Is the most frequently cited cultural value in explaining Jewish upward mobility

Ethn Racial Stud. Author manuscript; available in PMC 2011 Jul 1.

Published in final edited form as:

PMCID: PMC2882294

NIHMSID: NIHMS179065

Abstract

This article examines the debate between key theories of immigrant assimilation by exploring the effect of acculturation types – dissonant, consonant, and selective – on socioeconomic outcomes in young adulthood. Drawing on survey data from the Immigrant Second Generation in Metropolitan New York, we show that while all three types occur, dissonant acculturation is the exception, not the norm, among second generation young adults. Our results also suggest that neither the type of acculturation nor the level of ethnic embeddedness can account for the variation in mobility patterns both across and within second generation groups. These findings lead us to question assumptions about the protective effect of selective acculturation and the negative effect of dissonant acculturation.

Keywords: Straight-line assimilation, segmented assimilation, second-generation, acculturation, socioeconomic mobility, young adulthood

Introduction

The study of the “new second generation” – children born to post-1965 immigrants in the U.S. – has expanded rapidly as they have entered adulthood. Understanding how they are integrating into American society is both theoretically important and a key policy issue. Since the children of immigrants are now one tenth of the American population and one fifth of those under 18, their fate is enormously important to the future of the country. It is important not only to understand trends for the second generation as a whole, but why some of its members are succeeding and others doing poorly. What factors lead the children of immigrants to do better than their parents and what factors lead to downward social mobility? Drawing on data from the study of the Immigrant Second Generation in Metropolitan New York1, this paper examines how types of acculturation shape socioeconomic outcomes among young adult respondents.

Theories of straight-line and segmented assimilation revisited

Two major theories – straight-line assimilation and segmented assimilation – point to different processes underlying second generation outcomes. The standard assimilation theory is associated with the founders of the Chicago School of Sociology, who studied the integration of the first and second generation European immigrants in the early 20th century (Park and Burgess 1925). This model argues that assimilation processes will enable each succeeding generation to show upward social mobility in education and occupation, be more integrated into the American mainstream, and show less ethnic distinctiveness in language use, residential concentration, and intermarriage patterns (Warner and Srole 1945).

Segmented assimilation theory emerged as an alternative to this model in the 1990s and has been enormously influential. Formulated by Alejandro Portes and his collaborators and elaborated and tested empirically by Portes and Ruben Rumbaut (Portes and Zhou 1993; Portes and Rumbaut 2001), this approach argues that starkly different outcomes are possible for the second generation. Its members can end up “ascending into the ranks of a prosperous middle class or join in large numbers the ranks of a racialized, permanently impoverished population at the bottom of society” (Portes, Kelly and Haller 2005:1004).

Segmented assimilation theory posits three possible outcomes for the second generation: upward assimilation, downward assimilation, and upward mobility combined with persistent biculturalism. These paths correspond to three processes that summarize the relations between immigrant children, their parents, and the wider ethnic community – consonant, dissonant, and selective acculturation. Consonant acculturation occurs when the children and parents both learn American culture and gradually abandon their home language and “old country” ways at about the same pace. As these children enter the American mainstream, they achieve upward mobility with the support of their parents. Dissonant acculturation occurs when children learn English and adopt American ways far faster than do their immigrant parents. Portes and Rumbaut (2001) argue that this process can lead to downward assimilation when young people confront racial discrimination, bifurcated labor markets, and often nihilistic inner city young people on their own, without strong parental authority or community support. The third process, selective acculturation, leads to upward assimilation and biculturalism. This occurs when parents and children both gradually learn American ways while remaining embedded, at least in part, in the ethnic community. It is characterized by “preservation of parental authority, little or no intergenerational conflict, and fluent bilingualism among children” (Portes and Rumbaut 2001:52). Portes and his collaborators argue that selective acculturation is especially important for groups facing discrimination

…because individuals and families do not face the strains of acculturation alone but rather within the framework of their own communities. This situation slows down the process while placing the acquisition of new cultural knowledge and language within a supportive context. (Portes and Rumbaut 2001:54)

Segmented assimilation theory also stresses the importance of parental human capital (including parents' education and income), modes of incorporation (state definitions of immigrant groups, eligibility for welfare, degree of discrimination and antipathy towards immigrant groups), and family structure (single vs. married couple families as well as multigenerational vs. nuclear family living arrangements). Although less explicitly stated, the model also points to the varying degrees of transnational connection among immigrant groups as an important element of the context of reception.

This theory has inspired a large volume of work on immigrant incorporation. The concept of “modes of incorporation”, for instance, has been extremely useful in systematizing how varying political and cultural reactions to immigrant groups shape their individual experiences. Yet the most innovative causal mechanism of the theory – selective acculturation – has not been as closely examined as one would expect. This is unfortunate because this aspect of the segmented assimilation approach most clearly separates it from other accounts of immigrant incorporation. After all, standard sociological models of status attainment predict that children from two-parent households will have better outcomes, as whites will compared to Blacks and Hispanics, given the reality of ongoing racial discrimination in the U.S. It also predicts that the children of parents with high levels of education and income will do well, on average.

Where segmented assimilation departs from these standard interpretations is in predicting two specific outcomes as in part the result of intra-family dynamics – that downward assimilation occurs not because of the failure to Americanize, but of doing it too quickly (dissonant acculturation), and that upward mobility is possible for those with low income or poorly educated parents who stay at least partially tied to the “ethnic” community. In these two predictions, segmented assimilation stands the standard sociological account of assimilation on its head. For at least some immigrants, it argues that quickly coming to share American (or at least lower class American) ways is bad for the second generation, while holding on to immigrant distinctiveness can turn out to be an advantage.

In response to this approach, Alba and Nee (2003) formulated a new version of (more or less) straight line assimilation for the post-1965 immigrants. Retaining many key insights form earlier theorists, Alba and Nee predict that most members of the contemporary second generation will experience gradually increasing social integration and upward mobility. In contrast to the segmented assimilation model, they find little support for the notions that many will experience “downward assimilation” or that embeddedness in dense ethnic networks will prove beneficial. In contrast to earlier versions of straight line assimilation, however, Alba and Nee reject the overly prescriptive assertion that the second generation must adopt “American norms” and stress that the American “mainstream” is highly dynamic and heterogeneous. Drawing on segmented assimilation and other contemporary accounts, they understand the variation in immigrants' pre-migration backgrounds, current positions within a highly stratified American society and emphasize the importance of historically contingent contexts of reception.

Types of acculturation and socioeconomic mobility

Many studies in the U.S. and increasingly in Europe show support for the theory of segmented assimilation, including work arising from the Children of Immigrants Longitudinal Study (CILS) and important case studies of particular second generation groups including Vietnamese (Zhou and Bankston 1998), West Indians (Waters 1999; Vickerman 1999), Chinese (Zhou and Kim 2006), Salvadorans (Menjívar 2000), and Haitians (Stepick 1998; Zéphir 2001). There is close to universal agreement that American society is not an undifferentiated whole – and in that sense immigrants clearly assimilate into one of its segments. Virtually all studies show that the children of immigrants do not follow a single trajectory and that second generation outcomes are highly contingent on the segment of American society into which they are being incorporated (Greenman and Xie 2008).

That said, one key point of disagreement between Alba and Nee's reformulated assimilation theory and segmented assimilation theory is whether the processes and mechanisms that led the previous waves of European immigrants in the early part of the 20th century to successfully integrate will work for the current wave of immigrants and their children. Indeed, while segmented assimilation claims that the non-white racial status of most current immigrants and the very different economy they face in 21st century America puts them and their children at greater risk, Waldinger and Perlmann (1998), Perlmann and Waldinger (1997) and Gratton (2002) have all criticized segmented assimilation's characterization of past. Further, many studies of the contemporary second generation find little evidence of second generation decline or downward assimilation (Boyd, 20020; Farley and Alba (2002), Hirschman (2001), Smith (2003), Waldinger and Feliciano (2004), and Kasinitz et al. (2008, 2004, 2002).

Of course, segmented assimilation does not predict universal downward mobility any more than classic assimilation predicts universal upward mobility. Alba and Nee (2001) and Portes and Zhou (1993) both argue that some members of the second generation will do well compared to their parents, while others will not. Both theories posit that racial discrimination will make it much more difficult for those defined as non-white to achieve upward mobility in America's racially stratified economy. The difference between the two approaches lies in which mechanisms the theories suggest lead to successful outcomes for the second generation. Alba and Nee (2003) posit that similar historical processes will blur the differences between the immigrant groups and the mainstream. Segmented assimilation argues that, especially for non-white poor immigrants, maintaining ethnic differences with the American mainstream – selective acculturation – will lead to successful outcomes for the second generation.

Scholars have not yet put this key difference to a rigorous test. In a recent special issue on the European second generation, Crul and Thomson (2007:1036) conclude that “upward mobility through ethnic cohesion remains a limited phenomenon in the European context.” In the U.S., the lack of intergenerational data and the lack of a Census question on parents' birthplace have hampered quantitative research on this question, although evidence for selective acculturation has accumulated through qualitative case studies. Aside from the longitudinal data provided by CILS, no studies of the second generation have addressed this question.

This paper takes advantage of the Immigrant Second Generation in Metropolitan New York Study, which includes information on the ethnic embeddedness in neighborhoods and the labor market as well as detailed information on the language abilities of the first and second generation, to examine the relationships between these factors and second generation mobility pathways. Since predictions about the adaptation trajectories of the new second generation have mostly been speculative, this analysis provides one of the first tests of the effect of acculturation types on second generation outcomes.

Data and methods

The Immigrant Second Generation in Metropolitan New York Study is a random telephone survey of 3,415 respondents who live in New York City and the inner suburbs in New Jersey, Westchester, and Long Island. The study includes young adults aged 18 to 32 at the time of the interview in 1999-2000 who were born in the U.S. to parents who immigrated after 1965 (the second generation) or who were born abroad but arrived in the U.S. by age 12 and grew up in the U.S. (the “1.5 generation”). The study includes five second generation groups (West Indian, Dominican, Chinese, South American2 and Russian Jews) and three native-born comparison groups (white, black and Puerto Rican). This analysis is limited to respondents from the five groups whose parental language is not English and thus we do not include native whites, blacks, and West Indians, although we present descriptive data on them in Table 1.

Table 1

Socioeconomic Outcomes by Ethnic Group

The dependent variables include seven measures of socioeconomic outcomes in young adulthood. All measures are dichotomous, based on self-reported levels of education, occupation, employment status, teen parent status, as well as history of arrest and incarceration. Because we are interested in understanding how different acculturation types lead to negative and positive adaptation trajectories, we investigate seven outcomes: being a high school dropout, being unemployed, having been arrested, having been incarcerated, being a teen parent by age 18, being a college graduate by age 22, and being in a professional/entrepreneurial occupation by age 25. Table 1 provides the frequency distribution for each socioeconomic outcome by group. It highlights systematic variations across the eight groups. Chinese, Russian Jews, and whites report the lowest rates of dropping out of high school, unemployment, teen pregnancy, or having an arrest or incarceration record, whereas Dominicans, Puerto Ricans and African Americans report the highest rates. Chinese and Russian Jews are also the most likely to be college educated and in a professional occupation.

We use four clusters of independent variables to explain variations in socioeconomic outcomes. We operationalize these variables to best capture the three factors that segmented assimilation theory posits as important determinants of adaptation outcomes across and within groups: human capital, family structure, and mode of incorporation among the immigrant parents.

Demographic characteristics include national origin, age, gender, and immigrant generation status (1.5- generation vs. second-generation).

Acculturation type is a recoded categorical variable with three values (1=Dissonant acculturation; 2=Consonant acculturation and 3=Selective acculturation).

Ethnic embeddedness include whether respondent's parent or respondent works in an ethnically-concentrated industry, whether respondent belongs to an ethnic organization or watches ethnic media programs.

Background characteristics include respondent's high school GPA and highest level of education, parental education, number of earners in household while growing up (a proxy for the level of financial resources available), family structure, number of siblings, and residential stability which was measured by the times respondent moved between the ages of 6 to 18.

Across the seven outcomes, we apply five nested logistic regression models and report the odds ratios along with robust standard errors for ease of interpretation. For each outcome, the first model only includes ethnicity to establish the initial outcome differences across the five groups. We use Puerto Rican as the reference group because it is a native minority group. The subsequent four models introduce demographic characteristics, acculturation types, ethnic context, and background characteristics to further account for individual variations. Recognizing the potential heterogeneity of effects across groups, we ran the same nested models on each outcome by ethnic group as well as on a sample that excludes Puerto Ricans, but these additional analyses did not lead to substantively different results (full results available upon request).

Operationalizing types of acculturation

Portes and Rumbaut (2001) summarize their research on the second generation in Miami and San Diego by arguing that the dynamics between parents and children in immigrant families crucially shape second generation outcomes. They determine the type of acculturation by comparing the relative fluency of English and immigrant language among the first-generation parents and their second generation children (Portes and Rumbaut 2001:145). Because our study contains detailed information about what our respondents report about their own language fluency as well as that of their parents, we have the unique opportunity to operationalize this variable and explore its potential effects.

Figure 1 provides our detailed coding strategy for a three-category ordinal measure of parents' fluency of English: those who are not proficient, those who have limited proficiency, and those who are fluent. To code respondents' knowledge of the parental language, we rely on two questions on the respondent's ability to speak and understand the parental language. The parental languages include Chinese, Russian, and Spanish. Figure 2 outlines the coding strategy for the categorical variable on respondent's fluency of the parental language. Finally, we combine these two measures to construct a categorical variable for types of acculturation: dissonant, consonant and selective acculturation (See Figure 3). Our regression models use “dissonant acculturation” as the reference category because we are interested in testing the prediction that consonant and selective acculturations yield better socioeconomic outcomes.

Coding Strategies for Parents' Knowledge of English

Notes: 1. “P” refers parents' knowledge of English.

Coding Strategies for Respondent's Knowledge of Parental Language

Notes: 1. “R” refers respondents' knowledge of parental language. From a theoretical standpoint, understanding and speaking abilities are most relevant to the intergenerational dynamics within the immigrant family.

Generational Language Knowledge and Types of Acculturation1

Notes:

1. Source: Adapted from Portes and Rumbaut (2001:145), Figure 6.6.

2. We also coded this variable using only mother's knowledge of English or only father's knowledge of English, but none of these alternative coding strategies yield substantially different results

Table 2 presents descriptive statistics on acculturation types by group. These results make clear that selective acculturation is the norm whereas dissonant acculturation is quite exceptional. Respondents reported high rates of selective acculturation, with 69.5 percent both speaking and understanding the parental language well. An overwhelming majority of Russian Jews (78.9 per cent), South Americans (82.1 per cent), and Dominicans (83.7 per cent), and to a lesser extent Puerto Ricans (51 per cent) and Chinese (59.5 per cent), can be classified as having selectively acculturated. Puerto Ricans and Chinese reported relatively high rates of dissonant acculturation (20.8 per cent and 10.7 per cent) while Dominicans, Russian Jews, South Americans reported low rates (6.5 per cent, 5.2 per cent and 4.9 per cent). Finally, consonant acculturation is most common among Chinese (29.8 per cent) and Puerto Ricans (28.2 per cent) and least common among Dominicans (9.8 per cent) and South Americans (12 per cent).

Table 2

Types of Acculturation by Ethnic Group

In the following analyses, we test the effect of type of acculturation – the main causal mechanism suggested by Portes and Rumbaut (2001:145) – on a series of outcomes. We also examine whether measures of ethnic embeddedness – having parents working in the ethnic enclave, belonging to ethnic organizations, or consuming ethnic media – make a difference in second generation outcomes. While there are many ways to measure this concept, our choice of language practices of parents and children directly follows the path outlined by Portes and Rumbaut (2001).

Multivariate analyses

Table 3 presents logistic regression results on the likelihood of being a high school dropout. In the uncontrolled model, all the second generation groups are significantly less likely to drop out than Puerto Ricans. These initial differences remain significant even after controlling for demographic factors, acculturation types, the extent of ethnic embeddedness and other background characteristics. In the final model, a second-generation respondent is almost half as likely to drop out of high school as a 1.5-generation respondent (Model 5: OR=1.79, p<.05). Type of acculturation has no effect on the likelihood of being a high school dropout – while respondents with consonant and selective acculturation are slightly less likely to drop out than those with dissonant acculturation, the differences are not significant (Model 5: OR=0.9, ns; OR=.69, ns). Nor do any of the measures of ethnic embeddedness carry predictive power, suggesting that ethnic community is neither protective nor detrimental. As expected, parental education and family resources significantly protect against dropping out (Model 5: OR=.66, p<.01; OR=.76, p<.05; OR=.84, p<.05), whereas the number of siblings in the household while growing up is associated with a higher dropout rate (Model 5: OR=1.12, p<.05). This finding confirms previous research on the importance of parental education on socioeconomic outcomes among the second generation (Zhou and Xiong 2005; Abada et al. 2009).

Table 3

Logistic Regressions Predicting the Likelihood of High School Dropout

Table 4 shows that Russian Jews, Chinese, and South Americans are significantly less likely to be unemployed than Puerto Ricans. These differences remain strong after controlling for other demographic factors, whereas acculturation types and ethnic embeddedness have no significant effect. In the final model, women are twice as likely to be unemployed than men (Model 5: OR=2.08, p<.001), while those with more education are less likely to be out of work (Model 5: OR=.52, p<.001). Education explains the initial advantage among Russian Jews and Chinese over Puerto Ricans (Model 5: OR=.55, ns; OR=.53, ns), but not that of South Americans (Model 5: OR=.43, p<.05).

Table 4

Logistic Regressions Predicting the Likelihood of Being Unemployed

Table 5 presents results on the likelihood of having ever been arrested. Without other controls, Chinese and Russian Jewish males are less likely to have been arrested than their Puerto Rican counterparts (Model 1: OR=.30, p<.01; OR=.25, p<.01). Once again, acculturation types, level of ethnic embeddedness, and family background factors have no significant effect. Consistent with established findings in criminological research, age and educational level are strong predictors of having been arrested and account for the initial differences in arrest rate across the ethnic groups.

Table 5

Logistic Regressions Predicting the Likelihood of Being Arrested

Table 6 reports results on the likelihood of having ever been incarcerated and show that Chinese and Russian Jewish males have significantly lower incarceration rates compared to Puerto Ricans. However, these differences disappear after accounting for the Chinese and Russian Jews' relative advantage in education (Model 5: OR=.74, p<0.001; OR=.58, p<.01), whereas acculturation types and level of ethnic embeddedness have neither protective nor detrimental effect. In terms of background characteristics, residential stability clearly matters, as each additional move between the ages of 6-18 is associated with a higher likelihood of having been incarcerated (Model 5: OR=1.16; p<.001). (Of course, this pattern may stem from unmeasured family background factors that cause residential instability, rather than the mere fact of moving, but the destabilizing impact of moving frequently cannot be dismissed.)

Table 6

Logistic Regressions Predicting the Likelihood of Being Incarcerated

Table 7 presents logistic regression results on the likelihood of being a teen parent. Without any control, Chinese, Russian Jews, and South American females are significantly less likely to have a child before the age of eighteen than Puerto Ricans. Acculturation types and co-ethnic context hardly matter, though watching ethnic media is associated with a lower likelihood of being a teen parent (Model 5: OR=.67, p<.05). In terms of family background, the number of siblings and residential instability are associated with a slightly higher probability of teen parenthood (Model 5: OR=1.18, p<.05; OR=1.11, p<.05). Respondent's level of education also matters and explains away the initial difference between South Americans and Puerto Ricans, though major differences persist between Chinese/Russian Jews and Puerto Ricans, suggesting that other cultural factors are at play.

Table 7

Logistic Regressions Predicting the Likelihood of Teen Parent

Segmented assimilation also posits that selective acculturation should yield positive outcomes. Table 8 reports results on the likelihood of being a college graduate among respondents ages 22 or older. (If respondents went to college immediately after high school, they should be able to obtain their degree by age twenty-two.3) Without any controls (Model 1), the second generation groups display large ethnic gaps in college attainment compared to Puerto Ricans, with Chinese being 8 times more likely and Russian Jews 6 times more likely to have a college education. “Consonant” respondents are significantly more likely than “dissonant” ones to attain a degree whereas selective acculturation has no significant effect (Model 3: OR=2.32, p<.05; OR=1.86, ns). Respondents' high school GPA and mother's education account for this initial advantage among “consonant” respondents, however (Model 5: OR=1.31, p<.001; OR=1.49, p<.001). The number of siblings while growing is negatively associated with the likelihood of getting a college education (Model 5: OR=0.86, p<.05), suggesting that the presence of other siblings might result in fewer resources devoted to the respondent's own educational investment. In the final model, Chinese are the most likely to have attained a college education whereas Puerto Ricans are the least likely to do so. This divergence in outcome is striking in light of the fact that both groups reported high rates of dissonant acculturation – 10.7 per cent for Chinese and 20.8 for Puerto Ricans (see Table 2), suggesting that “dissonant acculturation” alone cannot account for differential educational trajectories across groups.

Table 8

Logistic Regressions Predicting the Likelihood of Being a College Graduate

Table 9 presents logistic regression results on the likelihood of being in a professional occupation by age twenty-five. Overall, Russian Jews and Chinese are significantly more likely than Puerto Ricans to be in a professional occupation but their educational advantage entirely explains these initial differences (Model 5: OR=1.58, ns; OR=1.37, ns). Acculturation types and the co-ethnic context have no effect, though respondents' participation in an ethnic organization is positively associated with professional attainment (Model 5: OR=2.06, p<.05). Consistent with past findings, respondent's education is the single most important predictor of occupational attainment5 (Model 5: OR=2.38, p<.001).

Table 9

Logistic Regressions Predicting the Likelihood of Being in a Professional Occupation

Discussion and Conclusion

Three key findings from our analyses of the effects of acculturation types on second generation socioeconomic outcomes in New York City should be highlighted. First, in light of the pervasive concerns that dissonant acculturation will have a negative impact on the second generation, we note that dissonant acculturation is the exception, not the norm, among the new second generation. Though groups vary, only 10 per cent of our respondents experience dissonant acculturation, whereas 20 per cent can be classified as consonant and another 70 per cent as selective acculturation. This alone should lessen concern about intergenerational dynamics within these groups because an overwhelming 90 per cent of our respondents have few problems communicating with their parents either in English or their native language.

Second, types of acculturation do not seem to matter much for socioeconomic outcomes among the second generation. The most interesting predictions of segmented assimilation are that dissonant acculturation leads to negative adaptation outcomes whereas selective acculturation leads to higher educational and professional attainments. Our multivariate analyses provide no support for either prediction. In fact, type of acculturation hardly matters for any of the outcomes that we examined.

Segmented assimilation also argues that the context and level of ethnic embeddedness matter, above and beyond what happens within the immigrant family between first-generation parents and their second generation children. Our analyses tried to capture these contextual effects with several clear measures of embeddedness – parents working in an ethnically concentrated industry or respondents participating in ethnic organizations or consuming ethnic media. Few of these measures matter and none explains away initial differences in outcomes either across groups or within them.6 These findings challenge the assumption among immigration researchers that the type of acculturation is a key mechanism explaining divergent outcomes across groups (Portes and Rumbaut 2001:145).

Further, segmented assimilation is thought to stem from a labor market bifurcated between high paid professional jobs requiring a great deal of education and low paid, low skilled, dead end jobs at the bottom. As Portes, Kelly and Haller (2005:1005) put it:

The promise of American society, which makes so many foreigners come, lies in the access it provides to well remunerated professional and entrepreneurial careers and the affluent lifestyles associated with them. At the same time, it is obvious that not everyone gains access to those positions and that, at the opposite end of society, there is a very unenviable scenario of youth gangs, drug dictated lifestyles, premature childbearing, imprisonment and early death. Immigrant families navigate between these opposite extremes seeking to steer their youths in the direction of the true mainstream.

Yet, among the second generation New Yorkers we studied, the large majority are neither high school dropouts nor medical doctors. More typical are young people with some college education working in white-collar clerical or service industries. Indeed, members of the second generation more closely resemble other New Yorkers their age than they do their immigrant parents.

The two groups whose outcomes are most surprising are Puerto Ricans and Chinese. The Puerto Ricans are doing worse in terms of their educations and occupations than might be expected from their parents' socioeconomic status and the Chinese are doing much better. Types of acculturation explain neither of these outliers. Both groups have high rates of dissonant acculturation, yet their socioeconomic attainment could not be more different.

Indeed, other aspects of segmented assimilation theory better explain the disparate mobility trajectories of the Puerto Ricans and the Chinese. Clearly, Puerto Ricans suffer from a negative context of reception – including racial discrimination in housing and the labor force, substandard inner city schools, and circular migration to the island facilitated by their American citizenship. The Chinese second generation benefits from greater class heterogeneity within the immigrant community, together with a strong sense of group boundaries and established ethnic institutions and media. While many members of the Chinese second generation have poorly educated parents, some have college degrees and professional jobs. Information is exchanged within the Chinese community among parents of different social class backgrounds and members of the second generation can benefit from advantageous social ties. This suggests that it is not the overall level of ties to the ethnic group or selective acculturation at the individual level that leads to better outcomes. Rather it is maintaining ethnic ties within those groups which have significant numbers of middle class, educated members that help children of poor immigrants. Ethnic embeddedness and social capital are helpful when they connect people to those with significant resources. They are of far less use for groups that are more uniformly poor.

Selective acculturation is an attractive concept. It recognizes the fear of many immigrant parents that their children are Americanizing too quickly. It also suggests an easily implemented policy solution. Instead of focusing on improving inner city schools or ending racial discrimination, the theory suggests that the lives of inner-city second-generation youths can be improved by strengthening the bonds of social capital within their ethnic communities, encouraging bilingual education, and strengthening family ties. While these goals may be worthwhile in their own right, our analyses suggest that they will do little to promote positive outcomes in the second generation.

Acknowledgments

We are grateful to Richard Alba, Maurice Crul, Nicole Deterding, Eva Rosen, Jens Schneider, three anonymous reviewers and participants of the Harvard Migration Workshop for their helpful feedback on previous drafts.

Notes

1For a detailed description of the study, refer to the Methodological Appendix in Kasinitz et al (2008).

2South American includes Columbian, Ecuadoran and Peruvian.

3Many respondents in our sample may take longer to finish, might be employed full-time or part-time while working towards their bachelor's degree, and might eventually get their bachelor's degree after the age of 22. However, using different cut-off ages does not affect our findings (results available upon request).

4Similarly, many individuals may still be enrolled in advanced degree program by age 25, but using different cut-off ages did not lead to substantively different results.

5One might argue that types of acculturation might have contributed to higher levels of educational attainment among the second generation and, as a result, indirectly contributed to professional attainment among the second generation. This would imply that the effect of acculturation on professional attainment is mediated by educational achievement. However, even if we remove respondent's level of education from our model (results not shown, but available upon request), type of acculturation still has no significant predictive power on professional attainment.

6Our measures of ethnic embeddedness could be crude and better measures would be desirable, but experimenting with contextual measures such as the percent of immigrant population and one's own ethnic group at the zip-code level did not yield significant results.

Contributor Information

Mary C. Waters, M.E. Zukerman Professor in the Department of Sociology at Harvard University. 33 Kirkland St., Cambridge MA 02138.

Van C. Tran, The Department of Sociology and Social Policy at Harvard University. 33 Kirkland St., Cambridge MA 02138.

Philip Kasinitz, The Department of Sociology at The Graduate Center of The City University of New York. 365 Fifth Avenue New York, NY 10016.

John H. Mollenkopf, The Department of Political Science and Sociology at The Graduate Center of the City University of New York. 365 Fifth Avenue, New York, NY 10016.

References

• Abada Teresa, Hou Feng, Ram Bali. Ethnic differences in educational attainment among the children of Canadian immigrants. Canadian Journal of Sociology. 2009;34(1):1–28. [Google Scholar]
• Alba Richard, Nee Victor. Remaking the American Mainstream: Assimilation and Contemporary Immigration. Cambridge, MA: Harvard University Press; 2003. [Google Scholar]
• Boyd Monica. Educational attainments of immigrant offspring: success or segmented assimilation? International Migration Review. 2002;36(4):1037–60. [Google Scholar]
• Crul Maurice, Thomson Mark. The second generation in Europe and the United States: how is the transatlantic debate relevant for further research on the European second generation? Journal of Ethnic and Migration Studies. 2007;33(7):1025–1041. [Google Scholar]
• Farley Reynolds, Alba Richard. The new second generation in the United States. International Migration Review. 2002;36(3):669–701. [Google Scholar]
• Gans Herbert J. Acculturation, assimilation and mobility. Ethnic and Racial Studies. 2007;30(1):152–64. [Google Scholar]
• Gratton Brian. Race, the children of immigrants, and social science theory. Journal of American Ethnic History. 2002;21(4):74–84. [Google Scholar]
• Greenman Emily, Xie Yu. Social Science Research. 1. Vol. 37. 2008. Is assimilation theory dead? the effect of assimilation on adolescent well-being; pp. 109–37. [PMC free article] [PubMed] [Google Scholar]
• Hirschman Charles. The educational enrollment of immigrant youth: a test of the segmented-assimilation hypothesis. Demography. 2001;38(3):317–36. [PubMed] [Google Scholar]
• Kasinitz Philip, Mollenkopf John H, Waters Mary C, Holdaway Jennifer. Inheriting the City: The Children of Immigrants Come of Age. New York: Russell Sage Foundation; 2008. [Google Scholar]
• Kasinitz Philip, Mollenkopf John, Waters Mary C. Becoming New Yorkers: Ethnographies of the New Second Generation. New York: Russell Sage Foundation; 2004. [Google Scholar]
• Kasinitz Philip, Mollenkopf John, Waters Mary C. Becoming American/becoming New Yorkers: immigrant incorporation in a majority minority city. International Migration Review. 2002;36(4):1020–36. [Google Scholar]
• Menjívar Cecilia. Fragmented Ties: Salvadoran Immigrant Networks in America. Berkeley, CA: University of California Press; 2000. [Google Scholar]
• Park Robert E, Burgess Ernest. The City. Chicago, IL: The University of Chicago Press; 1925. [Google Scholar]
• Perlman Joel, Waldinger Roger. Second generation decline? children of immigrants, past and present – a reconsideration. International Migration Review. 1997;31(4):893–922. [PubMed] [Google Scholar]
• Portes Alejandro, Fernandez-Kelly Patricia, Haller William. Segmented assimilation on the ground: the new second generation in early adulthood. Ethnic and Racial Studies. 2005;28(6):1000–40. [Google Scholar]
• Portes Alejandro, Rumbaut Ruben G. Legacies: The Story of the Immigrant Second Generation. Berkeley, CA: University of California Press; 2001. [Google Scholar]
• Portes Alejandro, Min Zhou. The New Second Generation: Segmented Assimilation and its Variants. The Annals. 1993;530(1):74–96. [Google Scholar]
• Smith James. Assimilation across the Latino generations. American Economic Review. 2003;93(2):315–19. [Google Scholar]
• Stepick Alex. Pride against prejudice: Haitians in the U S. Boston, MA: Allyn and Bacon Publishers; 1998. [Google Scholar]
• Vickerman Milton. Crosscurrents: West Indian Immigrants and Race. New York: Oxford University Press; 1999. [Google Scholar]
• Waldinger Roger. Did manufacturing matter? The experience of yesterday's second generation: a reassessment. International Migration Review. 2007;41(1):3–39. [Google Scholar]
• Waldinger Roger, Feliciano Cynthia. Will the new second generation experience ‘downward assimilation’? Segmented assimilation re-assessed. Ethnic and Racial Studies. 2004;27(3):376–402. [Google Scholar]
• Waldinger Roger, Perlman Joel. Second generations: past, present, future. Journal of Ethnic and Migration Studies. 1998;24(1):5–24. [Google Scholar]
• Warner W Lloyd, Srole Leo. The Social Systems of American Ethnic Groups. New Haven: Yale University Press; 1945. [Google Scholar]
• Waters Mary C. Black Identities: West Indian Immigrant Dreams and American Realities. New York: Russell Sage Foundation; 1999. [Google Scholar]
• Zéphir Flore. Trends in Ethnic Identification among Second-Generation Haitian Immigrants in New York City. Westport, CT: Bergin and Garvey; 2001. [Google Scholar]
• Zhou Min. Segmented assimilation: issues, controversies and recent research on the new second generation. International Migration Review. 1997;31(4):975–1008. [PubMed] [Google Scholar]
• Zhou Min, Bankston Carl L. Growing Up American: How Vietnamese Children Adapt to Life in the United States. New York: Russell Sage Foundation; 1998. [Google Scholar]
• Zhou Min, Xiong Yang Sao. The multifaceted American experiences of the children of Asian immigrants: lessons from segmented assimilation. Ethnic and Racial Studies. 2005;28(6):1119–52. [Google Scholar]
• Zhou Min, Kim Susan S. Community forces, social capital, and educational achievement: the case of supplementary education in the Chinese and Korean immigrant communities. Harvard Educational Review. 2006;76(1):1–29. [Google Scholar]