Ethnic identity changers 1991-2001

Robert Moore, University of Liverpool, Mary Hickman, London Metropolitan University

[Project number 30095]

Ethnic identities are more fluid than the framers of the 1991 'ethnic' question realised. Self-defined ethnicity may change with the contingencies of the life course. About 10 per cent of people changed ethnic category between censuses, 20 per cent of Black Caribbean and slightly less for other Black categories. In round figures there will be about 4,000 LS members in England who have changed their ethnic identity since 1991.

The formal entitlement to Irish nationality derives from birth in Ireland, or a parent or grandparent born in Ireland. The 2001 census offered 'Irish' as an ethnic identity for the first time. Field work by Hickman at the time of the census indicated widespread misunderstanding of Question 8 in the enumeration form. Survey results show that the 2001 census may undercount people who would normally claim to be 'Irish' when questioned in surveys by up to a factor of ten. There are two potential sources of change of identity for non-white groups; one will be random respondent error; other researchers have work in hand to estimate the extent of this. Secondly changes in the social circumstance of persons filling in the census, on their own behalf and for dependent children - changed domestic arrangements, social mobility etc. This proposal is for an study of ethnic identity changers. The social circumstances of the 'changers' will be examined at each census to discover the extent to which changes in social and domestic circumstances are associated with ethnic change. We can not address the political issues that might lead people to change their ethnic identity through the LS.

For the Irish we will compare the social characteristics (age, education, class, social mobility, UK region, etc) of people entitled to Irish citizenship who do and do not claim Irish ethnicity in the 2001 census.

The LS is the only source of data that will enable us to realise our aims given the size of the longitudinal sample. The British Household Panel Survey is the only survey with longitudinal data on ethnic identity and it will not offer a sufficiently large set of identity changers.