1. Introduction
The Oxfordshire housing-led population forecast provides resident population estimates for Oxfordshire by district council and small area (Middle Layer Super Output Area, MSOA). The data is provided by single year of age and gender. This model was developed by Oxfordshire County Council.
This model was developed to provide future population estimates for geographic areas smaller than those produced by the Office of National Statistics (ONS). This supports localised service planning, including school places, social care, and leisure provision. The model also matches the population to the available housing, so can constrain or increase population growth based on known housing growth in the local area. It is important to note that the rate of housebuilding is an input to the model and the outputs reflect the expected population if that level of housebuilding takes place. The forecasts, therefore, are not indicative of housing need, and should not be used or interpreted as such.
2. Methodology overview
The methodology is initially similar to the Office of National Statistics (ONS) methodology for their population projections but is extended to assess whether the projected population can be accommodated within the available housing stock.
The model uses a two-stage process:
The first stage is to calculate the population change based on life events (births and deaths) and migration. This is the same methodology as the ONS use for population estimates and projections.
The second stage checks whether the population projected in Stage 1 can be accommodated within the available housing stock by comparing housing demand with available dwellings.
3. Base population and ageing-on
The latest ONS mid-year estimates are used as the starting (base) population for the model. This is then used as the basis for applying changes in population rates (e.g. death rates) later in the model. For subsequent years, the base population is the previous year’s population after ageing-on.
Data used
The base population is based on the 2024 Mid-year estimates (MYE) from the ONS. With the Council geographies published on 30th July 2025 and Small Area (MSOA) published on 7th November 2025. The recent 2025 mid-year estimates are not used as they came out too late for this update and the MSOA data is not published until November.
4. Births
Births are modelled using Age-Specific Fertility Rates (ASFRs) for each age of mother, which are multiplied by the number of women in each age group within the area. ASFRs are calculated by dividing the total births by age of the mother by the total number of women in that age band. This is worked out for women of child-bearing age which is defined as 15 to 44 year olds. The ASFR may be adjusted for future years based on changes in past trends (e.g. women having babies later in life).
Data used
District-level data use the latest 2022-based Subnational Population Projections (SNPP) ASFRs as these reflect the national changes locally as women have babies later in life Methodology used to produce the 2022-based subnational population projections for England – Office for National Statistics. The next SNPP are due in 2028.
As national data are not available below local authority level, MSOA level ASFRs are calculated as a three-year average of live births divided by the female population in the MSOA from the Mid-year estimates. The age profile of the mother is based on the district breakdown but adjusted to match the total for each MSOA. https://www.ons.gov.uk/peoplepopulationandcommunity/birthsdeathsandmarriages/livebirths
5. Deaths
Deaths are modelled at each age and sex by multiplying the population in each age and sex group by the age-specific mortality rate (ASMR).
The ASMR is calculated for each age and sex by looking at historical deaths and dividing them by the total population in that group. This gives a likelihood of someone passing away at that age/sex.
The MSOA data used depends on the number of deaths in the area. For low numbers of deaths (less than 10 a year) ASMRs from a larger geography may be used (e.g. parent district or South East region, if the district is also low numbers). In all geographies the ASMR uses the average of the last 5 years mortality rates. Changes in ASMRs are accounted for by adjusting of the sum of the MSOA population to match the district population in the final step of the model.
Data used
The district level data is using the latest 2022-based SNPP ASMR as these reflect national trends that forecast increasing life expectancy, Methodology used to produce the 2022-based subnational population projections for England – Office for National Statistics. The next update of the SNPP is 2028.
The MSOA data is created from the ONS death registrations and the Mid-year estimates. Mortality statistics – underlying cause, sex and age – Nomis – Official Census and Labour Market Statistics
6. Migration
The migration component of population change relates to people moving into or out of an area. This can be from other parts of Oxfordshire, the rest of UK or internationally.
At District council level this is modelled using the methodology for the national population projections. For internal migration (within the UK and between areas), this is based on the average of the 10 years of historic component of change of the mid-year estimates. For international migration, council-level data follow the national projection methodology, which assumes migration gradually moves from the current higher levels towards the long-term pre-2022 average.
Migration rates at small-area (MSOA) is estimated by creating a model of historical data looking at the change in population of people born in the same year. Migration is the change left after removing the effects of natural change (births and deaths).
For example an area may have 50 people aged 30 and the next year have 60 people aged 31 but one person had died, this gives a migration of 11 people
Static migration relates to using a fixed number instead of the rates described above. This approach is used for areas that experience large migration changes in a single year compared to the total population. These are typically areas with large communal establishments e.g. boarding schools, university halls & military barracks.
Data used
For council migration the historical data is taken from the Mid-Year Estimates’ component of change Estimates of the population for England and Wales – Office for National Statistics. The methodology for the producing the forward trend for migration is based on the National population projections, migration assumptions: 2024-based – Office for National Statistics
Communal establishment data are compiled from the following datasets: Using CQC data – Care Quality Commission for care homes, Get Information about Schools – GOV.UK for boarding schools, Communal establishment residents, England and Wales – Office for National Statistics and Output Area level population from the Mid-year estimates (see base population).
7. Housing
Unlike trend-based ONS population projections, the County Council population model includes assumptions about future growth in housing within Oxfordshire, provided by the latest planning reports published by Oxfordshire’s district councils. These publications include Annual Monitoring Reports or 5-year land supply. Housing sites have been mapped to districts and middle layer super output areas to match the data available for the above steps. The number of available homes in each year equals the previous year’s housing stock plus any completed new-build dwellings.
To determine the relationship between the population and housing stock, we use Household Representative Rates (HRR). HHRs are the proportion of people within an age group and sex being the household reference person (HRP, formally called head of the household). The HRP is determined based on economic activity, age and/or sex of the household and is normally the oldest person who is working. The projected population from the above stages excludes people living in communal establishments as these don’t form households. The non-communal population is divided by the HRR to produce the number of homes they require.
The final population for the next year is calculated by taking the projected population and adjusted according to housing surpluses or deficits. This step moves people in or out of the area depending on the availability of housing. The Population Translation Multiplier (PTM sometimes called propensity to migrate) converts a housing surplus or deficit into an equivalent population adjustment. The PTM value is the number of people that would likely move into a house with the breakdown of the people by age and gender.
If the number of available homes exceeds the number required, the model moves people into the area. Conversely, if available homes are fewer than those required, the model moves people out of the area.
Data used
Communal establishments data see migration section
Annual Monitoring Reports and 5-year land supply for each District Council for new builds data
Five year land supply – South Oxfordshire District Council
Vale 5YHLS Statement December 2025 PDF
West Oxfordshire AMR 2024–25 PDF
Final steps
At this stage, the MSOA populations are adjusted so that they sum to the corresponding district council totals by age and sex. This adjustment is normally small but increases in later years as the model gets more uncertain but is less than 2% adjustment.