Adam Afriyie (Windsor, Conservative): To ask Chancellor of the Exchequer
(1) whether a racial impact assessment was carried out before the Office for National Statistics (ONS) changed the mid-year population methodology in 2002; and what consultation was undertaken by ONS before making this change;
(2) whether he has plans to change the methodology used to determine (a) inward and (b) international net migration for use in mid-year population estimates.
John Healey (Financial Secretary, HM Treasury; Wentworth, Labour): The information requested falls within the responsibility of the National Statistician who has been asked to reply.
Letter from Colin Mowl, dated 20 December 2005
The National Statistician has been asked to reply to your recent Parliamentary Questions asking (a) whether a racial impact assessment was carried out before making changes to the mid-year population estimates methodology in 2002; and what consultation was undertaken before making these changes and (b) whether any changes are planned to the methodology used to determine migration for use in the mid-year population estimates. I am replying in her absence.
No significant changes were made to the methodology used to calculate mid-year population estimates in 2002. The estimates published in 2002 were the mid-2001 population estimates and these were the first mid-year population estimates to use the 2001 Census results. The standard ‘cohort component’ methodology was used, in that Census data were updated to reflect births, deaths, migration, and ageing of the population between the date of the Census and the mid-year point (approximately 9 weeks). In a Census year the mid-year estimates are rebased on the most recent Census before the method is applied.
The most recent significant change to the way that the mid-year population estimates are compiled was made in 2000, and applies to the mid-1999 and subsequent estimates. This was in respect of the internal-migration component of the population estimates. This change was widely consulted on before introduction. It was extensively peer-reviewed both within the Office for National Statistics (ONS) and by external academic and local authority experts. The then Liaison Group on Population Statistics (now the Central Local information Partnership (CLIP) Population Subgroup) was also fully consulted on the change. This change, in simplified terms, introduced the use of patient records to allocate migration within (former) health authorities instead of electoral roll information. This was widely thought to be a more reliable method for calculating internal migration. This change was documented in Population Trends 101. No specific racial impact assessment was carried out as part of this methodological change. However, consultation confirmed the recommendation that the methodology should be implemented to improve the quality of population estimates.
A detailed description of the current methodology used to estimate population and the methodology used in a ‘Census year’ is published in Making a Population Estimate in England and Wales”. This document also describes the principal changes that have been made to the methodology over time.
ONS is currently undertaking a major project to research improvements to migration and population statistics (IMPS). Some of the major strands included in this project are the estimation of international migration at the national level, the sub-national distribution of international migration, and the estimation of internal migration. Currently ONS has no plans to implement significant changes to the methodology used in the calculation of mid-2005 population estimates as research and evaluation is ongoing. (Mid-2005 population estimates are due to be published in August 2006.) However, this current position is subject to possible change. Before any changes are implemented, they will be carefully researched, evaluated and where possible assessed for impact. Any changes will be introduced in a planned manner and they will be announced in advance, in accordance with the National Statistics Code of Practice.