You are in: Home › Statistics › Publications and Data › Population Estimates › Estimating council area migration ›

Estimating council area migration.

Estimating migration

Migration is the most difficult component of population change to estimate. The other components (births and deaths) are estimated using data from the civil registration system, which is considered to be virtually complete and therefore easy to estimate. In contrast, there is no comprehensive system, which registers migration in the UK, either moves to or from the rest of the world, or moves within the UK. Estimates of migration therefore have to be based on survey data and the best proxy data that exist.

Sources of data for estimating migration

Migration is derived from two key sources of data: The International Passenger Survey (IPS) for moves into and out of Scotland from outside the UK, and the National Health Service Central Register (NHSCR) for moves between Health Board areas within the UK. In previous years, electoral and school roll data were used to derive migration at council area level, but in 2002, these two sources were replaced by anonymised data from the Community Health Index (CHI).

Estimating council area migration

The CHI holds records of people registered with an NHS doctor in Scotland. Unlike the NHSCR, the records provided to GROS contain the postcode of the patient's address, which enables migration to be estimated for councils, and potentially for smaller areas. The approach used for estimating council migration involves matching CHI patient records from extracts taken a year apart.

This matching of two extracts, say A and B, will create three sets of patients:

A set of patients in extract A and extract B
A set of patients in extract A but not in extract B
A set of patients in extract B but not in extract A.

The set of patients in both extracts whose postcodes were not the same in each extract can be considered as within Scotland migrants, i.e. moved from one address in Scotland to another within Scotland between the dates of the two extracts. The remaining two sets of patients are either migrants to or from Scotland; babies born between the two extracts; deaths between the dates of the two extracts; movements to/from the Armed Forces; and a small number of records that are for the same patient but have different CHI numbers because they could not be matched when the GP registration was processed.

It is recognised that the GROS migration data derived from the NHSCR is the most reliable data available at health board level, so estimates from the CHI are controlled to ensure that they are consistent with the NHSCR data for moves across a health board boundary. CHI migration records for moves between health boards, moves to and from Scotland, and moves to and from the Armed Forces are controlled to agree with the NHSCR-derived health board migration data by randomly duplicating or deleting CHI records until the desired total is achieved. The controlling process produces a CHI dataset that is fully consistent with NHSCR-derived migration at HB level, by origin, destination, age and sex.

Further Details

Papers giving further details on using the CHI to estimate migration for council areas are available by clicking on the links below:- (Adobe Acrobat Portable Document Format)

PAMS (03) 10 - Improving the GROS Migration Data for Council Areas - Proposals for 2002
PAMS (03) 10(supplement) - Supporting Information -Validation of Proposed Methodology