cultural Identity Scales for Latino Adolescents (Félix-Ortiz, Newcomb, and Meyers, 1994). Lower scores indicate a lower level of proficiency in English while higher scores indicate a higher level of proficiency. US citizenship is a dichotomous variable (non-citizen or citizen).
Demographic factors
The following variables measure demographic characteristics of respondents: marital status (ever-married, or never-married), the presence of dependents aged under 18 (no dependents, or one or more dependents), age, and self-identified race. For the self-identified race, respondents were asked to choose a racial group which best describe their race. Dummy variables were created for whites, blacks, and other which mainly consists of Mestizo.
Results and Discussions
Descriptive Statistics
Table 1 shows the distribution of home country return visits for Latino and Asian samples. The level of return visits is generally similar between Latinos and Asians. About 40% of Latinos return to their countries of origin often or sometimes and 25% returned once or more in the previous one year. Similarly, about 30% of Asian immigrants return to the country of origin often or sometimes, and 25% returned in the previous year. The similar level of frequency of travel to home countries suggests that geographical distance alone may not explain the level of home country visits between Latinos and Asians.
Table 1. Descriptive statistics of home engagement by race/ethnicity (NLAAS 2002-2003).
| Latino Sample (n=2554) | Asian Sample (n=2095) | |||
|---|---|---|---|---|
| unweighted n | weighted % | unweighted n | weighted % | |
| “How often have you returned to your country of origin?” | ||||
| never | 978 | 35 | 730 | 34.2 |
| rarely | 623 | 25.4 | 663 | 31.8 |
| sometimes | 590 | 24.5 | 485 | 23.2 |
| often | 343 | 15.2 | 195 | 10.7 |
| missing | 20 | 22 | ||
| “How many times have you returned to your country of origin last year?” | ||||
| 0 | 1970 | 75.3 | 1592 | 74.6 |
| 1 | 365 | 14.8 | 406 | 21.7 |
| 2 | 82 | 3.9 | 49 | 2.4 |
| 3 | 25 | 1.2 | 10 | 0.7 |
| 4 | 17 | 0.8 | 5 | 0.2 |
| 5-10 | 27 | 1.7 | 6 | 0.3 |
| 12-25 | 14 | 0.8 | ||
| 27-36 | 20 | 1.4 | 2 | 0.1 |
| missing | 34 | 25 | ||
Tables 2 reports descriptive statistics of independent variables used in the analyses, disaggregated by Latino and Asian groups and by immigrant generations. The background characteristics of respondents show great variations between and within Latino and Asian samples. Compared to the Asian sample, Latino respondents tend to be younger, more likely to be married with children, and have lower socioeconomic resources. The socioeconomic difference is particularly pronounced among the first generation - while 44% of Asian immigrants are college-educated, only 7% of Latino immigrants have received the similar level of education. In terms of language proficiency, the second generation Latino respondents are more likely to retain their language of origin than Asians. In contrast, the first generation Asian respondents tend to have higher English proficiency than Latinos. For both Latinos and Asians, the first generation individuals are generally older than the 2nd and above generation individuals, and they are more likely to be married.
Table 2. Weighted median and % distribution of independent variables by race and immigrant generation (NLAAS 2002-2003).
| Latino Sample | Asian Sample | |||||
|---|---|---|---|---|---|---|
| 1st Gen n=1,265 | 2nd Gen n=728 | 3rd+ Gen n=561 | 1st Gen n=1,402 | 2nd Gen n=423 | 3rd+ Gen n=268 | |
| Age | 37 (s.e.=0.9) | 29 (s.e.=1.1) | 34 (s.e.=1.4) | 43 (s.e.=0.9) | 27 (s.e.=1.4) | 36 (s.e.=2.3) |
| English Proficiency (range 3-12) | 3 (s.e.=0.1) | 9 (s.e.=0.4) | 11(s.e.=0.1) | 8 (s.e.=0.3) | 11 (s.e.=0) | 11 (s.e.=0) |
| Spanish/Asian Language Proficiency (3-12) | 9 (s.e.=0.1) | 8 (s.e.=0.2) | 4 (s.e.=0.2) | 10 (s.e.=0.1) | 4 (s.e.=0.2) | 3 (s.e.=0) |
| Household Income per Family Member | 9,346 (s.e.=781) | 12,497 (s.e.=1,620) | 16,103 (s.e.=1,461) | 24,248 (s.e.=1,571) | 21,852 (s.e.=2,507) | 30,000 (s.e.=1,252) |
| Never Married | 0.22 | 0.38 | 0.35 | 0.12 | 0.55 | 0.41 |
| Male | 0.52 | 0.51 | 0.51 | 0.46 | 0.5 | 0.51 |
| Presence of Dependents under Age 18 | 0.45 | 0.41 | 0.42 | 0.35 | 0.28 | 0.32 |
| Citizenship | 0.29 | 0.8 | 1 | 0.56 | 0.86 | 1 |
| Race | ||||||
| White | 0.35 | 0.35 | 0.37 | |||
| Black | 0.09 | 0.07 | 0.04 | |||
| Other | 0.56 | 0.57 | 0.58 | |||
| Education | ||||||
| Less than high school (<12yrs) | 0.61 | 0.31 | 0.29 | 0.21 | 0.07 | 0.02 |
| HS grad and some college (12-15yrs) | 0.31 | 0.55 | 0.59 | 0.35 | 0.52 | 0.63 |
| College graduate and above (>=16yrs) | 0.07 | 0.14 | 0.11 | 0.44 | 0.41 | 0.36 |
| Ethnic Origin | ||||||
| Cuba (n=578) | 0.07 | 0.06 | 0.01 | |||
| Puerto Rico (n=500) | 0.06 | 0.17 | 0.12 | |||
| Mexico (n=895) | 0.56 | 0.57 | 0.68 | |||
| Other Latin Countries (n=542) | 0.31 | 0.2 | 0.19 | |||
| Vietnam (n=482) | 0.17 | 0.08 | 0 | |||
| Philippines (n=469) | 0.2 | 0.28 | 0.17 | |||
| China (n=568) | 0.33 | 0.25 | 0.14 | |||
| Other Asian Countries (n=444) | 0.3 | 0.38 | 0.69 | |||
Multivariate Analysis
Because of the dichotomous nature of the dependent variables, binary logistic regressions were used to analyze the relationships among the exposure to the US, countries of origin, socioeconomic resources, attachment, demographic backgrounds, and the home engagement. I first disaggregated the total sample into Latino and Asian subsamples, and conducted analyses on two dependent variables separately. In Tables 3 and 4, I compare the following five models for each dependent variable: the first model includes the exposure and countries of ancestral origin; the second model adds the demographic variables to the first model; the third model includes, in addition to the variables in the second model, the socioeconomic resources; the forth model includes English language proficiency and US citizenship; and the final model adds Spanish and Asian language proficiency as a measure of attachment to the country of origin. Multivariate tables display age, income, and language proficiency as continuous variables since preliminary analyses using detailed categorical variables replicated the results of continuous variables.
Table 3. Odds ratios predicting overall frequency of return visits (1=return often/sometime, 0=rarely/never) for Latino and Asian Americans.
| Latino Sample | Asian Sample | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Generation (ref=1st gen) | Model1 | Model2 | Model3 | Model4 | Model5 | Model1 | Model2 | Model3 | Model4 | Model5 |
| 2nd generation | 0.69** | 0.73* | 0.63** | 0.51*** | 0.63* | 0.23*** | 0.23*** | 0.23*** | 0.19*** | 0.27*** |
| 3rd generation | 0.35*** | 0.36*** | 0.30*** | 0.22*** | 0.37*** | 0.14*** | 0.14*** | 0.14*** | 0.11*** | 0.18** |
| Latino Origin (ref=Mexico) | ||||||||||
| Puerto Rico | 1.49* | 1.33 | 1.24 | 1.01 | 1.00 | |||||
| Other Latin Countries | 1.03 | 0.97 | 0.86 | 0.84 | 0.84 | |||||
| Cuba | 0.21*** | 0.16*** | 0.13*** | 0.12*** | 0.11*** | |||||
| Asian Origin (ref=China) | ||||||||||
| Other Asia | 1.53 | 1.50 | 1.44 | 1.23 | 1.19 | |||||
| Philippines | 1.18 | 1.21 | 1.20 | 0.93 | 0.92 | |||||
| Vietnam | 0.45*** | 0.46*** | 0.52** | 0.56** | 0.51** | |||||
| Age | 1.01 | 1.01 | 1.00 | 1.01 | 0.99 | 1.00 | 1.00 | 1.00 | ||
| Male | 1.33* | 1.35* | 1.32* | 1.42** | 1.02 | 0.96 | 0.94 | 0.95 | ||
| Ever Married | 1.19 | 1.17 | 1.19 | 1.12 | 1.24 | 1.05 | 1.04 | 0.98 | ||
| Dependents under 18 | 0.80 | 0.81 | 0.80 | 0.80 | 0.82 | 0.93 | 0.93 | 0.94 | ||
| Race (ref=white) | ||||||||||
| Black | 1.33 | 1.38 | 1.41 | 1.41 | ||||||
| Other | 0.99 | 1.01 | 1.03 | 1.06 | ||||||
| Education (ref=less than HS) | ||||||||||
| High School Graduate | 1.49* | 1.47* | 1.38 | 1.03 | 0.90 | 0.83 | ||||
| Some College | 1.64** | 1.63** | 1.49* | 0.86 | 0.67* | 0.62* | ||||
| College and above | 2.31*** | 2.47** | 2.20** | 1.39 | 1.03 | 0.93 | ||||
| Income | 0.98 | 0.97 | 0.97 | 1.13* | 1.11* | 1.12* | ||||
| English Proficiency | 0.95 | 0.94 | 1.12*** | 1.11*** | ||||||
| Citizenship | 2.17*** | 2.28*** | 0.98 | 1.01 | ||||||
| Spanish/Asian Language | 1.13*** | 1.08* | ||||||||
| N | 2496 | 2495 | 2494 | 2484 | 2484 | 1945 | 1943 | 1942 | 1941 | 1939 |
| -2 Log-Likelihood | 3207.0 | 3164.3 | 3127.5 | 3081.3 | 3035.9 | 2292.4 | 2284.7 | 2244.9 | 2223.1 | 2209.3 |
| DF | 5 | 11 | 15 | 17 | 18 | 5 | 9 | 13 | 15 | 16 |
p<.05,
p<.01,
p<.001 (two-tailed test)
Table 4. Odds ratios predicting past year return visits (1=returned at least once, 0=did not return) for Latino and Asian Americans.
| Latino Sample | Asian Sample | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Generation (ref=1st gen) | Model1 | Model2 | Model3 | Model4 | Model5 | Model1 | Model2 | Model3 | Model4 | Model5 |
| 2nd generation | 0.70** | 0.73* | 0.66** | 0.58** | 0.69 | 0.26*** | 0.20*** | 0.20*** | 0.21*** | 0.32*** |
| 3rd generation | 0.43*** | 0.44*** | 0.39*** | 0.32*** | 0.48* | 0.13*** | 0.12*** | 0.11*** | 0.12*** | 0.21* |
| Latino Origin (ref=Mexico) | ||||||||||
| Puerto Rico | 0.72 | 0.61** | 0.57** | 0.50** | 0.49** | |||||
| Other Latin Countries | 0.76 | 0.68* | 0.62* | 0.61* | 0.6* | |||||
| Cuba | 0.29*** | 0.23*** | 0.20*** | 0.19*** | 0.18*** | |||||
| Asian Origin (ref=China) | ||||||||||
| Other Asia | 1.02 | 0.99 | 0.96 | 0.86 | 0.83 | |||||
| Philippines | 0.72 | 0.75 | 0.74 | 0.64* | 0.64* | |||||
| Vietnam | 0.34*** | 0.35*** | 0.37*** | 0.41*** | 0.37*** | |||||
| Age | 1.00 | 1.01 | 1.00 | 1.00 | 0.99 | 0.99 | 1.00 | 1.00 | ||
| Male | 1.14 | 1.14 | 1.13 | 1.19 | 0.90 | 0.87 | 0.86 | 0.88 | ||
| Ever Married | 1.35 | 1.33 | 1.33 | 1.27 | 0.74 | 0.65 | 0.62 | 0.58* | ||
| Dependents under 18 | 0.63*** | 0.64*** | 0.64*** | 0.64*** | 0.81 | 0.88 | 0.92 | 0.93 | ||
| Race (ref=white) | ||||||||||
| Black | 1.54* | 1.6* | 1.63* | 1.62* | ||||||
| Other | 1.06 | 1.09 | 1.09 | 1.11 | ||||||
| Education (ref=less than HS) | ||||||||||
| High School Graduate | 1.19 | 1.16 | 1.11 | 1.01 | 0.99 | 0.90 | ||||
| Some College | 1.37* | 1.33 | 1.23 | 0.87 | 0.80 | 0.73 | ||||
| College and above | 1.65* | 1.71* | 1.56* | 1.15 | 1.00 | 0.90 | ||||
| Income | 1.00 | 1.00 | 1.00 | 1.09* | 1.09* | 1.09* | ||||
| English Proficiency | 0.98 | 0.96 | 1.06* | 1.05* | ||||||
| Citizenship | 1.57* | 1.63* | 0.69* | 0.72* | ||||||
| Spanish/Asian Language | 1.10*** | 1.09* | ||||||||
| N | 2482 | 2481 | 2480 | 2470 | 2470 | 1942 | 1940 | 1939 | 1938 | 1936 |
| -2 Log-Likelihood | 2698.0 | 2654.5 | 2643.2 | 2619.8 | 2599.0 | 2082.8 | 2064.7 | 2049.0 | 2034.8 | 2019.4 |
| DF | 5 | 11 | 15 | 17 | 18 | 5 | 9 | 13 | 15 | 16 |
p<.05,
p<.01,
p<.001 (two-tailed test)
Exposure
Generational status affects Latinos and Asians differently. For Asian respondents, the higher-order generation is strongly associated with lower odds of return visits: being a second-generation Asian, for example, reduces the odds of frequent home visits by almost 70%, compared to the first generation (Model 5 in Tables 3 and 4). Furthermore, generational status has a gross and net effect on home country visits for Asian Americans, instead of being mediated by socioeconomic resources or attachment. In contrast, Latino respondents' contact with the country of origin is less affected by the generational status when socioeconomic resources and attachment are controlled. The difference between the first- and second-generation Latinos is barely significant (Model 5 in Table 3) or non-significant (Model 5 in Table 4) at a .05 level. For Latino individuals, attachment to the original country mediates some of the negative effects of the generational status on return visits. For both Latino and Asian respondents, the negative relationships between generation and return visits become even stronger when attachment to the US is introduced in Model 4, suggesting that the effect of generational status was suppressed by their attachment to the US.
Countries of origin
Ethnic origin has a strong effect on return visits, regardless of individual characteristics, suggesting the importance of bi-national relationships between the US and the country of ancestral origin. Negative associations between Vietnamese or Cuban origin and home country visits are strong and consistent. Travel restrictions and the large proportion of Vietnamese and Cubans who left the country for political persecutions may account for the low propensity of visiting home for Vietnamese and Cuban respondents.
Resources
The educational attainment and household income significantly increase the odds of home country visits for Latinos and Asians, respectively. As a resource hypothesis suggests, financial resources may allow Asian respondents to purchase airplane tickets and increase the chance to visit one's home country for a visit. For Latino individuals, proximity to countries of origin and the low cost of traveling may enable them to visit their original countries even without financial resources. Instead, high-skilled jobs or valid visa status associated with higher educational attainment may enable Latino respondents to engage in frequent cross-national activities.
Attachment to the U.S.
For Latino respondents, the English language proficiency does not influence the frequency or the number of visits to the original country. As immigrant transnational literature suggests, attachment to the country of ancestral origin may be independent from the cultural adaptation to American society. Contrary to the assimilation hypothesis, citizenship has a positive impact on return visits for Latinos. Even after controlling for generation, being a citizen of the United States almost doubles the propensity to visit the country of ancestral origin. Citizenship for Latino immigrants may indicate a secure legal status is likely to ease the process of border-crossing for Latino individuals in the US. For Asian individuals, English proficiency has a positive impact on return visits. Establishing one's life in American society may motivate Asian Americans to visit the country of origin, net of financial resources.
Attachment to Countries of Ancestral Origin
Spanish and Asian language proficiency has consistent positive effects on return visits, mediating some of the negative effect of generational status. The effect is especially strong and consistent for Latino Americans. Latino individuals' everyday interactions may cut across the generational status, influencing the knowledge of Spanish language across generations (Waters and Jimenez, 2005). Although the immigrant generation appears to influence the frequent return visits, its effect is partially mediated by the degree of attachment to the country origin, especially for Latinos. Also, the effects of attachment to the US are relatively stable after introducing the attachment to the countries of origin, indicating the independent relationship between the home country attachment and the attachment to the US.
Demographic characteristics
Demographic factors, especially a family structure, appear to play an important, independent role in determining the odds of return visits. Having dependents under age 18 (for Latinos), marriage, and obtaining citizenship (for Asians) reduce the likelihood of return visits, especially within the past one year (Table 4). One possible explanation is that presence of immediate family members -- parents, siblings, children, or a spouse -- in the US reduces the need to return to the country of origin, regardless of levels of resources and attachment to the US. A closer temporal link between family structures and past-year return visits seems to explain the stronger associations between family and return visits in the past year. In addition, race influences the odds of return visits for Latino respondents. Self-identified blacks are more likely to visit the country of ancestral origin than self-identified whites, net of countries of origin. Those who experience discrimination in the US may be more likely to retain their racial heritage associated with the country of origin and thus engage in transnational activities (Itzigsohn and Saucedo, 2002; Waters, 1994).
Recall our discussion on the conceptual model informed by assimilation and transnational literature (Figure 2). The classical assimilation hypothesis proposes that exposure to the US is positively associated with assimilation and assumes that ties to the country of origin will decline as immigrants settle into a new country and grow attached to the new society. The immigrant transnational literature alternatively suggested that immigrants will remain attached to the country of origin regardless of their levels of attachment to the new society. In addition, the resource hypothesis drawn from behavioral transnationalism highlighted the positive relationship between socioeconomic resources and transnational involvement.
Data used in this study render some support for each hypothesis. In line with a classical assimilation hypothesis, much of return visits can be attributed to the generational exposure to American society especially for Asians. However, socioeconomic resources and attachment do not mediate the effect of exposure on return visits in the way assumed by the classical assimilation hypothesis. Educational attainment and income are found to increase the chance of return visits for Latinos and Asians, as predicted by the resource hypothesis. Also, the positive net effects of socioeconomic resources on return visits are largely independent of greater exposure to American society. Finally, the independent effects of attachment to the US and attachment to the home country (indexed by language) are consistent with the idea that immigrants remain connected to the country of origin regardless of their level of attachment to the US. Latino and Asian individuals appear to be capable of maintaining multiple attachments despite their settlement in the US. Attachment to the US and attachment to the country of their parents' or grandparents' origin may not be competing concepts.
Beyond the debates on assimilation and transnationalism, the results highlight different mechanisms for home country visits for Latino and Asian individuals residing in the US. First, because of proximity to most of Latin countries, Latino individuals' return visits are less likely to be affected by economic resources. Instead, a documented legal status which guarantees their return to the US appears to encourage their border-crossings. In addition, highly educated males with Spanish language proficiency are more likely to travel back and forth between two countries, suggesting that Latino return visits may be related to high-skilled jobs (Guarnizo et al., 2003). Second, compared to Asian Americans, the effect of generational status is less pronounced for Latinos. As Foner (2005) suggested, the low cost of airfare and the constant inflow of new immigrants from Latin America may sustain the transnational field for second and later generations. In contrast, geographical distance and cost of travel to Asian countries appears to make Asian Americans' return visits a financially costly activity. Despite the very strong, negative effect of generational status, household income consistently increases the odds of frequent home country visits. Financial resources may enable Asian individuals to purchase plane tickets and take vacation time from work to visit the country of origin. In addition, citizenship and marriage seem to indicate settlement in the US and constrain Asian individuals' return visits, whereas for Latino Americans, it is the presence of children under age 18 that limits their visits to their countries of origin.
Conclusions
Previous literature on immigrant transnationalism tended to posit the relationship between assimilation and transnationalism at opposite extremes. An assimilation perspective, for example, drew an image of immigrants' cross-national engagement declining as they became assimilated to the US. Portes and his colleagues' studies, on the other hand, suggested that the more assimilated immigrants are, the more likely they are to engage in transnational activities (Portes et al., 2002). Still others indicated that assimilation and transnationalism may be independent processes (Glick Schiller et al., 1995; Levitt, 2001). Because these perspectives did not engage much with each other, they were often treated as competing paradigms.
The first part of this paper analytically synthesized the three perspectives by distinguishing resources from attachment, instead of treating them as the same concept, i.e. assimilation. By acknowledging the two mediating paths between exposure to the host society and transnational home engagement (i.e. frequent return visits to the home country), the model incorporated the three types of relationships between assimilation and transnationalism: 1) having socioeconomic resources increases the chance of home country visits; 2) exposure to American society reduces the home country visits via attachment to the US; and 3) attachment to the country of origin influences return visits independent of the level of attachment to the US.
Empirical tests using the nationally representative sample of Latino and Asian Americans found some support for the three hypotheses. Hypothesis 1 was supported in a sense that the income and educational attainment increase the frequency of return visits for Asians and Latinos respectively. However, the effect of exposure in the US is not mediated by the positive effect of resources as assumed by classical assimilation literature. Hypothesis 2 was partially supported because generational status reduced the odds of return visits, especially for Asian Americans. However, data contradicted hypothesis 2 in a sense that most of the measures of attachment to the US had either non-significant or positive effects on return visits. In support of hypothesis 3, attachment to the country of origin increased return visits independently of the level of attachment to the United States.
There are several implications for the literature of immigrant assimilation and transnationalism. First, Latino as well as Asian individuals in the US are capable of maintaining multiple loyalties and attachments. This view is different from the image portrayed by the classical assimilation literature. The classical assimilation perspective assumed that immigrant groups' ties with the country of origin are incompatible with their new ties with the destination country (Gordon, 1964). This study suggests that contacts to the country of ancestral origin can be maintained without compromising attachment to their new society. Second, validity of generation as an index of exposure needs to be re-examined especially for Latino individuals. Given the high retention of the Spanish language among the second-generation Latinos, examining the generational and racial/ethnic compositions of neighborhood may shed some light on the nature of everyday interactions for Latino individuals.
Third, the analysis suggests the influence of national contexts on patterns of transnational engagement. Although immigrants and their descendents maintain connections across borders, this does not mean that the nation-state is irrelevant. Patterns of home country engagement are shaped by political and institutional arrangements between countries. Capturing immigrant experiences through a transnational lens is important, but it may still require a careful attention to the national-level institutions. Forth, determinants of transnationalism are more complicated than previous literature suggested. Analyses of transnationalism tended to simplify the relationship between assimilation and transnational involvement. Some have argued that immigrants who are settled in the US are more likely to engage in cross-national activities (Portes et al., 2002), while others have suggested that transnationalism in general is not associated with any particular types of immigrant individuals (Glick Schiller et al., 1995). The current study suggests a more detailed understanding of the relationship for different racial groups. While generational status reduces propensity of return visits especially for Asian individuals, attachment and socioeconomic resources have independent effects on home country visits: financial resources and educational level are important in increasing the home country visits for Asians and Latinos, respectively, while attachment to the US does not reduce levels of transnational home engagement for both Asian and Latino individuals. A nuanced framework may be necessary to recognize the various mechanisms of immigrant adaptation and transnationalism.
The limitation of this study provides prospects for future research. In particular, the current model needs to be examined beyond the United States. In the context of Europe, for example, immigrant networks and organizations across countries may be important resources in facilitating transnational activities, net of economic resources and attachment (Faist, 2003). Adaptation to the political process of destination society may also constitute another mechanism linking assimilation and transnational engagement. Political struggles over citizenship and representation of ethnic identity in Western European countries suggest that such institutional assimilation (Kastoryano, 2003) could facilitate frequent border-crossings and continued attachment to one's ethnic origin.
Footnotes
The data analysis conducted for this paper was made possible by the Robert Wood Johnson Foundation Research Grant # K23 DA018715-01A2. The project was also supported by NIH research grants R01HD049142, U01 MH62209, #P50 MH073469, P50MH073511, as well as the Substance Abuse & Mental Health Services Administration/Center for Mental Health Services and the Office of Behavioral and Social Science Research. The author would like to thank Charles Hirschman, David Takeuchi, Sara Curran, Julie Brines, and the anonymous reviewers for their helpful comments.
If there is more than one eligible respondent, a single respondent was randomly selected by the interviewer. For the detailed description, see Heeringa et al. (2004).
Other Latino ethnicity includes Dominican Republic among others.
Other Asian ethnicity includes Indian, Japanese, and Korean among others.
References
- Alba R, Nee V. Rethinking assimilation theory for a new era of immigration. International Migration Review. 1997;31(4):826–874. [PubMed] [Google Scholar]
- Alba R, Nee V. Remaking the American mainstream: Assimilation and contemporary immigration. Harvard University Press; Cambridge: 2003. [Google Scholar]
- Brubaker R. The return of assimilation? Changing perspectives on immigration and its sequels in France, Germany, and the United States. Ethnic and Racial Studies. 2001;24(4):531–548. [Google Scholar]
- Duval DT. Linking return visits and return migration among Commonwealth Eastern Caribbean migrants in Toronto. Global Networks-a Journal of Transnational Affairs. 2004;4(1):51–67. [Google Scholar]
- Faist T. The Border-Crossing Expansion of Social Space: Concepts, Questions and Topics. In: Faist T, Özveren E, editors. Transnational social spaces: agents, networks, and institutions. Aldershot, Hants, England; Burlington, VT: Ashgate; 2004. pp. 1–34. [Google Scholar]
- Felixortiz M, Newcomb MD, Myers H. A Multidimensional Measure of Cultural Identity for Latino and Latina Adolescents. Hispanic Journal of Behavioral Sciences. 1994;16(2):99–115. [Google Scholar]
- Foner N. From Ellis Island to JFK: New York's Two Great Waves of Immigration. Yale University Press and Russell Sage Foundation; New Haven and New York: 2000. [Google Scholar]
- Foner N. In a New Land: a comparative view of immigration. New York University Press; New York: 2005. [Google Scholar]
- Gans HJ. Toward a reconciliation of “assimilation” and “pluralism”: The interplay of acculturation and ethnic retention. International Migration Review. 1997;31(4):875–892. [PubMed] [Google Scholar]
- Glick Schiller N. Transmigrants and Nation-States: Something Old and Something New in the U.S. Immigrant Experience. In: Hirschman C, Kasinitz P, Dewind J, editors. The handbook of international migration: the American experience. New York: Russell Sage Foundation; 1999. pp. 94–119. [Google Scholar]
- Glick Schiller N, Basch L, Blanc-Szanton C. Towards a transnational perspective on migration: race, class, ethnicity, and nationalism reconsidered. New York Academy of Sciences; New York, N.Y.: 1992. [Google Scholar]
- Glick Schiller N, Basch L, Blanc-Szanton C. From Immigrant to Transmigrant - Theorizing Transnational Migration. Anthropological Quarterly. 1995;68(1):48–63. [Google Scholar]
- Gordon MM. Assimilation in American life: the role of race, religion, and national origins. Oxford University Press; New York: 1964. [Google Scholar]
- Guarnizo LE, Portes A, Haller W. Assimilation and transnationalism: Determinants of transnational political action among contemporary migrants. American Journal of Sociology. 2003;108(6):1211–1248. [Google Scholar]
- Heeringa SG, Wagner J, Torres M, Duan NH, Adams T, Berglund P. Sample designs and sampling methods for the Collaborative Psychiatric Epidemiology Studies (CPES) International Journal of Methods in Psychiatric Research. 2004;13(4):221–240. doi: 10.1002/mpr.179. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Itzigsohn J, Saucedo SG. Immigrant incorporation and sociocultural transnationalism. International Migration Review. 2002;36(3):766–798. [Google Scholar]
- Kasinitz P, Waters MC, Mollenkopf JH, Anil M. Transnationalism and the Children of Immigrants in Contemporary New York. In: Levitt P, Waters MC, editors. The Changing Face of Home: The Transnational Lives of the Second Generation. New York: Russel Sage Foundation; 2002. pp. 96–122. [Google Scholar]
- Kastoryano R. Negotiating identities: states and immigrants in France and Germany. Princeton University Press; Princeton: 2002. [Google Scholar]
- Levitt P. The transnational villagers. University of California Press; Berkeley: 2001. [Google Scholar]
- Levitt P, Dewind J, Vertovec S. International perspectives on transnational migration: An introduction. International Migration Review. 2003;37(3):565–575. [Google Scholar]
- Levitt P, Jaworsky N. Transnational migration studies: Past developments and future trends. Annual Review of Sociology. 2007;33:129–156. [Google Scholar]
- Louie A. Creating histories for the present: second-generation (re)definitions of Chinese American culture. In: Levitt P, Waters MC, editors. The changing face of home: the transnational lives of the second generation. New York: Russell Sage Foundation; 2002. pp. 312–340. [Google Scholar]
- Massey DS. Dimensions of the new immigration to the United-States and the prospects for assimilation. Annual Review of Sociology. 1981;7:57–85. doi: 10.1146/annurev.so.07.080181.000421. [DOI] [PubMed] [Google Scholar]
- O'flaherty M, Skrbis Z, Tranter B. Home visits: Transnationalism among Australian migrants. Ethnic and Racial Studies. 2007;30(5):817–844. [Google Scholar]
- Portes A, Escobar C, Radford AW. Immigrant transnational organizations and development: A comparative study. International Migration Review. 2007;41(1):242–281. [Google Scholar]
- Portes A, Guarnizo LE, Landolt P. The study of transnationalism: pitfalls and promise of an emergent research field. Ethnic and Racial Studies. 1999;22(2):217–237. [Google Scholar]
- Portes A, Haller WJ, Guarnizo LE. Transnational Entrepreneurs: An Alternative Form of Immigrant Economic Adaptation. American Sociological Review. 2002;67(2):278–298. [Google Scholar]
- Portes A, Rumbaut RG. Immigrant America: a portrait. University of California Press; Berkeley: 2006. [Google Scholar]
- Rumbaut RG. Severed or sustained attachments? Language, identity, and imagined communities in the post-immigrant generation. In: Levitt P, Waters MC, editors. The changing face of home: the transnational lives of the second generation. New York: Russell Sage Foundation; 2002. pp. 43–95. [Google Scholar]
- Sana M. Buying membership in the transnational community: migrant remittances, social status, and assimilation. Population Research and Policy Review. 2005;24(3):231–261. [Google Scholar]
- Thomas WI, Znaniecki F, Zaretsky E. The Polish peasant in Europe and America. University of Illinois Press; Urbana: 1984. [Google Scholar]
- Waldinger R. Between Here and There: How Attached Are Latino Immigrants to Their Native Country? The Pew Hispanic Center; Washington, DC: 2007. [Google Scholar]
- Waters MC. Ethnic and Racial Identities of Second-Generation Black Immigrants in New York City. International Migration Review. 1994;28(4):795–820. [Google Scholar]
- Waters MC, Jimenez TR. Assessing immigrant assimilation: New empirical and theoretical challenge. Annual Review of Sociology. 2005;31:105–125. [Google Scholar]
- World Bank. Global Economic Prospects: Economic Implications of Remittances and Migration. World Bank; Washington DC: 2006. [Google Scholar]
- Yang PQ. Ethnicity and Naturalization. Ethnic and Racial Studies. 1994;17(4):593–618. [Google Scholar]
- Zhou M. Segmented assimilation: Issues, controversies, and recent research on the new second generation. International Migration Review. 1997;31(4):975–1008. [PubMed] [Google Scholar]