Notes and Definitions
Notes and Definitions
The estimated population of an area includes all those usually resident there whatever their nationality. Students are treated as being resident at their term-time address. Members of HM and non-UK armed forces stationed in Scotland are included; HM forces stationed outside Scotland are excluded.
Population figures relate to 30 June of the year shown and ages relate to age last birthday.
Presentation of results
Although the populations are tabulated in units, this does not imply accuracy to that level. The data are presented in units for the convenience of users wishing to compile non-standard aggregations without encountering rounding problems.
Composition of health board areas
Health board area
|Argyll & Clyde||Argyll & Bute, East Renfrewshire (part), Inverclyde, Renfrewshire, West Dunbartonshire (part)|
|Ayrshire & Arran||East Ayrshire, North Ayrshire, South Ayrshire|
|Dumfries & Galloway||Dumfries & Galloway|
|Forth Valley||Clackmannanshire, Falkirk, Stirling|
|Grampian||Aberdeen City, Aberdeenshire, Moray|
East Dunbartonshire, East Renfrewshire (part), Glasgow City, North Lanarkshire (part),
South Lanarkshire (part), West Dunbartonshire (part)
|Lanarkshire||North Lanarkshire (part), South Lanarkshire (part)|
|Lothian||East Lothian, City of Edinburgh, Midlothian, West Lothian|
|Tayside||Angus, Dundee City, Perth & kinross|
|Western Isles||Eilean Siar|
The land areas used to calculate the population density information presented in Table 5 were derived from digital boundaries used for the 1991 Census.
The 2002 mid-year estimates use a new data source, the Community Health Index, for estimating migration at council area.
In addition to the new method and data source for estimating migration, two further adjustments were made to the migration estimates. These took account of information learned about the data sources used to estimate migration while preparing the 2002 population estimates and the population revisions.
1. Unattributable Population Change
The estimated civilian migration component in Table 4 includes an adjustment for recurring unattributable population change. This is assumed to be unmeasured migration. The adjustment is based on the 2001 Census, which showed that previous population estimates had overestimated the population of Scotland by some 50,000. This overestimate was largely due to errors in previous migration estimates in the 1980s and 1990s. To ensure that migration estimates do not continue to be overestimated an adjustment has been included in the 2002 mid-year estimates.
This adjustment is included within the net civilian migration column in Table 4 as it is assumed to be wholly comprised of unmeasured migration. At the Scotland level, this amounts to -2,600 - a breakdown by council area is here.
The 56,400 discrepancy between the census results and the previously rolled forward mid-year estimates comprised an over estimate of 23,670 for 1991 and a further accumulated overestimate of 32,730 in the 1991-2001 period. This was allocated to particular ages and years by the method described in Appendix A of the paper Revised Series of Population Estimates from 1982 to 2000 for Scotland. Investigations by the ONS indicate that 15,000 of this discrepancy can be explained by the method used to gross up the estimates of migrants from the International Passenger Survey (IPS) between 1992 and 1999. This total was subtracted from the 1990s discrepancy, because the current IPS estimates use the new method and therefore will not affect the estimates of international migrants from 2002 onwards. Allowing for this the level of unexplained discrepancy appropriate for 2001-02 was estimated to be 2,600.
The adjustment for unattributable population change was allocated to Council and Health Board areas and given an age/sex distribution based on the distribution of out migrants as estimated in the 2002 estimates of migration.
Further work is being undertaken to review the quality of the method and data sources used to estimate migration. In particular to explain the reasons for the discrepancy and therefore to reduce the level of unmeasured migration. This work will include analysing the detailed migration results from the 2001 Census due out later this summer. In addition, a National Statistics Quality Review of International Migration Statistics is being carried out by the Office for National Statistics. This review, to be published later this year, will recommend ways to improve the quality of international estimates of migration.
2. Record cleaning on the NHSCR
The National Health Service Central Register (NHSCR) records the movements of patients between NHS General Practitioner areas in the UK. Each time a patient transfers to a new NHS doctor in a different Health Board the NHSCR is notified and this is considered to be a migration move. The NHSCR also records migration to and from the UK. A patient registering with a doctor after moving into the UK from abroad will be recorded as an immigrant, and likewise, if a person de-registers with a doctor to move outside the UK the NHSCR will record him as an emigrant. These international moves recorded by the NHSCR are used to distribute Scotland-level international migration obtained from the IPS to Health Boards.
The NHSCR migration data used in the 2002 population estimates showed a particularly large outflow of migrants from Lothian Health Board to overseas. In 2001, emigration from Lothian comprised 24 per cent of the NHSCR international outflow, while in 2002 this proportion rose to 36 per cent. It emerged that cleaning exercises have recently been carried out in both Lothian and Grampian Health Boards to remove records of patients who moved abroad in previous years but whose records were not updated at the time of the move. This caused an artificially high outflow from Lothian between mid 2001 and mid 2002, and it was determined that an adjustment should be made to avoid distorting the international migration for these two areas. Consequently the international outflow for Lothian and Grampian was re-calculated by taking the average of the years 1998 - 2002.
In the longer term these errors will self correct, that is they will be correcting for under or over estimating of out-migrants in earlier years. However, an adjustment has still been applied as the self-correction will not occur given the recent rebasing of the population estimates using results from the 2001 Census.