Ethnicity's Shadows: Race, Religion, and Nationality as Alternative Identities among Recent United States Arrivals

Abstract

In recent decades, ethnicity has received increasing academic attention—and popular acceptance—as a core aspect of identity and one that is particularly relevant to immigrants. But is this notion of ethnicity salient to the immigrants themselves? This article provides a set of quantitative snapshots of the way new arrivals to the United States identify themselves and their children. The data—a set of annual surveys of recent refugee arrivals—suggest that in response to specific questions about “ethnicity,” newcomers often provide what from American perspectives are not “ethnic” responses. In particular, race and religion are frequently offered as ethnic identity. Furthermore, nationality can also override ethnicity. Parents, for example, sometimes categorize recent United States-born children as ethnically “American” even though the rest of the family retains its previous non-United States identity. Such identity choices reveal much about the starting point of migrant adjustment. They are also invaluable for the broader analysis of the logic, history, contingencies, and internal tensions of United States—and Euro-American—notions of ethnicity.

Key Words:

• ethnicity
• identity
• nationality
• immigration
• refugees

I thank the editors and reviewers of Identities for very helpful comments on the original submission, and Loren Bussert, Shawn Davis, and Leonid Guzman for assistance on the data and procedures for the Annual Survey of Refugees as used in this article.

Notes

1. See Jenkins (1997: 44–50) for sequential discussion of these two separate polarities of culture and biology, and the primordial versus the instrumental (distinctions that he sees as homologous to some degree), and for the specific point about race as imposed and ethnicity as chosen.

2. Jenkins (1997: 49) also supports the notion that “the anthropological celebration of ethnicity is a ‘good thing.’” Hirschman (2004), as another example, makes the same argument from the inverse perspective: that even at its very worst as ethnocentrism, it is still positive in comparison to racism. Nevertheless, the critique that ethnicity (and ethnocentrism) may effectively recreate race (and racism) also has its supporters—for example, Grosfoguel (2004).

3. On the one hand, then, ethnicity seems humane, flexible, and certainly not racist, even though the enduringness of ethnicity almost always hinges on both cultural and biological transmission. Culture, after all, travels especially well along bloodlines. On the other hand, ethnicity conveys a quite durable notion of a “peopleness,” even though we also know how often ethnicity is created, shaped, reshaped, invented, denied, imposed, rejected, negotiated, argued, and sometimes just ignored.

4. Omi and Winant (1986) stress the importance of an immigrant analogy in the very development of the concept of ethnicity, for it was immigrants who initially raised the questions of the possibility of change and thus the need for a kind of identity that was mutable.

5. Ethnicity often carries the sense of “peopleness” that used to be adduced by the word “nation” before that was wedded to the state. As an example, refugee law admits of no persecution for “ethnicity” per se, but only for race, religion, nationality, political opinion, or membership in a particular social group.

6. Karakayali (2005) provides a useful critique of the long-standing tendency to frame problems of the second generation according to a notion of children living in “two worlds.” Her own work on early immigrant autobiographies suggests that this notion—although not corresponding particularly to the experience of children—nevertheless was absorbed into their writing due to its pervasiveness as a concept in the society at large.

7. Consider, for example, the phrasing of a recent review of ethnic relations that begins with “Throughout its history, the United States has been inhabited by a variety of interacting racial or ethnic groups” (Frederickson 1999: 23). The “or” at the end of the sentence quite diffuses the difference between what might lead to a diatribe on slavery and what might lead to a paean to cultural diversity.

8. The literature is also clear that ethnic identity can have a positive transforming effect as well as a more negative and limiting one.

9. For recent examples of these four, see Grosfoguel 2004; Baker-Cristales 2004; Peek 2005; and Hussain and Bagguley 2005.

10. One of the interesting findings from both these studies is the tendency of people on the outside of the ethnic category to respond as if issues are universal to all people, while those within the ethnic category construe them as representing specific ethnic cultural tendencies.

11. This discussion is drawn from Benson 2005; Sanchez and Thorp 2005; Gentemann and Zhou 2005; Gunawardena and Findlay 2005; and Haines and Rosenblum 2005. The more general impact of immigrants on higher education has also received broad attention (Musil et al. 1995; Smith 1997; Gray, Vernez, and Rolph 1996; Grubb, Badway, and Bell 2003; Harklau 1999), including the effect of immigrants on other minorities (Hoxby 1998).

12. A more general example of this divide is provided by Tsai et al. (2000). It is worth noting that the General Ethnicity Questionnaire, which she uses for this particular piece of research, is framed toward ethnicity as culture and language. That is, “Chinese” is effectively stripped of its specifically national implications. It is thus not an issue of national origin in the strict sense. Nevertheless, as with an emphasis on national origin, ethnicity is here constructed as a single bounded dimension of identity, rather than a kind of holding place for multiple dimensions of identity.

13. See, for example, Schwartz and Montgomery (2002) and Tanaka (2003). This is an issue not only of identity but also of interaction (Chen et al. 2001) and of forging new paths to college for those who have not previously attended for various reasons, including gender (Lee 1997; Yiv and Secombe 1999; Simmons and Plaza 1998; Bankston 1995; Zhou and Bankston 2001).

14. Patel's specific reference is to the Indian census, but it applies rather well to the United States one as well.

15. These annual surveys had their origin in 1975 with what had become by 1979 a set of nine surveys of Southeast Asian refugees conducted and subsequently published by Opportunity Systems Incorporated. In 1981, the survey was substantially redesigned and in 1984 transformed into a panel design (Gordon 1989). These surveys were influential in early discussions of the economic adjustment of Vietnamese refugees (Stein 1979). The redesigned surveys were also used extensively for discussions of general adaptation (Haines 1983; Bach et al. 1983; Bach et al. 1984) and for more specific issues of household composition (Bach 1985), continuities in the labor force activities of women (Haines 1987), the effects of type of sponsorship (Bach and Carroll-Seguin 1986), and the degree of continuity in household formation between Vietnam and the United States (Haines 2002). Over time, and with changes in the United States refugee program, these surveys have expanded to include a very wide range of countries of origin. They thus provide data not only on the Southeast Asian refugees for whom the surveys were first designed (still crucial in the 1995 survey to be discussed in the text) but also the surge in arrivals from Eastern Europe and the former Soviet Union (well represented in both the 1995 and 2001 surveys) and also the greatly increased number of Africans (who are particularly apparent in the 2001 survey—and some of the more illustrative data used from the 2002 survey).

16. More descriptive reports of the survey's methodology and findings are provided in the annual reports of the Office of Refugee Resettlement. The data presented here are calculated separately from the public use data files for the 1993–2002 surveys.

17. There are two kinds of translation involved: formal written translation of the instrument into a foreign language for the largest of the refugee language groups (e.g., Chinese, Lao, Hmong, Khmer, and Vietnamese for the 1995 survey) and ad hoc translations by staff when that is necessary during actual survey calls for less common languages.

18. The surveys are conducted by phone based on lists of refugees resettled. Those lists include initial addresses. Although there are some potential problems with phone interviews, they do have some advantages in general (ease of having multiple interpreters available) and possibly some advantages for this particular topic by taking statements about ethnicity out of direct visual encounter. The answers that will appear about race, for example, might be handled differently by phone and in person.

19. This is partially an artifact of the situation at the time these surveys were redesigned in 1980. As part of the team that did not redesign, I would note the strong commitment at that time to move away from nationality as a classifier and toward a set of Asian ethnic categories not terribly different in concept from those used in the contemporary United States. The two driving forces in that process were the presence of large number of Chinese among those from Vietnam, and the arrival of numerous highland groups from Laos (especially Hmong). The mixing of these populations with ethnic Vietnamese on the one hand and lowland Loa on the other had truly bizarre effects in statistical reporting. Any statement about how “Laotians” were doing simply ignored the very disparate situations of ethnic Lao and Hmong. The differences were less sharp for ethnic Vietnamese versus Chinese from Vietnam, but still significant.

20. Concepts of ethnicity and actual linguistic codes for ethnicity are related. Whether interviewees are faced with an English word whose parameters are unclear to them or whether a written or verbal translation actually shifts the meaning of “ethnicity” is not always clear. Whether it is the translator or the interviewee who makes the “translation,” the same problem (and potential) remains in the research to show a divergent notion of the jointly conceptual and linguistic term “ethnicity.” This is not always, however, a serious problem. As an example, in the Vietnamese case, “ethnicity” is usually translated formally as “dan toc”—literally “people-clan.” That is a reasonably good gloss for ethnicity, though it actually rings more with the spirit of the late nineteenth and early twentieth century European notion of “nationality” (in its original sense of “peopleness”) rather than with the current link of that term to the state apparatus of governance.

21. Even with continuing groups, there are shifts in numbers (e.g., increasing numbers of Bosnians but shrinking numbers of those from the former Soviet Union).

22. These ethnic labels are not always fully self-explanatory. One might wonder, for example, how many Eritreans were among the 55 Ethiopians, or how many Chinese-Vietnamese might have (for a variety of reasons) identified themselves as Vietnamese. That is, the ethnicity options not taken are unavailable in these surveys. Here, decennial census data are more reliable since minority status (e.g., Chinese-Vietnamese) can be triangulated among country of origin, self-described “race,” ancestry, and home language.

23. See Gilman (1991) for the specific issue of Jews, but see also Roediger (1991) and Jacobson (1998) for the more general issue of being white in America and Takaki (1989) for a good review of the stages by which white and Asian were ever more carefully distinguished in American federal and state law. Despite the general intransigence of race (as compared to ethnicity), it is worth reiterating that race can change: Jews became white (Sacks 1994) and Latinos became people of color (Melville 1988: 74).

24. In interim years, the number of responses for the hybrid category was, in fact, quite large. In 1996, for example, it was actually the most frequently used ethnic identity option for those from Bosnia.

25. Technically, the terms “ethnic” and “tribal” should be distinguished, with the latter having a quite specific meaning of linked by kinship. In common discussion, however, very strong ethnic consciousness in a group might well be noted as “tribal” and certainly a tribal identity would meet any of the (less restrictive) criteria for shared ethnicity. Both terms partake of the mix of culture and biology.

26. Qian (2004) provides a very interesting analysis of a similar issue: how the race of biracial children is assessed by their parents in response to census questions.

27. This decline is gradual over time, not sudden. The 2001 survey was in fact conducted during Fall 2001, and that certainly might have affected how some people responded. But the pattern itself was already established.

28. The exception involves their other young United States-born, United States-citizen siblings.

29. The numbers were also higher before 1995. The survey in 1995 thus appears anomalous.

30. There were also 13 people identified ethnically as Burmese. They were from Bosnia both by birth and citizenship.