Close
Christel Kesler, Colby College, Maine and Stephen Jivraj, UCL
(Project no. 0201465, previously 20146)
Our project concerns ethnic identification across successive generations of immigrants' descendants. We wish to explore potential determinants (particularly socioeconomic status and ethnic intermarriage) of ethnic identification among those with an immigrant background, and thereby to examine links among socioeconomic, familial, and identificational dimensions of 'assimilation'. We will also, using 2011 data, explore the links between ethnic and national identity, and the role of language use and formal nationality in shaping both ethnic and national identity among immigrants' descendants. We will use basic descriptive statistics as well as simple multivariate models.
The goal of the study is to examine ethnic identification and its correlates among the children and grandchildren of immigrants. We are specifically interested in the links (as have often been theorised in the classic social science literature on immigration but far less often empirically studied) among socioeconomic outcomes, patterns of partnership formation with the ethnic majority, and ethnic identificational 'assimilation' (i.e., identification as white British). In the proposed study, we extend on earlier work using the LS (Project Number 30090), in which we studied the ethnic identification of the immigrant second generation (children of immigrants) in adulthood, and particularly how their identification is shaped by socioeconomic factors. This prior work resulted in a forthcoming article in the International Migration Review. In the new study, we hope, first, to use additional information from the 2011 Census to better understand the correlates of ethnic identification among the second generation. The 2011 Census contains several new variables that are of interest. In particular, we hope to consider in our analysis the interaction of ethnic identification with national identity, language use, and formal nationality (as indicated by passport holding), and to incorporate also information on religion available in both 2001 and 2011. Second, we wish to extend the generational purview of our study, by analysing the ethnic identificational patterns of the third generation (grandchildren of immigrants). We expand on both of these facets of the project below.
We will rely on a sub-sample on all LS members who are present in 1971 and 15 or younger and living with one or both parents. We will primarily analyse LS members who remain in the sample in any/all of the subsequent years 1991, 2001, and 2011 (the Censuses for which ethnicity information is available), and who have at least one parent born outside the UK (i.e., the children of immigrants). However, we are interested in knowing what proportion of these young children in 1971 can be traced into adulthood in later years (and how they compare to those who cannot be traced), and also in comparing the children of immigrants to their counterparts from non-immigrant families, so we do not wish to initially limit the sample by these latter two criteria.
Our previous analysis, using the same basic sampling criteria just described to select the children of immigrants in 1971, focused primarily on ethnic identification data from 2001, when these children of immigrants were adults. The analysis indicated that higher social class in childhood, for some (specifically non-European) immigrant groups, leads to a greater propensity to identify as white British in adulthood, whereas for all groups irrespective of their parents' geographic origins, higher attained social class in adulthood (or educational attainment, which is closely correlated) corresponds to greater minority ethnic identification. With the 2011 data and the same basic sampling framework, we wish to explore the correlates of national identity among the second generation, and also the interaction between ethnic and national identity. We will also, as is relevant, explore the extent to which immigrants' children hold passports from their parents' countries of origin and how this affects their identificational patterns. Similarly, though we expect English to be the main language for the vast majority of the adult children of immigrants, in cases where it is not, we will look at the implications for ethnic and national identity. To the extent possible, we also hope in this extension of our earlier analysis of the second generation's identificational patterns to leverage the longitudinal dimension of the LS more to better pinpoint causal ordering, particularly of the education/ethnic identification association we found in our earlier study.
In a separate but related analysis, we will explore the identificational patterns of the grandchildren of immigrants, which will be the co-resident children of the second generation sub-sample we describe above, after these sub-sample members have formed their own households in later censuses (1991, 2001, and 2011). The analysis will necessarily focus on the third generation when they are still co-resident with their parents, in order that we can link information about all three generations, so we will not be able to take into account the third generation's own attained socioeconomic characteristics, nor, most likely, to examine their own self-reports of ethnic identification, since in most cases, the census information will have been completed by parents. However, we will have full information about the geographic origins and socioeconomic characteristics of grandparents on either mother's or father's side, as well as information about resident parents' socioeconomic characteristics and ethnic identification. This will permit us to consider how socioeconomic characteristics and ethnic intermarriage over the course of two prior generations affect the ethnic identification of children, a considerable advance over previous studies. The complexity of changes over time in who lives with the LS member (partnering, separation, re-partnering, children moving out of the household, etc.) will make analysis of data from all three of the most recent census years (1991, 2001, and 2011) advantageous, but we recognize that the ethnic identification variable will not be strictly comparable across censuses.
We will extract or derive the following key variables from the 1971 sub-sample of LS members we described above: LS member's gender; age; whether LS member was born abroad or in the UK; foreign-born parent's or parents' country/region of birth; whether the second parent is absent, UK-born, or foreign-born; parents' education; parents' social class; and co-resident sibship size, gender composition, and the respondent's position in it.
From the 1991, 2001, and 2011 Censuses, we will extract or derive the following key variables: LS member's ethnic identification, educational attainment, social class attainment, and marital status. For the analysis of the third generation, we will further include variables that indicate the LS member's partner's ethnic identification, educational attainment, and social class attainment and the ethnic identification, age, and gender of any co-resident children. As the variables are available, we will also analyse, as described above, information about religion, national identity, passports held, and language use for the LS member, and (if present) spouses and children.