Investigation of first onward moves of international migrants

Emma Wright, Giles Horsfield, Alison Whitworth and Nargis Rahman, Office for National Statistics

[Project number 20085]

International migration is a key component of population change - it currently accounts for about 80% of UK annual population change, and also accounts for a large proportion of population change within specific local areas, especially London. However, it is also the hardest component of population change to measure, and as a result, a National Statistics Quality Review (NSQR) of International Migration Statistics was published in 2003 (see www.statistics.gov.uk/about/data/methodology/quality/reviews/population.asp), the recommendations of which the ONS Population & Demography Division (P&D) are currently taking forward.

One of the recommendations of the NSQR was to investigate ways in which the geographical distribution of UK international migration inflows could be improved, both at the UK country/GOR level and at lower geographical levels such as local authority. The current basis of our estimates of International Migration is the International Passenger Survey (IPS), but research comparing the geographical distribution of migrants in the IPS with other sources such as the 2001 Census and the Labour Force Survey (LFS) indicate that the IPS may overestimate the number of international migrants settling in London and underestimate the numbers settling elsewhere in the UK. As a result, the Migration Statistics Unit (MSU) of P&D and the Demographic Methods Centre (DMC) have been researching the potential for combining IPS and LFS data to improve the geographical distribution of international in-migrants.

Estimation of internal migration within the UK is based on the NHS Central Register and patient register data held by individual health authorities. As part of the work on improving the geographical distribution of international in-migrants, it is essential to consider the interaction of estimates of international migration with estimates of internal migration, and in particular, the first onward internal moves of international migrants within the UK. For example, if an international migrant arriving in the UK informs the IPS that they intend to live in area A, but then in reality quickly move on to area B, this is not a problem in terms of population estimation if they register with an NHS GP in area A, as their subsequent re-registration in area B would be included within internal migration estimates. However, if the same migrant did not register with a GP in area A, and their first registration with the NHS was in area B, in terms of population estimation the population of area A would be overestimated and the population of area B would be underestimated.

The LS permits analysis of the timing (month and year) of immigration, and the usual place of residence of the sample of international migrants based on immigration records from the NHSCR during the 1990s. The sample will include those who subsequently embarked and/or re-entered and will be analysed by age, sex and, where the information is available, country of birth recorded at the 2001 Census. Linked 2001 Census information will also provide information on internal migration over several analysis periods:

  • between the point of immigration and the year before Census (comparing immigration records and 2001 Census 1-year migration indicator)
  • between the point of immigration and Census (comparing immigration records and usual residence on Census night 2001)
  • in the year before Census (using 1 year migration indicator at 2001 Census)
  • between the point of immigration, the year before Census and Census night (comparing all three records of usual residence)

Analysis of usual place of residence will be undertaken at the Government Office Region and local authority levels.

For LS members who appeared in the Study for the first time in 2001 and for whom we have no immigrant record, internal migration in the year before Census will be identified.

The analysis will be repeated for immigrants during the 1980s (as in stage 1 above) and LS members found for the first time at the 1991 Census (stage 2). Analysis of those found for the first time at the 1991 Census will be supplemented with information on place of usual residence from immigration records for those whose entry to England and Wales was subsequently recorded at NHSCR, after GP registration.

Top of page