Close
David Coleman, Sylvie Dubuc and John Haskey, University of Oxford
(Project no. 30060)
The first aim is to update and to improve the estimates of vital rates according to ethnic origin and birthplace provided by the ONS LS in study 30001 'Post War Migration and the United Kingdom' by extending the coverage to 2001. The previous data were not linked to the 2001 Census. Updated data will enhance sample size, be more timely and more useful for the study of ethnic differential and trends in vital rates and in the production of population projections. The second is to use the question on religion in the 2001 Census to help to determine levels and trends of fertility by religious affiliation, at least for the 1990s up to 2001.
The level and trend of vital rates according to ethnic origin and birthplace is well known to be important directly on grounds of health and welfare and indirectly because of the implications of differentials for the future growth and distribution of ethnic minority populations at national and local level. No direct measure of the ethnic origin of births is available except for incomplete hospital records, and there is none at all relating to mortality. Instead estimates of fertility for population projections must be made indirectly from various methods such as the 'own child' method or the alternatives adopted by Large (2006) and Rees (in press). These do not yield results that are closely comparable, either to each other or to the more recent years (post 1991) of the ONS LS data that were obtained previously. Preliminary population projections, not using ONS LS-based data, have already been prepared and published in elementary form. So far attempts to create life-tables for ethnic minority populations have stalled for want of sufficient numbers of linked deaths at younger ages.
The introduction of a question on religious affiliation in the 2001 Census opens up a completely new dimension of analysis of demographic characteristics. There has been much speculation about the relationship of vital rates, especially fertility and family building, to religious affiliation, especially in respect of non-Christian religions. At least for the period 1991-2001, new information should be available to test expectations about the different levels of 'modernisation' of fertility behaviour in respect of the timing and the parity progression of births, and to allow preliminary projections to be made on the future population according to ethnic origin, on lines already pioneered by (e.g.) Skirbekk, 2005.
The ONS LS is the only source of data whereby vital events to ethnic minority populations can be estimated directly, albeit through linkage, and on an individual basis and also linked with the rich variety of variables from the census as well as basic data such a birthplace, age and marital status. It is the only possible source, in particular, of data on parity progression. Most particularly, no other routine source of data with a large sample size offers any information on religious affiliation able to be linked to demographic characteristics. In that its contribution is unique.