![Sample our Social Sciences journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/110822/TOC-Subject-Sample-2021-Social-Sciences.jpg"})
ABSTRACT
Socio-cultural and ethnic origin can be a powerful predictor of social attitudes and behaviours but, unlike the situation in the classical countries of immigration such as Australia, Canada and the USA, there is no standard measure in Europe for measuring ethnic background. The paper reports a new measure and classification, developed for the ESS and trialled in the ESS wave 7 (2014/2015). It describes the underlying theoretical concepts, structure and classification criteria and reports a range of substantive findings. The paper shows that the new measure of ethnic origins has both criterion and predictive validity: it predicts whether respondents identify themselves as belonging to an ethnic minority and whether they feel that theirs is a group which is discriminated against. It also predicts strength of national identity and attitudes towards immigration. A particular strength of the new measure is that it identifies both indigenous and (sub)national minorities as well those with a migration background. The paper shows that in some countries subnational minorities are quite distinctive, for example in their feelings of being discriminated against and in their low levels of national attachment.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 The ABS stated that ‘For the purposes of ASCCEG, “ethnicity” refers to the shared identity or similarity of a group of people on the basis of one or more factors. These factors were enunciated by the 1986 Population Census Ethnicity Committee (ABS Cat. No. 2172.0), chaired by the late Professor W.D. Borrie CBE, in The Measurement of Ethnicity in the Australian Census of Population and Housing report to the Australian Statistician (the Borrie Report). ‘The Committee considered that the most enlightening attempt to define an ethnic group is that contained in a United Kingdom Law Lords statement’ (reported in Patterns of Prejudice, Vol 17, No. 2, 1983). The Law Lords noted the key factor as being the group regarding itself and is regarded by others, as a distinct community by virtue of certain characteristics, not all of which have to be present in the case of each ethnic group’ (Australian Bureau of Statistics Citation2011). Incidentally the Law Lords’ judgement established Sikhs as a recognized ethnic group for the purposes of protection from discrimination under English law.
2 But see the groupings suggested by the UN at http://unstats.un.org/unsd/methods/m49/m49regin.htm. There are also several solutions for aggregating countries to the rather simplistic ‘global south’ (i.e. developing or low-income countries) and ‘global north’ (i.e. developed or high-income countries) (International Organization for Migration Citation2013).
3 Overseas territories of some nation states (e.g. Bermudan, Aruban) also receive their own category, because they are specific in that their populations are often ethnically and culturally highly mixed and contain substantial numbers of indigenous peoples (and, for example in the case of the Caribbean, descendants of slaves) with quite distinct cultural and ethnic roots from the ‘national’ majority that the territory belongs to.
4 For further detail on this variable, see Heath, Schneider, and Butt (Citation2016). For purposes of anonymization, Norway simplified the ancestry variables and is thus excluded from all analyses.
5 In a relatively small number of cases, respondents gave origins from two different broad groups. If one of these was European, we coded the respondent to the European broad category. In other cases, we made the arbitrary decision to code according to the first ancestry named.
6 Sub-national means 5th digit categories not referring to the main national group, as well as 4th digit categories that are autochthonous to the country in question but cross border into another country (e.g. Frisian for Germany and the Netherlands), or are classified in a different narrow or broad group altogether (such as Central Asian minority groups in Russia).
7 For simplicity, in the small number of cases where there were missing values we used mean substitution. Separate analyses of the three component variables showed broadly similar patterns to the overall measure.
8 The comparison of coefficients from logistic regressions across samples has been criticized (see e.g. Mood Citation2010; Karlson, Holm, and Breen Citation2012). However, these critiques rely on the assumption that the logistic regression is modelling a binary version of an unobserved latent continuous variable (Kuha and Mills Citation2018). We do not believe that this assumption is valid in the cases which we deal with and we therefore think that comparing the magnitude and significance of the coefficients across the three sets of countries is a legitimate undertaking. We report results using average marginal effects (AMEs) anyway in the online appendix for the interested reader. In almost all cases, AMEs follow the pattern of the logit effects.
9 Also, some of their effects are picked up by the measures of discrimination and belonging to an ethnic minority group now included in these models. The remaining ancestry effects apply only to those respondents who don’t feel discriminated and don’t feel they belong to an ethnic minority group. We did not test for interaction effects in these models.
10 The notions of ‘allochthonous’ and ‘autochthonous’ are of course rather problematic in the case of Israel. In our measure, respondents in Israel with only Jewish/Israeli origins are defined as autochthonous only without subnational origin, even though they may be immigrants or 2nd generation in many cases. This latter information is, in our view, better captured in the variable on generational status.