2001-2004 Small Area Population Estimates, Scotland

Contents

• Background
• Project progress
• Release of statistics and user consultation
• Future workplan
  • Feedback
  • Development of methodology
  • Related products
  • Further documentation
• Tables
• Methodology

Background

The National Records of Scotland (NRS) has been involved in work to produce small area population estimates for a number of years. Initially this work was motivated by resource allocation work in the Scottish Government Health Department, but it was latterly funded by the Scottish Neighbourhood Statistics project.

This work resulted in ward based Small Area Population Estimates (SAPE) published for 1999 and 2000, using an apportionment method based on the Community Health Index (CHI). The 2001 Census gave an opportunity to quality assure this work but it also provided a new baseline for the application of the cohort component method. (In this method a baseline population is modified by ageing it and adjusting it with information on births, deaths and migration). The conclusion of this work was that the apportionment method was unsatisfactory, and that the preferred method was the cohort component method. A ward based report had been made available on the web, although there were quality issues identified as part of an exercise comparing the migration component of SAPE with the census migration data.

NRS has been developing the cohort-component methodology to produce a data zone SAPE since June 2004, and have now produced estimates for mid-year 2001, 2002, 2003 and 2004. A brief description of the methodology is outlined in section 2 of the methodology paper '2001-2004 Small Area Population Estimates- Methodology'.

Project progress

The project has been guided by a Small Area Population Estimates Working Group. Papers from meetings of the working group are available on this website.

Release of statistics and user consultation

Small area population estimates consistent with the Registrar General's mid-year population estimates for Council areas are available for data zones by gender and five-year age groups [Footnote 1].

The data have been derived from the 2001 Census using the cohort-component method where data zones were created by aggregating census output areas. More information on the data zone geography can be found in the 'Scottish Neighbourhood Statistics Data Zones Background Information' page on the Scottish Government website. A CD containing a look up from data zone to various other geographies including council area can be obtained by contacting Neighbourhood Statistics.

Quality assurance of the data has taken place by consulting with local authorities and assessing the population trends against trends in the assessor dwelling counts and the NRS postal address file. Refer to 'A Users Guide to NRS Geography Products' on this website. This process has resulted in adjustments being made to the populations of about two per cent of data zone areas, particularly those with high transient populations such as student areas. The main reason for these adjustments is that the data source used to estimate migration (the community health index) does not accurately pick up all student moves, especially for small areas such as data zones.

In addition to the mid-year estimates for Council areas, NRS also makes available population estimates for Health Board areas. The small area population estimates presented here have been made consistent with Council area totals. However, these estimates may not be completely consistent with Health Board totals, as neither Census Output Areas nor data zones nest exactly into Health Board areas. Users should be aware of the inconsistency and avoid amalgamating the data zones presented here to estimate the population for Health Board areas. Health board populations for 2004 can be found in the publication 'Revised Mid-2004 Population Estimates, Council and Health Board Areas'.

As we have some concerns about how the methodology fares in areas with a high percentage of students and armed forces populations, feedback or comments on them will be greatly appreciated. Any queries about the statistics in this dataset, or feedback on the quality of the estimates, should be directed to Statistics Customer Services using our Contact us page.

Future workplan

Following this publication, the project moves into a new phase of obtaining feedback and investigating alternative methodologies for future publications. In addition, a number of related outputs will be produced. Details of the future work proposed follows:

Feedback

Comments on the quality of the estimates will be sought through a web based questionnaire which accompanies this publication. In addition, we will identify key users to contribute to the feedback process through our working group. 

It is planned to ask for feedback for at least 12 weeks following publication and to report on this in March 2006.

A copy of the feedback questionnaire, 'Consultation on NRS mid-2001, mid-2002, mid-2003 and mid-2004 data zone population estimates' (Adobe Acrobat Portable Document Format (PDF) 345 Kb), can be found on this website.

PDF files require Acrobat Reader. Download Acrobat Reader free

Development of methodology

NRS will carry out research to compare the SAPE with a range of administrative datasets. This research will compare both the levels of the estimates and the relative changes in these data. The administrative datasets to be considered will include Community Health index based patient data, child benefit data, school census data, and the 'Super Older Persons Database' derived from individual DWP databases for Attendance Allowance, Disability Living Allowance, Widows Benefit, State Pension, Incapacity Benefit, Winter Fuel and Minimum Income Guarantee, and covers persons aged 65 years and over.

The purpose of this research will be to:

1. provide guidance for users on the strengths and weaknesses of the current estimates;
2. identify the potential of these sources for use in the ratio change method. This is the preferred method used by the Office for National Statistics in their SAPE project, which takes a  base population and applies changes to it based on the changes in a range of administrative sources. 

Related products

A number of  related products for which there would be customer interest have been identified by the working group and from previous consultation. These include:

• Single year of age estimates - required to allow more flexible age groups to be derived;
• Postcode counts – required to allow the production of population figures for non-data zone based areas including settlements and areas exposed to particular environmental pollutants;
• Measures of turnover and decline in an area – to identify and monitor characteristics of areas.

The project will investigate the feasilbility of producing these outputs and prioritise this work. This will involve development of the methodology, as the current approach which uses the Community Health Index cannot be used directly. The current agreement with the health service which governs the use of the CHI by NRS, requires that estimates are produced for age bands. In addition, the method is increasing unreliable at very low levels of disaggregation.

Further documentation

The methodology paper '2001-2004 Small Area Population Estimates - Methodology' can be found on this website.

Work undertaken by the Office of National Statistics (ONS) to produce small area population estimates for England & Wales can be found in the 'Postcensal small area population estimates' page on the ONS website. 

Tables

Note:
On 27 July 2007 the Mid-year population estimates for 2003-2006 were revised for the six Council areas of Aberdeen City, Aberdeenshire, Angus, Dundee City, Moray and Perth & Kinross. The statistics available on this page for 2003 and 2004 were produced using the revised population estimates. These revised statistics replace the previously published figures. More details on the 'Revised Mid-year Population Estimates 2003-2006' are available from this website.

The estimates have been made available as Excel and CSV (Comma separated Value) spreadsheet files. When opening the spreadsheet files your browser, depending on how it has been configured, will prompt you to either open the file or save it to disk.
Maximum file sizes are Excel 1.6 Mb and CSV 652 Kb.  

Footnote

1. An exception is the age range 10-19 which is split into 10-15 and 16-19 age groups. This has been done to allow the creation of the broad age groups children (0-15), working age (16-59/64) and pensionable age (60/65 and over).