| National |
| Product |
|Subnational population projections for England|
| Data |
|Use existing data|
|Frequency||Every two years (variants currently ad hoc)|
| How |
|Administrative data, survey data and projections methodology|
| Geographic |
|Local authority, region and clinical commissioning group|
| Last |
|9 April 2019|
| Related |
| Variant subnational population projections for England: |
Subnational population projections for England: 2016-based
Methodology used to produce the 2016-based subnational
population projections for England
This quality and methodology report contains information on the quality characteristics of the data (including the European Statistical System five dimensions of quality) as well as the methods used to create it.
The information in this report will help you to:
understand the strengths and limitations of the data
learn about existing uses and users of the data
understand the methods used to create the data
help you to decide suitable uses for the data
reduce the risk of misusing data
Subnational population projections provide statistics on potential future population levels, based on the continuation of recent demographic trends and assumptions used in the national population projections.
Projections relate to the usually resident population and do not include people who come to or leave the country for less than 12 months.
They are not forecasts and do not attempt to predict the impact of future political and economic changes, or local development policies.
Projections are used in the household projections and can be used for planning the provision of services such as healthcare and education; they are sometimes used in the assessment of local authority need and the funding formula.
Since projections are produced in a consistent way, they can be used as a common framework for informing local-level policy and planning; local areas are advised to supplement them with any local information they have.
Users should be aware that projections become increasingly uncertain as they go forward into the future, particularly for smaller geographical areas and detailed age and sex breakdowns.
Subnational population projections give an indication of the possible size and structure of the future population, based on the continuation of recent demographic trends. They are produced every two years and project the population for each year of a 25-year period from the base year.
They are produced using the cohort component methodology and are based on the local authority mid-year population estimates. The cohort component method is a standard demographic method that uses high-quality data sources to inform the three major components of population change: natural change (births, deaths and ageing), migration and special populations.
Assumptions made about future fertility, mortality and migration at local authority level are based upon recent observed trends from the components of change, which are published with the latest mid-year population estimates. The assumptions use five years’ worth of trend data except for the 10-year migration variant where 10 years of data are used for the migration components. They are constrained to the equivalent national population projections for England. This means that all local authorities are scaled proportionally such that they sum to the equivalent national population projections for England.
Each new set of subnational population projections supersedes the previous set. Comparisons can be made with earlier sets to show how the projections have changed over time and also to assess their accuracy by comparing projections with the final population estimates for a given year. However, these comparisons are not straightforward. More information can be found in the section on comparability over time.
Variant subnational population projections are produced using broadly the same methods as the main projections, with specific differences dependent on the variant in question.
Projections produced by other organisations may not be comparable with these projections since they will use different methodologies, may not be constrained to national population projections and, in some cases, use additional local data.
More information on the methodology used for the subnational population projections can be found in the section methods used to produce the subnational population projections data and in the methodology document.
Uses and users
Subnational population projections are produced in a consistent way across all areas and use a robust methodology so that they are relevant for all types of users. They are used in a number of ways, including: for local planning of health, education and other service provisions; as a basis for household projections; and as a basis for projections produced by other organisations. Dependent on timing of central government planning rounds, they are also sometimes used in the assessment of local authority needs and the funding formula.
Strengths and limitations
These data provide users with an indication of the potential size, and age and sex structure of the future population of local authorities in England.
Each set of subnational population projections uses the same methods for projecting the population for all local authorities in England, so that data for one local authority are comparable with data for other local authorities.
The assumptions used in subnational population projections are based on past trends. However, demographic behaviour is inherently uncertain, so projections become increasingly uncertain the further they are carried forward.
The subnational population projections may be subject to additional inaccuracy if any of the components such as births, deaths, internal migration and international migration used to produce the projections are inaccurate.
For the 2016-based subnational population projections there were methodological improvements and changes to source data; this is explained in Section 6. A list of methodology changes is also contained within Annex A of the methodology document.Back to table of contents
(The degree to which the statistical output meets users’ needs.)
Subnational population projections are demographic, trend-based projections indicating potential levels of the future population and are used primarily for planning purposes. While the national population projections inform policy and planning at the national level, the subnational population projections inform policy and planning at the local level. Uses include, but are not limited to:
allocation of resources from central government to local areas
healthcare and education provision
emergency service provision
household projections and business development
calculation of local rates, measures and indicators
academic and market research
A robust and objective methodology is employed to create subnational population projections that are relevant for all types of users.
The projections take no account of local development aims, policies on growth, capacity to accommodate population change, or economic factors that could impact the population in the future. As with the national population projections, they also do not try to predict any potential demographic consequences of future political or economic changes, including the UK’s pending withdrawal from the European Union.
Subnational projections are currently produced up to 25 years ahead from the base year. This provides a sufficiently long time series to enable analysis and planning but avoids going too far into the future when values become increasingly uncertain. They can be combined with the population estimates to create a time series from 1971.
Subnational population projections were an important variable used by the Ministry of Housing, Communities and Local Government (MHCLG) in the last assessment of local authority need in financial year ending March 2014 and, depending on the outcome of the Review of local authorities’ relative needs and resources, may form part of a new funding formula from financial year ending March 2021 onwards.
The advantage of using population projections for planning is that the time element is built in. For example, if an area is increasing or decreasing in population over time, then this will be accounted for. However, there is a limitation in that the projections are demographic and trend-based, taking no account of the growth policies of an area. In addition, they do not predict changing demographic patterns over time – they simply provide an indication of population levels arising if the underlying assumptions are realised.
The subnational population projections are also used as an input into the household projections, which were previously produced by MHCLG but are now produced by the Office for National Statistics (ONS). These are used to help regional house planning and monitoring. The household projections themselves are also trend-based projections; therefore, the subnational population projections are suitable for this purpose.
The Department of Health and Social Care (DHSC) uses subnational population projections for resource planning and healthcare provision. The age and sex structure of the projections is important for this purpose. Subnational population projections are supplied at clinical commissioning group level for DHSC purposes.
Local authorities use subnational population projections as a starting point for local level planning and monitoring. Since the population projections do not consider local growth policies, local authorities are advised to also use any local information in producing their plans. The population projections are used in the calculation of local rates and measures, which provide indicators for future requirements of local services.
Part of the usual production of subnational population projections involves consulting with users on the content published in the release, and on the methods used to produce the projections. This process helps us to better understand user requirements.
The production of the 2016-based subnational population projections was unusual as a number of changes were made to the methodology and source data (detailed in this section) and, at the same time, a new production system was introduced. To ensure that the subnational population projections were published to their usual timetable, it was decided not to carry out a full consultation in the usual manner. This approach was also in line with the tighter climate around advance access to data. Instead, a paper was published, which informed users of the changes to the methodology and source data and asked them for their feedback.
The Population Projections Unit, as part of the Centre for Ageing and Demography within ONS, routinely considers what user needs are not being met by their published statistics. This is done using evidence from user engagement activities and contact with users. This process enables the division to understand unmet user needs and, if appropriate, consider the inclusion of new outputs in the divisional workplan.
The 2016-based variant subnational population projections for England were based on views from users as outlined in Section 7 and reflect the most common requests for alternative migration scenarios.
Accuracy and reliability
(The degree of closeness between an estimate and the true value.)
Subnational population projections are demographic, trend-based projections indicating likely size and age structure of the future population if the underlying trends and assumptions about future levels of components of change are realised. They are based on levels of births, deaths and migration observed over a five-year reference period leading up to the base year (or a 10-year reference period for the migration components in the 10-year migration variant).
However, projections are not forecasts and, because of the inherent uncertainty of demographic behaviour, any set of projections will inevitably differ from actual future outcomes to a greater or lesser extent. As such, the subnational population projections should be used as a starting point and supplemented with local information for planning purposes.
The subnational population projections use the latest available population estimates and are inevitably dependent on the accuracy of these estimates. The most recent release of 2016-based subnational population projections is based on the revised mid-2016 population estimates and uses a revised back series of population estimates and component data. If revisions are made to the estimates following a projections release, the projections would be expected to be less accurate as the revisions would not be reflected in the base population used in the projections.
For further information on the accuracy of the population estimates, refer to the Quality and Methodology Information (QMI) report for the population estimates.
Fertility and mortality rates are based upon registration data and are believed to be relatively reliable as the basis for projection. The internal and international migration components of change are more difficult to estimate and project, and this is particularly the case for the local distribution of international migrants.
We ran a programme of work to improve the subnational allocation of both internal and international migrants to and from local authorities. Improvements from Phase 1 of the Migration Statistics Programme (MSIP) were introduced in May 2010 and Phase 2 of the MSIP produced further improvements in November 2011. Subsequent projections all use data produced by the improved methods. Further information on all aspects of the MSIP is available.
More recent improvements in the estimation of emigration at local authority level have been implemented in the revised mid-2012 to mid-2016 population estimates and are documented in the QMI for the population estimates.
In addition, the components of population change are constrained to match those of the national population projections, so the local trends will be scaled. This means that there may be differences between recent trends locally and the assumptions used in the subnational population projections.
For example, five years’ worth of data are used to create the local international migration trends in the subnational population projections, except for the 10-year migration variant where 10 years of data are used. However, the national population projections use different models and a much longer time series in setting the national long-term assumptions. Therefore, it is likely that the assumptions made for international migration will be at a different level to a simple average of the last five (or 10) years’ worth of local data.
Different users may have different opinions on whether five years or 10 years of migration data are most representative of the likely long-term local trend. There is no right or wrong in this; it will vary by area and the users’ individual perspectives.
Users should also be aware, regardless of whether five or 10 years of trend data are used, that any change in actual population dynamics (for example, a speeding up of growth) during the trend period might not be fully reflected in the assumptions being used for future years, as the averaging process has a dampening effect on such changes.
The constraining process also means that any factors that cause the actual population to differ from the projections at the national level will also impact on the projections at the subnational level. Further information on the assumptions used in the national population projections is available in the latest national population projections release. Further information on the accuracy of the national population projections can be found in the QMI for national population projections.
The projections take no account of the following factors, which could lead to differences between the projections and actual population change:
local development aims
policies on growth capacity of a given area to accommodate population change
political or economic factors that could impact the population in future
any international factors that may affect the UK population
Rigorous quality assurance is carried out at all stages of production. Specific procedures include:
scrutinising input data to investigate the accuracy of any abnormal values
scrutinising trends in the total population and components of change projected over time for plausibility
comparing current projections with previous projections and population estimates, to see where large changes are taking place and understand the reasons for these
examining sex ratios to find any areas of imbalance
checking output tables to ensure that there are no errors or inaccuracies during the creation of published tables
Comparisons with previous sets of projections have been carried out to examine how they eventually differed from the final mid-year population estimates for a given year. This gives an indication of how close the projections have been to the estimates when they were published.
An accuracy report was published in August 2015, which considers the accuracy of the 2004-based, 2006-based, 2008-based and 2010-based subnational population projections when compared with the mid-year population estimates for 2011 based on the 2011 Census. Information on the results of this work are summarised in the later section on Comparability over time.
A more recent report was published in August 2016, which compares the mid-year population estimates for 2015 with the 2014-based subnational population projections for 2015. The most recent accuracy assessment of the national population projections was published in July 2015 and considers their accuracy over the last 40 years. The findings take account of the revisions to population estimates following the 2011 Census and are largely similar to the findings of a previous review carried out after the 2001 Census.
Unattributable population change
Following the 2011 Census, a component of population change referred to as unattributable population change (UPC) was defined. This was the remaining difference between the rolled forward 2011 population estimates and the 2011 Census-based population estimates, once methodological changes and estimated errors in the components had been taken into account. The UPC for England was 103,700.
To produce the revised series of population estimates for the last decade, the UPC was apportioned across each of the 10 years using the cohort component method, which takes account of the fact that individuals age as the decade progresses. An explanation of this methodology can be found in the article on the methods used to revise the population estimates.
No adjustment for UPC was made in the 2012-based and later sets of projections or in the series of population estimates based on the 2011 Census. This was because the UPC specific to the previous decade was unlikely to be replicated in continuing subnational trends. Further information can be found in the consultation paper on unattributable population.
Quality assurance of administrative data
In January 2015, the UK Statistics Authority issued the Regulatory Standard for the quality assurance of administrative data. This standard applies to all official statistics where administrative data are used in the production of these statistics. All producers of official statistics that use administrative data need to implement this requirement, by embedding good practice into their production to assure the quality of the data.
In response, the ONS’s Population Statistics Division has published Quality assurance of administrative data (QAAD) reports for all of the administrative datasets that underlie its products. Many of these underlie the subnational population projections either directly or indirectly (via the population and migration estimates).
Issues relating to accuracy of particular components
As far as possible, the projections are produced using a standard method for all areas. In a very small number of cases it is reasonable to believe that mortality, fertility or migration observed in an area over the trend estimation period is a poor indicator of the true underlying trend that should be projected. In such cases, adjustments to improve the likely accuracy of the projections are considered. For the 2016-based principal and variant projections, these adjustments were:
any instances of exceptionally high fertility rates (by single year of age of mother) and mortality rates (by single year of age and sex) were capped at five times the corresponding national figure
Isles of Scilly fertility rates (by single year of age of mother) and mortality rates (by single year of age and sex) were set at the national rates
City of London mortality rates (by single year of age and sex) were set at the national rates
internal out-migration probabilities for Oadby and Wigston males aged 19 to 25 years were set equal to the corresponding probabilities for females, to address the issue of student migration described in this section
All of these adjustments have precedents in previous sets of projections.
Areas with large numbers of students present particular issues in estimating internal out-migration probabilities at ages 20 to 22 years. This is partly because of the known issue of students, especially males, delaying re-registering with a General Practitioner (GP) when they move out of an area at the end of their studies. Particular care should be taken in using or interpreting age distributions in the early 20s for local authorities with substantial student populations.
Coventry and Warwick
A specific adjustment is made in the mid-year population estimates to allow for internal migration to and from a large university campus, which is allocated, based on its postcode, to Coventry, but has halls of residence on both sides of the border between Coventry and Warwick. The subnational projections reflect this adjustment in the base population for the projection. However, the adjustment is not replicated within the projections themselves. The impact of this process is complex but is liable to have a minor impact over time on the size and age structure of Coventry and Warwick’s projected populations.
City of London and Isles of Scilly
Reliable projection of trends is particularly difficult for the two very small local authorities, City of London and Isles of Scilly. As noted previously, national mortality and fertility rates have been used in place of the observed local authority rates where thought appropriate. Users are advised to take particular care when using projections for these areas.
Clinical commissioning groups
Projections for clinical commissioning groups (CCGs) are derived from the projections for local authorities. Where a CCG contains only part of a local authority, it is implicitly assumed that rates of population change in that part of the local authority are the same as across the local authority as a whole. Users of CCG projections are advised to consider how any deviation from this assumption might affect their conclusions from analysis of the projections.
The 2016-based population projections by CCG have an additional issue regarding accuracy. Revised mid-2016 Lower layer Super Output Area (LSOA) population estimates needed to produce the projections by CCG were not available at the time of production, so unrevised mid-2016 LSOA population estimates published in October 2017 were used. Because the 2016 values in the projections by CCG are created by aggregation and/or apportionment of the projections by local authority, they differ from the CCG population estimates published in October 2017. They also differ from the revised mid-2016 CCG population estimates that were published in October 2018. However, the difference in the mid-2016 CCG values between the projections and revised estimates was sufficiently limited that it was decided not to revise the projections.
Coherence and comparability
(Coherence is the degree to which data that are derived from different sources or methods, but refer to the same topic, are similar. Comparability is the degree to which data can be compared over time and domain, for example, geographic level.)
The subnational population projections are consistent with the mid-year population estimates at local authority level, which form the starting point (base year) for projecting forward. They are also consistent with the national population projections for a given base year.
“Consistent” in this instance means that the total population and components of population change across all local authorities in England sum to those presented in the national projection for England. The exception to this is the component for international inflow. This is because people from Syria granted humanitarian protection, as well as armed forces returning from Germany and their dependants, are not included in the international inflow as, unlike in the national projections, they are added separately. However, these groups are included in the total projected population and so the subnational population projections remain consistent with the national population projections.
Usually, the population projections are consistent with the mid-year population estimates at clinical commissioning group (CCG) level for the base year of the projection. The 2016-based subnational population projections are an exception to this and do not match the mid-year population estimates for CCGs. This is because revised population estimates at lower geographies, needed to produce the projections at CCG level, were not available when production was taking place. A solution was developed to ensure that the 2016-based subnational population projections for CCGs were published (as described previously).
Subnational population projections across the four nations of the UK have broadly similar methods. Like those for England, the projections created by National Records of Scotland and the Northern Ireland Statistics and Research Agency are also constrained to the national population projections. However, those projections created by the Welsh Government are not constrained to the national population projections.
Further details about the differences between the subnational projections of the different countries of the UK can be found in the article on subnational population projections across the UK. This article reflects the recent changes to the methodology and data sources in the 2016-based subnational population projections for England and any changes made by the other countries of the UK since the paper was last updated in 2019.
Subnational population projections may be generated by other organisations. For example, some local authorities and academics produce their own set of subnational population projections. However, these often use a different methodology from that used for the ONS projections and therefore produce different results. Important methodological differences include:
not taking account of moves between areas in the same way
trying to take account of economic factors or local policy decisions about issues such as future population levels, housing development and economic initiatives
not needing to apply a consistent and objective methodology for all areas within England
not needing to constrain to the national population projections nor follow the assumptions associated with them
using different definitions of migration and population
When attempting to compare ONS projections with projections created by external organisations, careful attention must be paid to any differences in assumptions, methodologies and definitions being used.
The Greater London Authority produces projections of both the size and characteristics of the future resident population for all local authorities in London. In addition, they now produce projections for all local authorities in England. This is primarily to assist in strategic planning across the wider region and to help those authorities that border London to understand the implications of their projections. Users should be aware that these projections are produced using a different methodology to that used by ONS.
A challenge of producing projections and a time series based on historical data is that the quality of the data may not be consistent over the designated time period. For example, the 10-year migration variant was produced with input data whose methodology changed during the 10-year period. This means that any changes over time are liable to reflect a combination of both real world and methodological differences.
Comparability over time
Each projection is internally consistent, so it can be interpreted as a consistent time series of projected populations for each year of the projection period.
Comparability over time can also be used to describe the comparability of different sets of projections. Each set of subnational population projections is unique and is produced using trends based on the best data available at that time, including the latest population estimates. Therefore, each new set of projections supersedes the previous set. Although projections are broadly comparable over time, like-for-like comparisons are not straightforward. However, it is possible to observe what effect the most recent demographic trends, when built into projections for the future, have on the possible future population of local areas.
Subnational population projections dating back to 2010-based are available. Earlier sets of projections are available on the archived website and additional sets are available from the Stakeholder Engagement Team at firstname.lastname@example.org.
There are a number of issues that need to be considered when different sets of subnational population projections are compared. These issues are summarised as follows:
until the 2003-based subnational population projections, the full 25-year period was only available every four or five years; in intervening years, short-term 10-year projections were produced using the same assumptions on levels of components of change as the previous set of long-term projections, but updating the base-year populations to take account of latest population estimates
the 2003-based subnational population projections were an interim set that was produced when methodological changes in the population estimates resulted in a revised back series
the 2004-based subnational population projections used a different methodology for the distribution of international migration than previous sets of projections; the method used was brought more into line with that used for the production of the official mid-year population estimates
the revised 2004-based subnational population projections updated the 2004-based projections to take account of revised population estimates using new methods introduced in 2007; the same methodology was used for the 2006-based projections
the 2008-based subnational population projections used a different methodology for the distribution of internal and international migration than previous sets of projections as they incorporated further developments of the Migration Statistics Improvement Programme (MSIP), and no longer used the Rogers-Castro curve when calculating internal migration
the 2010-based subnational population projections used a different methodology for the distribution of international immigrants, which in turn affected estimates of emigrants, and also included improvements to internal migration of students; further details are available from the release by MSIP
the interim 2011-based subnational population projections used the mid-2011 population estimates rolled forward from the 2011 Census results as the base, but the assumptions made on future trends were the same as those used in the 2010-based projections
the interim 2011-based, the 2012-based and the 2014-based subnational population projections incorporated information from the 2011 Census so care is needed if comparisons are made with earlier sets of projections; in addition, the new internal migration methods were introduced and used in the 2012-based subnational population projections onwards
the 2016-based subnational population projections include the revised back series of population estimates and components of change, published in March 2018; they also include a number of changes to the methodology and source data, which are described in the information paper published in January 2018
An accuracy report was published in 2008 that considered the accuracy of the projections from the 1996-based subnational population projections to the revised 2004-based subnational population projections. An updated accuracy report considers the accuracy of the 2004-based, 2006-based, 2008-based and 2010-based subnational population projections when compared with the mid-year population estimates for 2011 based on the 2011 Census. The report concludes that it is difficult to determine what is causing the differences seen, because of the many changes in methodology and resulting revisions that occurred during the last decade, outlined previously.
Accessibility and clarity
(Accessibility is the ease with which users are able to access the data, also reflecting the format in which the data are available and the availability of supporting information. Clarity refers to the quality and sufficiency of the release details, illustrations and accompanying advice.)
Subnational population projections are available online and can be downloaded free of charge in Microsoft Excel and CSV format. Graphs, textual background information and supporting documents are provided as part of each release.
Any additional enquires regarding subnational population projections can be made via email to email@example.com or by telephone on +44 (0)1329 444661. Additional data requests can be met where this is possible. Metadata describing the limitations of additional data are provided with individual requests. These requests are also published on the ONS website.
Our recommended format for accessible content is a combination of HTML webpages for narrative, charts and graphs, with data being provided in usable formats such as CSV and Excel. We also offer users the option to download the narrative in PDF format. In some instances, other software may be used, or may be available on request. For further information please refer to the contact details at the beginning of this report.
For information regarding conditions of access to data, please refer to the following:
Timeliness and punctuality
(Timeliness refers to the lapse of time between publication and the period to which the data refer. Punctuality refers to the gap between planned and actual publication dates.)
Subnational population projections are typically published in May, around two years after the base year. For example, the 2016-based subnational population projections were published in May 2018. This timeframe occurs because the subnational population projections are based on input data that are not available until late in the year after the base year. The time between then and May is needed for the production and quality assurance of the projections.
This timetable ensures that the projections are available to two important customers, the Ministry of Housing, Communities and Local Government (MHCLG) and the Department of Health and Social Care (DHSC), for their use in resource allocation and planning. The timetable is similar to projection timetables used in other countries. In terms of punctuality, all previous sets of projections have been published to schedule on pre-announced dates.
In special circumstances, interim sets of projections may be produced, using modified methods and the most recent data available. An example of this is the interim 2011-based subnational population projections, which were produced shortly after the 2011 Census results. This set of projections was published to satisfy a strong user requirement for projections that took on board the results of the 2011 Census, sooner than the normal publication timetable.
For more details on related releases, the release calendar provides 12 months’ advance notice of release dates. In the unlikely event of a change to the pre-announced release schedule, public attention will be drawn to the change and the reasons for the change explained at the same time, as set out in the Code of Practice for Statistics.
Concepts and definitions (including list of changes to definitions)
(Concepts and definitions describe the legislation governing the output and a description of the classifications used in the output.)
The overview of population and migration statistics explains the concepts and definitions used in population projections.
A conceptual framework for population and migration statistics (including the population estimates) is available.
Population estimates and projections estimate the “usually resident” population only. This is the standard United Nations definition and includes only people who reside in a country for 12 months or more, making them usually resident in that country. As such, visitors and short-term migrants are excluded.
Components of change
These can be described as population changes between one year and the next year due to various components. The two categories of population change are natural change (the difference between births and deaths) and net migration, covering movements of people between England and the various countries of the world (international migration), between England and the other countries of the UK (cross-border migration) and between local areas within England (internal migration). By projecting estimates of these various individual components of the population, their effects can be added together to provide a projection of the population at selected points into the future.
This is 30 June of any given year, where the period from one mid-year to the next is from the first day of July year x until 30 June of year x+1.
Geography (including list of changes to boundaries)
The 2016-based subnational population projections provide information at region, county, local authority, clinical commissioning group and NHS region levels, using the geographic boundaries, names and codes existing as at mid-2016. All three migration variant projections use the same geographies and boundaries as the 2016-based subnational principal population projection. They do not incorporate any change made after 2016.
This report provides a range of information that describes the quality of the data and identifies the issues that should be noted when using the output.
We have developed Guidelines for measuring statistical quality based upon the five European Statistical System (ESS) quality dimensions. This report addresses the quality dimensions and important quality characteristics, which are:
timeliness and punctuality
coherence and comparability
output quality trade-offs
assessment of user needs and perceptions
accessibility and clarity
Subnational population projections give an indication of the possible future resident population in each administrative area (regions, counties, local authorities and health areas) in England by sex and single year of age. It is currently not possible to calculate projections for any further breakdowns such as ethnicity, marital status or lower-level geographies, because of limitations in the availability of data and the lack of a robust methodology required for such projections.
The 2016-based subnational population projections are published unrounded, by sex and single year of age, to enable users to carry out further analysis. However, the projections are not presented as reliable at this level of detail and so users are advised to aggregate the data to at least five-year age groups and round to the nearest 100 people if quoted in any documentation.
Occasionally there may be significant user demand for a set of projections at a point in time when only some of the required input data are available. For example, when new census results are published, there is a time lag before rebased population estimates are available to provide base data and a further lag until a revised historic series becomes available to calculate the trend data.
Depending on the timing of the projections, this may affect the quality of the output produced. For example, for the interim 2011-based subnational population projections, the population estimates were updated with census results making the base more accurate but as trend data were not updated at the same time, the resulting projections of future change may not have been as accurate as they could have been.Back to table of contents
How we collect the data, main data sources and accuracy
Subnational population projections are calculated using the internationally-recognised cohort component method and are consistent with the national population projections based on the same base year. They use the national population projections, the local authority population estimates and associated components of change, which inform projected births, deaths, migration (internal, cross-border within the UK and international) and asylum seekers. More information on how these components are derived can be found in the Population estimates Quality and Methodology Information (QMI) report.
In addition, the projections use data on home and foreign armed forces and data on the dependants of foreign armed forces. Data on UK armed forces returning to England from Germany and their dependants are also used, as are data on people from Syria granted humanitarian protection.
The 2016-based projections take the local authority 2016 mid-year population estimates as their starting point. Change in population is calculated by modelling trends. Data for five preceding years are used, so for example, the 2016-based projections trends are based on data from the years ending mid-2012 to mid-2016. The projections based on these trends are constrained to the assumptions made for the principal national population projection for the equivalent base year.
The projections model splits population between the armed forces and civilian population and treats them differently. The population of armed forces is treated as a “static population” whose size, and age and sex structure is assumed to be constant over the projection period. These armed forces include home and foreign armed forces based in England, UK armed forces returning to England from Germany and dependants of US armed forces. Dependants of home armed forces, which include dependants of UK armed forces returning to England from Germany, are treated as part of the civilian population.
A summary of the major steps in the production process is as follows:
start with the mid-year population estimate for the base year
remove the static population to produce a civilian population
age on the population by one year
apply local fertility and mortality rates to the aged-on population to calculate projected numbers of births, which are added to the population, and projected deaths, which are taken out of the population
adjust the population for internal migration (movement between areas within England) and cross-border migration (movement between England and the other countries of the UK)
adjust the population for international migration (movement between England and countries outside of the UK); this includes asylum seekers, people from Syria granted humanitarian protection and dependants of returning armed forces from Germany
add back the static population
constrain the sum of local authority population projections to the population projection for England for that year
the population at the end of each cycle becomes the base population of the next cycle
As an example, for the 2016-based subnational projections, the change projected for mid-2016 to mid-2017 is applied to the mid-2016 base to produce a population projection for mid-2017. This process is repeated for each year of the projection period. Each component (except internal migration) is constrained to its respective total in the 2016-based national population projection for England.
The 2016-based subnational projections included, for the first time, variant projections featuring different levels of migration. The high international migration and low international migration variants are produced broadly using the same methods as the 2016-based subnational population projections main release (the principal), except that the totals are constrained to match those in the 2016-based high and low migration variant national population projections for England.
The 2016-based 10-year migration variant uses 10 years (years ending mid-2007 to mid-2016) of input data for international migration (including asylum seekers), internal migration and cross-border migration. The 10-year migration variant is consistent with the principal subnational population projections in that all migration components are constrained to the principal 2016-based population projection for England. However, because it uses 10 years of migration input data rather than five, the distribution of migration at local authority level is different.
More detailed information can be found in the subnational projections methodology report.
Recent changes to the methodology and data sources
A number of changes were made to the methodology used to produce the 2016-based subnational population projections. These are summarised in this section; more detail can be found in the subnational projections methodology report.
Estimates of emigration for local authorities in the population estimates are created by using a statistical model based on data from the International Passenger Survey and other data sources to create robust estimates. The statistical model has been improved (by modelling rates rather than counts, using different explanatory variables and no longer constraining to new migration geographies for outflows) and these improvements have been implemented in the revised population estimates and consequently in the trend data for emigration used in the projections.
In addition to this change in the source data, the methodology to project emigration has changed. Previously, the methodology for projecting emigration used a model that applied a complex weighting system to the emigration estimates for the last six years and used the new migration geographies, which are no longer used in the emigration estimates. The new methodology removes the new migration geographies from the process and projects emigration based on a simple five-year average of trend data (or 10 years in the 10-year migration variant).
While the methodology used to project immigration remains unchanged in the subnational projections, some of the source data used to project immigration have been revised. Estimates of immigration for local authorities in the population estimates are created by distributing migration estimates directly from the national to local authority level using administrative data sources.
Supply of more up-to-date administrative data and a change in processing environment have meant that the immigration estimates, used to project immigration, have changed for the years ending mid-2015 and mid-2016. This is discussed in more detail in the documentation published in March 2018 to support the release of the revised back series of population estimates.
Dependants of US foreign armed forces
The revised population estimates treat the dependants of UK armed forces as a special population rather than including them in the civilian population. This is because dependants of UK armed forces are recorded in the census but have relatively little interaction with the standard administrative sources used in the population estimates process.
Dependants of US armed forces are now treated as a static population in the subnational population projections alongside other foreign armed forces. An adjustment is made for the US armed forces dependants aged under one year to prevent double-counting of births. Treating the dependants of US armed forces as a static population counters the inaccurate ageing on of women of childbearing age and reduces the imbalance in the sex ratio in local authorities with a large US armed forces presence.
Returning armed forces from Germany
Recent sets of national population projections have made an allowance for the planned return of armed forces and their dependants from Germany by 2020. In the 2014-based subnational population projections, returning UK armed forces and their dependants were distributed to local authorities in England based on the distribution of international immigration to local areas, with the exception of Wiltshire where we had very clear evidence on the likely net effect of the return on the armed forces population in the area. The 2016-based subnational population projections have used data from Ministry of Defence and British Forces Germany, to allocate returning UK armed forces and their dependants to the local areas where their units are due to be based.
The approach to projecting cross-border flows (that is, flows between countries in the UK) was changed in the 2014-based national population projections to be based on rates of the population rather than absolute numbers. Although rates are not used in the projected cross-border flows at subnational level, they are constrained to the cross-border flows in the national projections. Therefore, the improvements in the national population projections have some impact at the subnational level.
In the 2016-based subnational population projections, the methodology was further improved. The trend in cross-border flows was previously calculated using recorded moves between, for example, England and other countries of the UK. This trend, using five years of cross-border migration data (or 10 years in the 10-year migration variant), is now calculated at individual country level. This means that, for example, cross-border flows between England and Scotland in the subnational projections are calculated, which are then constrained to the cross-border flows between England and Scotland in the national projections.
People granted humanitarian protection
The 2016-based national population projections included people from Syria granted humanitarian protection under the Vulnerable Persons Resettlement Scheme. Currently, the national population projections contain around 10,000 people expected to be granted humanitarian protection in England between mid-2016 and mid-2020. These people have been included in the 2016-based subnational population projections by using the distribution of Home Office data on where these people were received in each local authority over the period between mid-2015 and mid-2017.
In earlier sets of projections, the most recent year of data on asylum seekers at local area level was used and constrained to the total in the national population projections for the first two years of the projections. The projected number of asylum seekers was then assumed to remain constant for the remaining years of the projection. The 2016-based projections are the first set to use a five-year average of data on asylum seekers at local area level; this is constrained to the total in the national population projections for the entire projection period. The 2016-based 10-year migration variant uses 10 years of asylum seekers data in the same way.
How we quality assure and validate the data
The 2016-based subnational population projections use data from various sources including population estimates, births, deaths, internal migration, cross-border migration, asylum seekers, international migration and people from Syria granted humanitarian protection. All of these have undergone quality checks by data suppliers. Within the projections process, each local authority projection is quality assured and validated through a range of quality assurance processes to ensure the components are aligned with what is expected for each area and the total figures are constrained to the national figures.
How we disseminate the data
The subnational population projections and variant subnational population projections are disseminated through the ONS website. Stakeholders are alerted to new releases through various media channels, email and newsletters.Back to table of contents
Assessment of user needs and perceptions
(The processes for finding out about uses and users, and their views on the statistical products.)
As stated in the Relevance section previously, it has been normal practice to consult with users as a standard part of the release cycle. Evidence from these consultations, along with any other information from user engagement, provides an opportunity for new requirements to be considered and for any problems with the projections to be highlighted and reviewed.
The Office for National Statistics (ONS) Projections Team regularly provides advice and support to customers on how to use projections, and in doing so the team keeps a record of user requirements. Such needs are considered during production of the statistics and, where possible, provision is made to accommodate new customer needs. Any unmet user need is considered as part of the future divisional work plan.
In 2016, we consulted with users on how we could reinvest resources to develop the skills and technology needed for the UK's future statistical needs. This consultation included a review of our outputs and included the proposal to produce subnational population projections once every three years rather than every two years.
The majority of respondents said that the proposal would have a high impact, because of the importance of population projections for town planning, policy development and allocation of funding and resources. Some respondents highlighted that they notice significant changes in local populations in a short period of time, so extending the time between editions of subnational population projections would be problematic. As a result, it was decided not to change the frequency of population projections. The response to this consultation is available.
The most recent Centre for Ageing and Demography user engagement events were held in February 2016 and April to May 2018. These events updated users on the Centre’s work and gave users an opportunity to provide feedback.
The 2016-based variant subnational population projections for England were based on feedback from the 2014-based subnational population projections consultation, the 2016-based subnational population projections user engagement exercise and other user forums on changes to methods and new data sources. A review of the responses from the user engagement events and activities led to the production of three migration variants: high international migration, low international migration and the 10-year migration variant, as these reflected the most common user requests for alternative migration scenarios.
Variant subnational population projections for England: 2016-based. Supporting documentation includes a methodology document and a quality and methodology information report (this report).
2016-based subnational population projections: frequently asked questions provides answers to questions on the 2016-based subnational population projections.
Subnational population projections for Scotland are available from the National Records of Scotland. The 2016-based subnational projections were published on 28 March 2018.
Subnational population projections for Wales are available from Stats Wales.
Subnational population projections for Northern Ireland are available from the Northern Ireland Statistical Research Agency.
Subnational population projections across the UK: a comparison of data sources and methods summarises the methodologies used to produce the subnational population projections across the UK.
The latest subnational population projections for Scotland and Northern Ireland are 2016-based as, like England, they operate on a two-year cycle. They were published on 28 March and 26 April 2018 respectively. The latest subnational population projections release for Wales is 2014-based as they operate on a three-year cycle. The 2017-based subnational population projections for Wales are due to be published in summer 2019.
In January 2017, the responsibility for household projections was transferred to the ONS. The most recent set of household projections for England are the 2016-based, which include Household projections in England: 2016-based and Household projections in England – household type projections: 2016-based. Previous releases of household projections can be found on the Ministry of Housing, Communities and Local Government (MHCLG) website. Queries on household projections should be directed to firstname.lastname@example.org.
A variant subnational population projections paper was published in August 2016, detailing experimental work on “proof of concept” 2014-based variant subnational population projections.Back to table of contents
Contact details for this Methodology