ABSTRACT
Cross-ethnic friendships are at the heart of the immigration experience. We examine the incidence and change of co-ethnic friendship groups among 1.5 and second-generation immigrant youth in Canada using the nationally representative Ethnic Diversity Survey. We find that visible minority immigrant youth friendships are more co-ethic than those of non-visible minority immigrant youth. We also find visible minority individuals’ friendship groups become more co-ethnic over time. In addition, we find differences in the pattern and direction of change across different visible minority immigrant youth groups. We interpret these findings in light of theories of assimilation and segmented assimilation.
Disclosure statement
No potential conflict of interest was reported by the author(s).
Correction Statement
This article has been republished with minor changes. These changes do not impact the academic content of the article.
Notes
1 See http://www23.statcan.gc.ca/imdb/p2SV.pl?Function=getSurvey&SDDS=4508 for details.
2 The sample includes all residents of Canada including citizens, landed-immigrants, and non-permanent residents. Those under 15 years of age, residents of collective housing, living on Indian reserves or of Aboriginal background, and those living in remote territories are not part of the sampling frame. Population weights are used for all descriptive statistics reported in this paper. Boostrap weighting techniques (using the STATA pweights function with bootstrap options) are used for each model reported in order to account for sampling design and generate unbiased variance estimates.
3 The EDS allows up to eight multiple ethnic ancestry responses, and asks the network question in reference to the two ancestries ranked most important by the respondent. We explored measures of co-ethnic friendships that prioritized first, second or combined ancestries. While acknowledging the empirical reality of multiple ethnicities, we do not present the findings here.
4 This recollection and change variable is essential to the analysis below, but we are aware of the limitations of retrospective questions. Bauman and Ennett (Citation1994) observe that respondents are biased towards perceiving similarity among friends. We do not, however, see any reason to expect this bias to vary across our key comparisons in the analysis to come.
5 The EDS separates East Indian’s from Punjabis, Pakistanis, and Sri Lankans which are commonly considered South Asian (Statistics Canada Citation2003). We follow the common practice of considering all these groups South Asian.
6 We use the age category distinctions made available to us through the EDS.
7 We use the binary outcome variable indicating a co-ethnic friendship group because ordinal regression models did not meet the parallel regression assumption and the assumption of equal distances between categories is not obvious. Results from the ordinal regression models do not contradict the findings presented. Models using different cut off points for measuring a diverse network also do not contradict the findings presented.