Executive summary and overview of Census 2021: General Report for England and Wales

The census of population and housing has been at the heart of population statistics for England and Wales for over two centuries. It has been the biggest statistical exercise the Office for National Statistics (ONS) undertakes, involving contacting every household and communal establishment every 10 years. The resulting statistics from each census have been the most detailed available on the population, its characteristics, and where and how we all live. These statistics are widely used by government, local authorities, charities, private organisations, and the public.

Census 2021 was the first to be primarily completed online. It achieved a high response rate, with 97% of all households completing the census, and all local authority areas reached response rates of at least 88%. It also saw a timelier release of data than ever before, with the first results released a year after the data-collection operation and all major England and Wales releases published by the end of 2023. Our Census 2021: General Report for England and Wales describes the preparation and delivery of Census 2021, from the decision to hold a online-first census through to the release of those outputs in 2022 and 2023.

Any census is an ambitious and complex undertaking. In addition, the preparation and delivery of Census 2021 was affected by the coronavirus (COVID-19) pandemic. The pandemic and lockdown restrictions meant that some people temporarily lived elsewhere, while other people's work and behaviours changed, including increased use of the internet. This presented both challenges and opportunities for the collection of census data.

Alongside Census 2021, the ONS has led a programme of research to improve its population and migration statistics by making the most of the potential of administrative data (information collected by other public bodies) to produce up-to-date statistics about the population more often and more quickly than the census.

Important features and innovations of Census 2021

Online-first approach

Most people completed the census online (including 89% of households), which provided opportunities for features such as automatic quality checks and ensuring that people were shown only the questions that were relevant to them. 

Innovative use of technology

New and improved use of technology featured throughout the census, including the electronic questionnaire, timely data on responses to inform decisions on follow-up activity, and increased computing capacity for data processing.

Inclusivity, accessibility, and engagement

As well as providing a wide range of support online, by phone and in person, a major part of our communications and engagement focussed on overcoming barriers to taking part. The addition of new questions provided new data to address gaps in our understanding of society.

Use of administrative and alternative data sources

Throughout Census 2021, we made greater use than ever before of administrative data and other data sources. This included:

• in the data-collection operation, such as for the address frame (listing the residential address for initial contact and response tracking) and planning resources for those who needed help

• in our processing and quality assurance

• in the production of integrated outputs that used census data and data from administrative sources

Flexible and accessible outputs

People could access Census 2021 data in a range of ways, including through our innovative Create a custom dataset tool as well as ready-made datasets, interactive digital products, analysis articles, and commissioned tables. This range of outputs enabled users to explore census data more flexibly than ever before.

A three-year period of question development established what data users needed from the census and how we would collect them. This began with our Topic Consultation in 2015 and led to the recommended census questions being set out in the 2018 census white paper. This work and our wider preparations for Census 2021 are outlined in Chapter 2 of our Census 2021: General Report for England and Wales. We also consulted on census outputs in 2018 and 2021, and these consultations helped to shape the suite of products described in Chapter 9.

Two new voluntary questions required a new Act of Parliament, which gained Royal Assent in 2019. This was followed by the census secondary legislation: an England and Wales Census Order and separate regulations for England and for Wales, which confirmed the questions and other details of the census.

While it was digital-first by design, Census 2021 was also multi-mode, with both online and paper responses. Most households received initial contact letters with online access codes, but our paper-first strategy identified 10% of areas to receive paper questionnaires, which also included a code to complete online, to aid responses. Our question and questionnaire development optimised the design for both online and paper completion.

Our commitment to delivering a bilingual census in Wales meant that the census website, our materials and services were available in English and Welsh. Users could toggle between English and Welsh versions of each webpage, including within the questionnaire for respondents in Wales. In Wales, initial contact letters sent to addresses were bilingual and paper-first addresses received questionnaires in English and Welsh. A Welsh-language option was also offered by our freephone contact centre, and the field force included Welsh speakers.

In early March 2021, we sent out 24 million letters and 3 million paper questionnaires to households, ahead of Census Day on 21 March. We exceeded our target of responses received without follow-up activity, with over 70% of households responding by the end of Census Day and nearly 79% by 26 March. As well as the quality and reach of our communications and the ease of completion both online and on paper, the coronavirus (COVID-19) pandemic lockdown conditions may also have helped this early success.

Much work was needed to ensure that we reached our target of 94% of households taking part. The pandemic made this follow-up activity more difficult, but we exceeded both our response rate and variability targets. Chapter 4 describes the data-collection operation, including the work of the field force, the electronic questionnaire, and the support provided to respondents. Chapter 5 describes the communications, media, and engagement work that was essential to delivering a successful operation with high response rates.

Major components of the data-collection campaign

Support and overcoming barriers to taking part

We provided guidance and offered practical support to help people to complete the census. The public could access a range of online and in-person support, including a freephone public contact centre that provided guidance and help by telephone, and digital services such as web chat and social media. Our contact centre received almost 4.5 million calls, and over 180,000 census responses were completed through the telephone capture service. Information about the census was available in 49 languages through the census website, the contact centre, and local engagement staff, and in a range of formats such as braille and large print. We also provided in-person support for online completions through census support centres, though this service was affected by the coronavirus (COVID-19) pandemic and lockdown restrictions.

We recognised that some people may be less likely to take part without additional interactions taken to publicise the census, underline its use and value, and offer support. Communications and advertising were produced in different formats and ran across numerous channels. Our network of engagement staff worked with communities at regional and local levels. They helped people to overcome barriers to completion and offered more tailored communications and support.

We worked closely with local authorities and community partners well before March 2021, including deploying local engagement staff from autumn 2020. Building trust and advocacy from community leaders and representatives meant they could help to build awareness, trust, and willingness to complete the census within those communities. Our local engagement staff worked with population groups less likely to respond to the census without additional interactions, for example through community events and local media. These staff included 200 Census Engagement Managers working with local authorities and communities across England and Wales, and 115 Community Advisers working with specific population groups.

At our request, almost every local authority nominated a member of staff to take on the role of Census Liaison Manager or Assistant Manager. The support of local authorities, with their knowledge and understanding of their local areas and communities, was very important. These partners' contributions to coordinating activity across their councils and their commitment to achieving a successful local census was essential.

Completion of the census by students was a particular challenge; this was made harder by the pandemic, as many students lived away from their normal term-time addresses. Our student communication campaign focused on the message that they had to be included at their term-time address, if they had one, and at their out of term-time address, such as a family home.

Each of the individual measures we took helped to ensure that as many people as possible could engage with the census and access the information that they needed to take part.

Communications and public relations

Our communications campaign was integrated with the phases of contact with households. Our campaign began with preparing audiences for the upcoming contact, and moved on to encouraging completion before and after Census Day, which included targeted advertising for areas and populations with lower response rates. The campaign used a range of communications channels to reach a large proportion of the population. This ranged from TV and radio to social media and digital, including social media influencers and paid media partnerships. The communications campaign also focussed on messaging for groups facing barriers to taking part. For these targeted communications, we used different languages and bespoke photography representing different population groups, while keeping our "it's about us" message and a common style to the whole campaign.

The public relations campaign included promotional activities such as "community heroes", who received 1 of 22 purple plaques; a purple-light switch-on that lit up over 200 buildings across the country; and extensive media activity resulting in national coverage and over 430 pieces of coverage across community-specific media outlets. National coverage included appearances by ONS staff and spokespeople on every major news channel, an EastEnders storyline, and a partnership with Gogglebox. We made use of social influencers to reach both mass and niche audiences. Census communications and engagement were also supported by successful programmes for primary and secondary schools.

Field force operations

Our household field force of 20,000 people made over 16 million visits to over 5 million households during the five weeks after Census Day, alongside us sending more than 11 million reminder letters. During these visits - 4.5 million of which resulted in contact with the household - census officers were able to offer advice on the doorstep and direct householders to support channels, to enable householders to complete their census.

Flexibility in the design of the field operation was a vital part of its success. This included the use of mobile teams in the field force and the ability to redeploy staff in a flexible way. Technology played an important role, with up-to-date data on responses and information from field staff used to make timely decisions about where to target resources to optimise response.

A further 1,300 field staff supported communal establishments and other "special population groups". Communal establishments are generally sites that provide managed residential accommodation, such as care homes and student halls of residence. Other "special population groups" were largely people who lived in accommodation types that were more difficult to access and enumerate, such as people living in armed forces family accommodation, marinas, or traveller communities. As expected, the final response rate was lower for communal establishments, at 79%. The return rate for "special population groups" was also lower, at 76.4%.

After the main campaign, a further field operation of nearly 4,000 staff conducted the Census Coverage Survey, a vital part of producing accurate estimates for Census 2021.

The coronavirus pandemic affected these operations in different ways, and the safety of the public and census staff was a major consideration in the operations.

Following the successful data-collection campaign, we completed a rigorous series of steps to produce accurate estimates of the whole population of England and Wales. There were four main stages of this work.

Pre-processing

The first stage involved cleaning and coding responses, resolving duplicates, and removing false responses. This resulted in a dataset of 58.6 million residents and 26.3 million households.

Edit and imputation

The edit and imputation stages produced a fully populated, clean, and consistent census database by resolving missing values and inconsistencies within or between persons and within households.

Estimation and adjustment

In the estimation and adjustment stage, we produced accurate estimates of the whole population. We did this by taking account of both undercoverage and overcoverage (using the Census Coverage Survey) and adjusting the census dataset to represent the estimated population of 59.6 million usual residents.

Quality assurance

We carried out extensive quality assurance to assess the quality of the previous stages and compared the census results with other available information. Population estimates were quality assured at national, local authority, and sub-local authority levels, including inviting all local authorities to contribute to our quality-assurance process. We also carried out detailed assurance of the census results for each topic. This included comparisons with 2011 subnational patterns and with other available data, and consideration of other issues raised by users or in the quality-assurance process.

Chapter 6 describes each stage in more detail. Processing the data from Census 2021 benefitted from the digital-first approach. This was because collected data were quickly available to test, update, and quality assure the processing stages. Processing also benefitted from a substantial increase in computation processing power since 2011. Estimation of large communal establishments also made use of administrative data and bespoke surveys carried out after the census.

We estimated the usual resident population in England and Wales on Census Day to be 59,597,542, of whom 56,490,048 lived in England and 3,107,494 in Wales.

As for previous censuses, we had a range of measures and targets to assess the quality of those estimates. The main strategic aim was to achieve at least a 94% overall response rate and to minimise local variation, with all local authorities having a response rate above 80%. We exceeded both, achieving a 97% overall response rate and over 88% in all local authority areas with more than half above 97%.

Chapter 7 details other measures of the quality of the census population estimates. These include:

• response rates by ethnic group, all of which were higher or the same as in 2011, with less variation between groups because each had a response rate of at least 92%

• question completeness, which was generally higher than in 2011

• confidence intervals, which are a statistical measure of uncertainty

Our 95% confidence interval had a lower limit of 0.16% below the England and Wales population estimate and an upper limit of 0.07% above. This was well within our target of plus or minus 0.2%, and a narrower overall width of the interval than in 2011.

Chapter 7 also outlines some issues that were specific to Census 2021. The coronavirus (COVID-19) pandemic and lockdown restrictions had some specific impacts. These were most notably on the place of usual residence for some people, and on labour market and travel to work data. The chapter also describes issues relating to the new questions and our joint workplan with the Welsh Government on the coherence of Welsh language statistics.

Our strategic aims for Census 2021 outputs were flexibility, timeliness, accessibility, and relevance. The first release of Census 2021 data on 28 June 2022 provided rounded population estimates of usual residents, along with estimates by sex and five-year age bands, down to local authority area level. We published more detailed releases over the remainder of 2022 and 2023, supported by data visualisations, interactive content, and information on quality and methods. We released our innovative "Create a custom dataset" tool in March 2023, enabling users to find and download multiple combinations of data they were most interested in. This was followed by specialist data and additional geographies, including origin-destination flows, alternative population bases, and bespoke datasets containing data about specific small population groups (defined by ethnic group, religion, national identity, main language, or country of birth). With the first results released a year after the data-collection operation, and the main England and Wales census data published by the end of 2023, this was a timelier release of census data than ever before. Chapter 9 describes the full range of Census 2021 outputs.

Our suite of outputs supported users with a range of experience and needs. These included expert users who wanted machine-readable datasets and APIs (application programming interfaces), and general users who used our interactive content, data visualisation, storytelling, and games to bring the data to life. We also produced a suite of analysis articles using multivariate data to explore different topics and populations.

The safety of census respondents' personal data is a top priority for the Office for National Statistics (ONS). All our systems, staff, and suppliers must protect the confidentiality of census data by law, including the:

• Data Protection Act 2018

• UK General Data Protection Regulation (GDPR)

• Census Act 1920

• Statistics and Registration Service Act 2007

Chapter 8 sets out how security and confidentiality were built into Census 2021 storage, processing, and publication. This included our approach to statistical disclosure control to protect personal data in our published datasets. An Independent Information Assurance Review provided independent assurance.

Our Census and Data Collection Transformation Programme delivered both Census 2021 and evidence to enable the UK Statistics Authority's recommendation on the future of population and migration statistics. The census cost £539 million as part of this £820 million programme. More detail on the costs can be found in Chapter 3.

As in previous censuses, external suppliers provided several of our services, including:

• printing and posting census materials

• the communications campaign

• questionnaire management

• the public contact centre

• field force people services

More information on these and other contracts is available in Chapter 3 and on our Census 2021 suppliers webpage.

Ahead of the census, the total benefits to central government, local government, and private sector users were expected to be £5.5 billion over a 10-year period. Our assessment of Census 2021 benefits is due to be published in summer 2025.

Chapter 10 reflects on the experience of planning and delivering Census 2021 and reflects on both positive lessons and things that could have been improved. It identifies the main factors that made Census 2021 successful, including some of the features and innovations described earlier in this Executive Summary. These factors include:

• a good statistical design for the census and a strong, but flexible, operational design to deliver it

• very high response rates by the end of the census weekend, reflecting the quality of our communications campaign, engagement, and media relations, and well-designed routes to completion

• exceeding our household response and local variability targets, thanks to effective follow-up activities and communications, and local, national, and community engagement

• improvements in our processing stages, assisted by early access to live census data and greater processing power

• improvements in our estimation and quality assurance, including the Local Authority Insight Initiative

• high-quality outputs that were more timely and more flexible than ever before, providing data and resources to meet a wide range of user needs

• a legacy beyond the census itself, with many of the technologies, methods, and lessons used across our wider work

Chapter 10 addresses each stage of Census 2021, from planning through data-collection and processing to outputs dissemination, highlighting lessons learned. While some are specific to the delivery of a census, others may be applicable to our wider work, including producing population estimates based primarily on administrative data and other surveys.

We ran Census 2021 as part of a wider programme focused on producing essential, up-to-date statistics about the population and migration. This work follows our 2014 commitment to deliver an online-first census in 2021 and to use administrative data and surveys to improve annual statistics, reducing reliance on a census every 10 years. We are using a wider range of data sources - including administrative data, such as data from tax, benefits, health, border flows, and education systems - to produce up-to-date statistics about the population and migration more often.

Our headline international migration estimates have changed from using survey-based data, where we would ask people if they were intending to move to or from the UK, to primarily using administrative data. These new data and methods mean we are using a wider range of data to provide improved estimates that capture long-term international migration based on actual behaviours, rather than stated intentions.

Chapter 11 summarises the programme's work beyond the census. This includes the UK Statistics Authority's June 2025 recommendation to government that the Office for National Statistics (ONS) should be commissioned to deliver a questionnaire-based census of the whole population in 2031, alongside the ongoing development. It also includes the ongoing development of an admin-based population and migration statistics system, supported by improvements in the supply and quality of administrative data. This will enable us to combine the strengths of both sources of data.

Delivering a census in 2031 is an exciting opportunity to build a system of population and migration statistics that combines the strengths of whole-population data collection with the additional value that administrative data can provide. This combination would create a resilient and future-focussed population statistics system, which will help to provide users with the detail they require, while also providing an improved understanding of how society is changing in the years between censuses.