Census 2021 outputs: content design and release phase proposals

Closes 5 Oct 2021

Section 5: Population-base specifications

Main population bases

In The Census 2021 – Initial view on content for England and Wales consultation in 2015, we assessed that we needed to produce the same output bases as in 2011. We’ve included details of that assessment in the report Outputs and enumeration bases.

Definitions for most Census 2021 population bases will remain largely the same as those we used in the 2011 Census. We’ve included details of these changes in the report Output and enumeration bases: residential address and population definitions for Census 2021.

The main population bases for Census 2021 are:

  • usual residents
  • households
  • usual residents in households
  • communal establishments
  • usual residents in communal establishments
  • household reference persons (HRP)

We propose that users will be able to build their own tables using flexible datasets, based on these populations. A proposed list of possible variables that we could make available in each flexible dataset is in the Draft proposals for outputs data content (XLS, 610KB).

Secondary population bases

We will provide outputs based on the following population bases.

  • Dwellings
  • Families
  • Dependent children

These populations will not initially be available as flexible datasets because of their complexity. However, we will provide them as ready-made tables. There may be scope for developing flexible datasets based on these populations as the outputs develop over time.

Place of residence

In all of these population bases, usual residents are counted at their place of residence.

A usual resident of the UK falls into one of the following categories. They’re anyone who, on 21 March 2021:

  • is in the UK and has stayed, or intends to stay, in the UK for 12 months or more
  • has a permanent UK address and is outside the UK and intends to be outside the UK for less than 12 months

The statistics show usual residents at the location they considered to be their usual place of residence on Sunday 21 March 2021. A UK resident’s usual place of residence is generally their permanent or family home or the address in the UK at which they spend most of their time.

We counted students at their term-time address, with some details also collected at their home address. We advised students that a term-time address was the address that they intended to stay at regularly during term time in the 2021/22 academic year, even if they were not there on Census Day. Further information is provided in “Section 6: Taking a census during a period of change”, in the “Place of residence” subsection.

Alternative population bases

Alternative population bases allocate those counted in the census to different geographical locations where individuals could be counted. The data on alternative population bases that we made available following the 2011 Census are accessible on the Alternative population statistics pages of the ONS website.

Alternative population bases are those that:

  • move usual residents to different locations based on their characteristics, and/or
  • report on a population that is not usually resident

We discuss each alternative population base in more detail in the following paragraphs of this consultation document. There’s the potential for disclosure to be higher for alternative population bases. As a result, we need to carefully assess what additional information we can publish and for what geographical levels.


The workplace population is an estimate of the population working in an area. It includes usual residents aged 16 to 74 years whose usual place of work is in the area. We include people who work mainly at or from home, or do not have a fixed place of work, in their area of usual residence.

The workplace population of an area excludes anyone in one of the following population groups. These include:

  • those living in England and Wales but working in Scotland, Northern Ireland, outside the UK or on offshore installations
  • those with a place of work in England and Wales but who are not usually resident in England and Wales
  • short-term residents

The definition for the workplace population remains unchanged. However, we’re anticipating that because of the coronavirus pandemic more people will be working from home, on furlough or otherwise not working. We will still count people at their workplace address who state that they’re “temporarily away from work”. There may also have been an increased number of jobs without fixed workplaces. One example is delivery jobs that are also counted at the respondent’s home address.

We’re proposing to produce outputs on the workplace population at Middle Layer Super Output Area (MSOA) level and above. These will include some information on basic characteristics of the population.


The workday population is an estimate of the population during the working day. It includes everybody who works in an area, wherever they usually live, and all respondents who live in the area but do not work.

The workplace population of an area excludes anyone in one of the following population groups. These include:

  • those living in England and Wales but working in Scotland, Northern Ireland, outside the UK or on offshore installations
  • those with a place of work in England and Wales but who are not usually resident in England and Wales
  • short-term residents

Workday population estimates are also a geographic redistribution of the part of the usual resident population who are in work, allocated to their place of work. The data include the workplace estimates plus those usually resident in the area who are not working.

The definition for the workday population remains unchanged from 2011. It’s defined as the non-working and homeworking usual residents in the area, plus those working in the area but resident elsewhere. The coronavirus pandemic may have changed this distribution because of larger numbers of people staying at home to work, being on furlough or not working.

We’re proposing Output Area (OA) level outputs. These will include some information on basic characteristics, creating a similar series of estimates to 2011.

Out-of-term population

This is the usually resident population, redistributed to their out-of-term address if they have one. As a result, the difference relates to the location of some students and schoolchildren.

We will define this population in almost the same way as in 2011. The only slight difference will be because of a change in how we count flexi-boarders. For 2021, only schoolchildren at boarding school who stay at their term-time address for four or more nights per week in general should be recorded as usually resident at both their term-time address and their home address. Those who stay for fewer than four nights a week should be recorded as usually resident at their home address only.

Because of the pandemic, there may be more students without term-time addresses and fewer students entering or leaving England and Wales to study. This will potentially lead to less difference between the usual resident and out-of-term populations than in other years.

We’re proposing OA-level outputs. These will include some information on basic characteristics, creating a similar series of estimates to 2011.

Short-term population

Outputs for short-term residents provide characteristics of people who were not born in the UK and who intend to stay in the UK for 3 to 12 months. Because of the restrictions on travel that the coronavirus pandemic caused, this population may be much smaller than in 2011.

As a result of the anticipated small counts, we expect to provide users with a series of regional short-term resident tables for 2021, with accompanying analysis. These will include some information on basic characteristics.

Second address

Second-address estimates provide information on the number of people who have a second address outside of the local authority where they are usually resident.

This population is redistributed to addresses that respondents recorded as their second address. This is an address they spent more than 30 days a year at, including:

  • holiday homes
  • partners’ addresses
  • armed forces bases
  • students’ out-of-term-time addresses

Because of the possible small counts, we’re proposing mostly local authority outputs for 2021. These will provide some basic characteristics. To ensure we protect confidentiality, it may be necessary to use higher geographies.

We’d like to know what data tables and analysis you need for alternative population bases.

Small populations

Following the 2011 Census, we offered small population tables. These tables provided information on the key characteristics of people in specific small population groups where numbers may not have supported multivariate analysis at smaller geographies. These groups were defined by their cultural background. Examples of these characteristics were individuals of:

  • an ethnic group
  • a country of birth
  • a religion
  • a national identity

These small population tables reflected identities about which users have specific needs. This may have been to better understand that small population or to analyse potential inequalities between that small population and the wider population. We published the small population tables on the Nomis Small population – 2011 Census page.

We only produced these small population data for geographic areas in which the small population being counted exceeded a set threshold. We only included the areas in which the population exceeded those thresholds in each table.

At a minimum, we aim to produce tables for the six small populations that we produced tables for in 2011. We will make these available at the MSOA level. In 2011, we produced 17 tables for each of the six populations. The populations in 2011 were:

  • National identity - Cornish
  • Ethnic group - Kashmiri
  • Ethnic group - Nepalese
  • Ethnic group - Sikh
  • Religion - Ravidassia
  • Religion - Jain

For the Sikh and Jain tables in Census 2021, the small population outputs will use new derived variables. The variables include responses in either the ethnic group or the religion question.

We will also produce local-authority-level tables by five-year age bands by sex, for a total of 30 population groups for local authorities. We will do this for the same 30 population groups as in 2011. These were all ethnic groups or countries of birth.

For ethnic group, these were:

  • Afghan
  • Filipino
  • Greek
  • Greek Cypriot
  • Kurdish
  • Latin Central South American
  • Nepalese
  • Polish
  • Somali
  • Sri Lankan
  • Tamil
  • Turkish
  • Turkish Cypriot

For country of birth, these were:

  • Bangladesh
  • Bulgaria
  • Cyprus EU
  • France
  • Ghana
  • India
  • Ireland
  • Jamaica
  • Nigeria
  • Pakistan
  • Philippines
  • Poland
  • Romania
  • Somalia
  • Sri Lanka
  • South Africa
  • Turkey

For Census 2021 outputs, we aim to make more information available on small populations. This includes consideration of increasing the volume of tables produced on small populations in three ways.

Firstly, we could provide data on additional small populations defined based on their ethnic group, religion, country of birth or national identity. The data could use responses from multiple questions to determine these populations. The data could also include small populations that are now possible because of changes to the questions, such as “Roma” or ethnic groups within the “African” subgroup.

Secondly, we could provide further detail on small populations. This detail could be variables such as different cross-tabulations or geographical breakdowns.

Thirdly, we could provide data on groups defined using other characteristics. An example of such a characteristic could be, for instance, the population who have British Sign Language (BSL) as a main language.

However, because of potential for disclosure being higher for small populations, we need to carefully assess what additional information we can publish.

We’d like to know what data tables and analysis you need in relation to existing and potential new small populations.