Close
Mathieu Ichou, INED and Anthony Heath and Jan O. Jonsson, University of Oxford
(Project no. 0301766, previously 30176)
This project aims at studying the structural assimilation of immigrants and their children in England and Wales, especially on the labour market, using individual-level cross-sectional and longitudinal statistical techniques.
Britain has long been a country of immigration. According to the ONS, 1 in 8 UK residents was born abroad and 1 out of 4 babies born in the UK is a child of immigrants. Rigorous empirical results on the assimilation of immigrants and their children would therefore be of both scientific value and have public policy implications.
As part of my postdoctoral research fellowship at Nuffield College (University of Oxford), I intend to use the ONS Longitudinal Study to analyse the structural assimilation of immigrants in England and Wales. Using data from the 1971, 1981, 1991, 2001, and 2011 censuses, I will analyse the socio-economic trajectories of immigrants and their children, and compare them to those of natives.
I intend to make four main contributions. The first three are substantive, the fourth is primarily methodological. First, I would like to assess the impact of immigrants' initial characteristics on their later socio-economic trajectories in England and Wales. I would especially like to investigate the effects of immigrant educational selectivity (i.e. their relative educational attainment compared to non-immigrants from their country of origin), for which I created an original measure in a previous research project. The construction of this variable for immigrants in the ONS LS would require matching data from an external source of data (the Barro-Lee dataset on educational attainment distribution around the world) into the LS. The matching will be conducted at the individual level. The matching variables (existing or creatable in both the LS and Barro-Lee datasets) would be year of birth (in 5-year bands), age at migration (in 5-year bands), sex and country of birth. To my knowledge, the empirical measure of immigrant educational selectivity at the individual level and its impact on assimilation has never been achieved and would therefore constitute an important step in understanding the influence of immigrants' pre-migration characteristics on their post-migration outcomes.
Second, I would like to examine trends in the level of socio-economic diversity among the immigrant-origin population. Research has traditionally been interested in describing and explaining average characteristics of groups (through measures of central tendency), but attention is seldom devoted to describing and explaining dispersion within groups. This is what I intend to do by using measures of dispersion as dependent variables and by adopting both a repeated-cross section and a longitudinal perspectives.
Third, I intend to study the intergenerational social mobility between immigrants and their children. By linking young LS members to their cohabitating parents, I will derive parents' occupation at the beginning of the period (1971 and 1981). I will then take a measure of occupation of the children 30 years later (in 2001 and 2011). My goal is to compare the statistical relationship between parents' and children's occupation within immigrant and native families. I would therefore be able to show whether or not the process of intergenerational mobility is similar or different between the majority and minority groups.
Fourth, from a methodological point of view, I plan to compare the findings regarding immigrant assimilation obtained from a true longitudinal approach with traditional findings obtained from single cross-sections or repeated cross-sections. If, as I expect, the results differ significantly, I will examine the source of these differences in potential biases related to differential selectivity levels between arrival cohorts and differential return migration.
In addition to descriptive statistics, the methods used will include classification techniques (e.g. clustering and latent class analysis) to produce typologies of socio-economic trajectories, basic multiple regression and panel models (for both continuous and categorical dependent variables), and age-period-cohort models.