1. Executive summary

What is a census and why is it important?

Every 10 years since 1801 (except in 1941 when no census was taken due to the World War 2) the nation has set aside one day for the census – a count of all people and households. It is the most complete source of information about the population that we have.

Every effort is made to include everyone, and that is why the census is so important. It is the only survey that provides a detailed picture of the entire population and is unique because it covers everyone at the same time and asks the same core questions everywhere. This makes it easy to compare different parts of the country.

The information the census provides allows central and local government, health authorities and many other organisations to target their resources more effectively and to plan housing, education, health and transport services.

Who will carry out the 2021 Census?

The Office for National Statistics (ONS) will carry out the next census in England and Wales in March 2021. Our approach to preparing for the 2021 Census has been guided by the lessons learned from the previous census in 2011.

The 2011 Census for England and Wales was conducted successfully in line with international best practice and provided a sound basis for understanding the nature and diversity of the population. We carried out an internal evaluation of the 2011 Census, the results of which were published in 2011 Census General Report for England and Wales. The report noted there were some important successes and a number of challenges and lessons to be learned for any census operation in the future.

What are the ONS’s aims for the 2021 Census?

Our aims for the 2021 Census are:

  • to produce high-quality results

  • to generate outputs that meet the needs of users

  • to maximise online response rates for the census

  • to produce timely outputs to maximise benefits

  • to protect, and be seen to protect, confidential information

  • to conduct it in a cost-effective way

  • to make recommendations for the approach to future censuses in England and Wales

Who monitors the running of a census?

The Office for Statistics Regulation (OSR) is the regulatory arm of the UK Statistics Authority. It assesses official statistics for compliance with the Code of Practice for Statistics, and reports on system-wide issues and on how statistics are used, celebrating when the standards are upheld and challenging publicly when they are not.

The conduct of the entire 2021 Census operation will be scrutinised by the OSR against the Code of Practice for Statistics, which comprises elements of trust, quality and value (known as pillars). This assessment will determine whether the outputs produced from the 2021 Census will retain the “National Statistics” status we achieved for the 2011 Census. If we are successful, our accreditation will be confirmed before the first census outputs are released in 2022.

By complying with the Code, it gives users confidence that our statistics are of public value, are high quality and are produced by people and organisations that are worthy of trust. Read more about the Code of Practice for Statistics.

How will the ONS achieve its aims for the 2021 Census?

We are building on the success and valuable experience gained from the 2011 Census and developments and lessons learned in international census taking. A comprehensive programme of research and testing has underpinned our proposals for the 2021 Census, including our major census test in 2017 and upcoming rehearsal in October 2019.

The White Paper outlining the recommendations for the content and conduct of the 2021 Census for England and Wales was published in December 2018. It sets out:

  • the proposal that, for the first time, the census will be predominantly online

  • the proposed date for the 2021 Census in England and Wales

  • the topics recommended for inclusion in the 2021 Census and the consultation process that informed these recommendations

  • confidentiality provisions for personal data collection and digital security

  • plans for the publication of, and access to, 2021 Census data

The report that follows addresses in detail our progress so far in designing and preparing for the 2021 Census against the three pillars in the Code of Practice for Statistics. We have submitted it to the OSR to be used as evidence, as part of the accreditation process. The main areas within the report are as follows.

Collection is being designed to make it easy and clear for respondents

The 2021 Census for England and Wales will be predominantly online for the first time. In 2021, we have a target to achieve a 75% online return rate. We will provide a wide range of assistance to the public to encourage online completion, including a dedicated census contact centre, work with community groups, and work by census field staff on the doorstep.

We recognise, though, that some people may struggle with this. Therefore, we are proposing comprehensive arrangements to make sure that everyone can complete their form, whether online, on paper or even by phone.

Our aim is to deliver better quality, to make it easier for people to fill in the form and all the while being more efficient.

For more information on data collection, please see Data collection and Improving the quality of the census in Section 5, and Developments in technology and census data, 2017 Test and Improvements for the 2021 Census in Section 6.

Keeping personal data secure from collection through to publishing results is fundamental

Protecting the confidentiality of every individual’s personal information is and will continue to be of paramount importance to us. Every person’s identity will be protected, not only through secure handling and storage of their data, but also by ensuring that our statistical publications do not identify individuals.

Census records are released to the public 100 years after the census took place. They provide a fantastic source of information to find out more about our ancestors.

For more information on how we are working to protect people’s privacy, please see Confidentiality principles in Section 4 and Developing output products to meet user needs in Section 5.

Outputs will meet user needs and be timely to maximise benefits

We have worked and consulted extensively with community groups, charities, governmental bodies and businesses in preparation for this census, to make sure the statistics it produces are of benefit to everyone. This census will not only shed light on long-term trends, but will also reflect the society we live in today, with a mixture of traditional and new questions.

For more information on how we are engaging with users to provide relevant data, please see Understanding the needs of census data users and Development of proposed new questions in Section 5, and Public consultations, roadshows and meetings and Using the right data sources in Section 6.

It takes time to process the vast amounts of data that result from the census, but new approaches and investment should allow quicker results than in the past. The aim is to release all the initial findings within 12 months, and all data within 24 months. Access to census results should be improved, with users able to design alternative tables to supplement some published standard tables.

For more information on how we are working to provide more timely and accessible data, please see 4.3 Principle T3: Orderly release in Section 4, Developing output products to meet user needs and 2021 Census results and findings in Section 5 and Public consultations, roadshows and meetings and Accessible outputs in Section 6.

Assurance through internal and external sources

In addition to the OSR assessing the conduct of the census, we are subject to other forms of assessment and review. The work we do has to adhere to an assurance strategy, which will demonstrate to stakeholders that the Census and Data Collection Transformation Programme (CDCTP) will achieve its scope, time, cost and quality objectives to realise the benefits of the 2021 Census.

We involve experts from inside and outside the ONS to provide assurance that methods and processes will produce robust results.

For more information on how we are assuring the census, please see Internal and external review in Section 4, and Developing sound methods and Working with other national statistics institutes in Section 5.

Back to table of contents

2. Introduction

At the Office for National Statistics (ONS), we are responsible for the planning, undertaking and dissemination of outputs from the 2021 Census in England and Wales.

The conduct of the entire 2021 Census operation will be scrutinised by the Office for Statistics Regulation (OSR), the regulatory arm of the UK Statistics Authority, against the Code of Practice for Statistics. This report forms part of the OSR assessment, that will determine whether the outputs produced from the 2021 Census can be designated as “National Statistics”.

The OSR will also consider evidence from a variety of sources including their own research, information given to them by us and importantly feedback from users and stakeholders connected with the census. Capturing the views of users and potential users forms an important part of the OSR’s judgement about the statistics, including how they could be improved. If you would like to provide feedback or speak to the OSR, please see detail of the assessment on its website.

The Code of Practice for Statistics provides producers of official statistics with the detailed practices they must commit to when producing and releasing official statistics. When producers of official statistics comply with the code, users of statistics, and citizens, should have confidence the statistics are “of public value, are high quality and are produced by people and organisations that are worthy of trust”.

The 2011 Census outputs were designated as National Statistics. We are aiming to retain the National Statistics designation for 2021 Census outputs.

Assessment of the 2021 Census

The OSR will evaluate the conduct of the 2021 Census from the initial planning stages that started in 2014, through to the production of our outputs.

The assessment is split into three phases during which we must demonstrate how we are being trustworthy, of high quality and of value to users. The OSR will also review other evidence alongside the report and users’ views.

Each phase of the assessment occurs as follows:

  1. the ONS compiles and submits the report
  2. the ONS publishes the report on the website
  3. the OSR assesses the report, speaks to census teams and census data users, and publishes requirements
  4. the ONS acts on the requirements

This process is repeated for all three phases of the assessment.

At the end of the phase 2 assessment period, the OSR will consider whether the 2021 Census for England and Wales has sufficiently complied with the Code of Practice for Statistics to retain the badge of National Statistics Accreditation. If we are successful, our accreditation will be confirmed before the first census outputs are released in 2022.

Undertaking the National Statistics Accreditation and laying out our processes, procedures and practices for assessment by the OSR allows us to continually improve and ensures census outputs serve the public, and meet the highest levels of trustworthiness, quality and value. It also ensures that users of our statistics, and citizens, can have complete confidence in our 2021 Census.

Code of Practice for Statistics framework

The framework for this report is based upon the framework of the Code of Practice for Statistics. The Code is based on three pillars. These are:

  • trustworthiness – having confidence in the people and organisations that produce statistics and data

  • quality – using data and methods that produce assured statistics

  • value – producing statistics that support society’s needs for information

Each pillar contains a number of principles and detailed practices. The report addresses each of these principles and provides evidence demonstrating how we have met the Code during phase 1 and addressed the recommendations from the 2011 Census. Table 1 shows the pillars and principles of the Code.

What’s included in this report?

Section 3 of this report outlines the recommendations made by the UK Statistics Authority in the Special Assessment of the 2011 Censuses in the UK (PDF, 1.2MB) and provides evidence to demonstrate where we have addressed these recommendations.

In the subsequent sections, we explain how practices, processes and procedures of the 2021 Census for England and Wales have met, and will continue to meet, the pillars of trustworthiness (Section 4), quality (Section 5) and value (Section 6). Section 7 explains what will happen next following the submission of this report. Sections 8 to 14 are appendices.

2021 Census in Scotland and Northern Ireland

The National Records of Scotland (NRS) and Northern Ireland Statistics and Research Agency (NISRA) are responsible for conducting the 2021 Census in Scotland and Northern Ireland respectively. NRS and NISRA are also responsible for providing evidence to demonstrate how their census undertaking is meeting the Code of Practice for Statistics.

Back to table of contents

3. Response to recommendations from the UK Statistics Authority’s Special Assessment of the 2011 Censuses in the UK: Phase 3 report

Phase 1 of the National Statistics Accreditation requires us to respond to recommendations made by the UK Statistics Authority in the Special Assessment of the 2011 Censuses in the UK: Phase 3 report (PDF, 1.2MB). This report reviewed our statistical practices in three main areas: user engagement, dissemination and accessibility of the data and statistics. The main outcomes of the review, informed by feedback about the experiences of both users and the statistical producers, as well as from the published documentation and census outputs, concerned:

  • the effectiveness of engagement with users

  • collaboration and partnership

  • accessibility and dissemination

  • coherence across the UK

Recommendations

The Authority provided five recommendations (PDF, 1.2MB) to the Office for National Statistics (ONS).

Recommendation 1

Shares their experiences of effective engagement gained on the 2011 Census with other statistical teams.

Recommendation 2

Further develop partnerships with key users and census data distributors to:

a) facilitate the wider use of the data and statistics by other users, such as through collaboration with, and support provided to, voluntary sector groups, local government, census data resellers and data application developers

b) consider secondments to develop knowledge and a common understanding between the statistical producer, data suppliers and user organisations

c) identify opportunities for data suppliers and users to contribute to the quality assurance of official statistics and the development of outputs

Recommendation 3

Discuss with users the potential for:

a) enabling the quicker release of data, including for small areas such as the travel to work data

b) widening access to anonymised samples such as by users in the commercial and local government sectors

c) greater flexibility in creating univariate and multivariate tables for bespoke geographies and population groups

Recommendation 4

We (the Authority) suggest that each statistical producer implement ways of informing users of substantive revisions alongside the statistics in the context of the developments in online dissemination.

Recommendation 5

Identify (together with Welsh Government):

a) the areas of success and how to build on them

b) the areas of practice where alternative approaches would strengthen the delivery of the census in each country, as well as in delivering comparable and consistent UK census data and statistics

c) the needs of users in relation to cross-UK census data and how well these were met by the 2011 Census

The remainder of Section 3 sets out how we have addressed the Authority’s recommendations in the planning of the 2021 Census for England and Wales. The full details are available in the Special Assessment of the 2011 Censuses in the UK: Phase 3 (PDF, 1.2MB) report.

Addressing the Authority’s recommendations

Addressing Recommendation 1

Following the release of the 2011 Census results, the ONS carried out a significant amount of engagement to share information on lessons learnt around stakeholder engagement.

An assessment was provided in the general report for England and Wales.

A dedicated webpage was set up to share information on benefits, including a range of case studies that were prepared in association with data users to demonstrate the benefits.

As well as publishing information, we carried out extensive face-to-face engagement with other government departments, the third sector, businesses and other organisations.

Addressing Recommendation 2a

We carried out a lot of work following the 2011 Census to derive benefit from the data. As well as those mentioned previously, there was an end user satisfaction survey and microdata teaching file, and a secondary schools programme with modules on using census data.

Building on what went well for the 2011 Census, lessons learnt detailed in the general report are being used to develop wider engagement for the 2021 Census. They can also be used by other statistical teams (for example, social survey) to ensure a coherent relationship with stakeholders that continues beyond the census. This includes building wider relationships with communities, other organisations and groups, as well as maintaining relationships with our traditional stakeholder base. Through these partnerships, we aim to gain assistance in reaching specific groups to improve response, minimise variability in response and understand what data analysis is required from the census results to meet their needs.

This direct engagement approach is aimed at improving the quality of the data, building trust with a wider range of users and improving the value of the outputs.

We are maintaining the usual engagement methods such as, consultations, Census Advisory Groups (CAGs), Working Groups (for example, Outputs and Dissemination User Group (Appendix 1), Origin-Destination (Appendix 2), Microdata (Appendix 3) and Data Distributors Working Group (Appendix 4)) and newsletters.

Addressing Recommendation 2b

We used the lessons learnt (PDF, 498KB) from the 2011 Census to inform our design of the 2021 Census. We work with a number of user groups on an ongoing basis to help develop our innovative dissemination approach.

We acknowledge we could benefit from a deeper understanding of the needs of census data distributors in light of our new dissemination approach and secondments could be an option.

Further guidance from the Authority on meeting this recommendation would be welcomed.

Addressing Recommendation 2c

Frequent and open engagement with users was a major contributory factor in improving the quality assurance (QA) methods (PDF, 1.1MB) and building user trust in the results (details in Section 5). Engaging with users on the 2021 Census continues to be a critical exercise. This is being achieved in several ways including:

Addressing Recommendation 3a

We have set up the Origin-Destination Working Group (Appendix 2) and held a public consultation to gather user views on these data.

We have used the lessons learnt from the 2011 Census to inform the design of the 2021 Census outputs. This includes developing the statistical disclosure control methodology, which we hope will enable us to disseminate data quicker than in 2011. We aim to release all the initial findings within 12 months, and all data within 24 months.

The primarily online-first design and new technologies will improve the efficiency and timeliness of the data collection and processing operations and therefore dissemination of results.

Addressing Recommendation 3b

We are committed to making data available as widely as possible and are responsible for maximising benefits from the 2011 Census. As part of this, we released several different microdata products alongside samples from the 2001 and 1991 censuses. The products strike a balance between detail and security, allowing a flexible toolset for casual users as well as professional researchers.

In 2017, additional microdata samples from the 1981, 1971 and 1961 censuses were made available as part of a project to make better use of existing data and to advance research based on census data.

Users can access the census microdata samples through our Secure Research Service (SRS). This facility provides secure access to sensitive detailed data to Approved Researchers, including those in the commercial and local government sectors, for work on defined and approved projects that serve the public good.

Improvements to the SRS infrastructure have been made in recent years, which make it easier for users to access the data from additional locations. Further work is planned to allow greater connectivity and increase remote access. Users are kept informed of developments through a programme of events.

We consulted with users in the outputs consultation (PDF, 239KB) about access to microdata samples (see pages 42 to 43 of the consultation).

Addressing Recommendation 3c

We have consulted with users through the outputs consultation, roadshows, CAGs and Outputs and Dissemination Working Group to gather their needs for creating bespoke geographies and population groups.

Addressing Recommendation 4

In the event of a correction being required to census data after publication, we follow the revisions policy, which outlines the process for amending data and notifying users of the correction.

To keep users alerted to any revisions or updates, we produced an issues and corrections page on our website for the 2011 Census for England and Wales, to ensure all revisions and corrections were publicly available as soon as they occurred.

Where 2011 Census data were published through other channels, such as the Neighbourhood Statistics Service (NeSS) and Nomis, issues were reported separately on these websites.

Addressing Recommendation 5

In preparing for the 2021 Census, full account has been taken of the lessons learned from the 2011 Census through an internal evaluation titled 2011 Census General Report for England and Wales (PDF, 1.2MB).

In recognition of the importance of providing consistent outputs across the UK, the National Statistician and the Registrars General for Scotland and Northern Ireland have signed an agreement on the conduct of the census in 2021 (PDF, 165KB). This includes the production of harmonised statistics, with specific attention being given to:

  • common questions
  • population bases
  • definitions
  • disclosure control methodologies
  • output formats

Professor David Martin, of the University of Southampton and Deputy Director of the UK Data Service (UKDS), wrote to the UK Census Committee (UKCC) on the topic of UK harmonisation and attended the UKCC meeting in June 2017. The main points of the letter were that:

  • the 2011 Census made progress towards harmonised output but did not fully realise the vision

  • there was a lack of coherence in coding, metadata and supporting information from the three census offices

  • there was no single point for multivariate data for small areas

The user need for UK data was recognised and the UKCC will be a vehicle for setting direction and monitoring progress towards common databases, data models and metadata as required. All three census offices will work in collaboration with the commercial users and UKDS.

We are working with National Records of Scotland (NRS) and Northern Ireland Statistics Agency (NISRA) to to deliver harmonised census outputs for the UK where possible. There is a project looking at providing a single point of access for harmonised results for areas across Great Britain and the UK where the data collected are consistent across the census offices of the UK.

We have already started work with our colleagues in Scotland and Northern Ireland to enable the harmonisation of outputs and dissemination of UK data.

Back to table of contents

4. Trustworthiness

Confidence in the people and organisations that produce statistics and data

Trustworthiness means that statistics are produced using trusted processes, are free from any vested interest and that they represent the best professional judgement of statisticians acting in accordance with the highest standards.

4.1 Principle T1: Honesty and integrity

People in organisations that release statistics should be truthful, impartial and independent, and meet consistent standards of behaviour that reflect the wider public good. For more information, see Principle T1 in the Code of Practice for Statistics.

Our employees are expected to carry out their role with dedication and a commitment to the Civil Service and its core values set out in the Civil Service Code and the Better Statistics, Better Decisions strategy (PDF, 1.4MB). The core values are:

  • integrity – putting the obligations of public service above your own personal interests

  • honesty – being truthful and open

  • objectivity – basing your advice and decisions on rigorous analysis of the evidence

  • impartiality – acting solely according to the merits of the case and serving equally well governments of different political persuasions

We must adhere to several Office for National Statistics (ONS) policies such as General Conduct Policy, Conflicts of Interest Policy and our data and security policies. All staff who work on the census must sign a Census Confidentiality Agreement and Conflicts of Interest form.

4.2 Principle T2: Independent decision making and leadership

Organisations should assign a Chief Statistician or Head of Profession for Statistics who upholds and advocates the standards of the Code, strives to improve statistics and data for the public good, and challenges their inappropriate use. For more information, see Principle T2 in the Code of Practice for Statistics.

The Census and Data Collection Transformation Programme (CDCTP) fully understands the role of the Head of Profession (HoP) for Statistics in terms of the guidance and support they can provide in implementing the Code of Practice for Statistics. It is understood how to engage the Statistics HoP, Debra Prestwood, and how the Statistics HoP Office can be used to advise on judgements about methods, standards and data release, as well as ways to improve the statistics.

We meet quarterly with National Records of Scotland (NRS), the Northern Ireland Statistics and Research Agency (NISRA) and the Welsh Government (WG) at the UK Census Committee (UKCC) (Appendix 5) meetings. This enables each agency to keep abreast of plans and progress at a strategic level, to discuss ideas and risks, and to support work on harmonising censuses in the UK. These meetings are complemented by regular discussions at an operational level. The HoP from each devolved administration is represented and the meeting is chaired by the National Statistician.

4.3 Principle T3: Orderly release

Organisations should commit to releasing their statistics in an open and transparent manner that promotes public confidence. For more information, see Principle T3 in the Code of Practice for Statistics.

We outlined our plans for the dissemination of statistical outputs from the 2021 Census to the public as part of the 2021 Census outputs strategy consultation (PDF, 1.4MB), which ran from February 2018 to May 2018. We published a report on the findings (PDF, 83KB) together with recommendations in December 2018. Dissemination of outputs will follow a pre-arranged timetable, with the aim of releasing population and household estimates by March 2022 as part of the first phase of outputs, followed by all remaining pre-planned outputs by March 2023. This includes microdata and origin-destination data. The intention is for the outputs to be published within a shorter timescale compared with the 2011 Census, that is, 12 months rather than 16 months.

There will be a single point of access to 2021 Census outputs through a dedicated website that meets the needs of users. Furthermore, we are proposing the development of a flexible dissemination system, allowing users to create their own tables that have been appropriately protected from a statistical disclosure control perspective.

These results will be complemented by a range of other outputs which will be made available online as they are completed. An analytical work programme including topic-based reports and visualisations will support the release of the data.

The published statistics from the 2021 Census will include detailed metadata along with quality and methodology documentation. Contact information will also be provided that enables users to direct queries or seek further guidance from a relevant ONS statistician.

4.4 Principle T4: Transparent processes and management

Organisations should have effective business processes and appropriate resources to support their statistical functions and be open about their plans, priorities and progress. For more information, see Principle T4 in the Code of Practice for Statistics.

As part of being open about plans, priorities and progress, we published the 2011 Census General Report and laid it before Parliament in March 2015. The report reviewed the whole of the 2011 Census operation from the early planning and consultation through to the production and dissemination of outputs and evaluation. The report noted there were some important successes and a number of challenges and lessons to be learned for any census operation in the future.

Overall, however, the 2011 Census for England and Wales was conducted successfully, in line with international best practice, and provided a sound basis for understanding the nature and diversity of the population.

Deputy National Statistician, Iain Bell, is responsible for delivering the 2021 Census for England and Wales. The census spans three directorates. These are:

  • the Transformation directorate, which is led by Frankie Kay and includes Statistical Design and Service Design, which is responsible for the successful design of the census
  • the Operations directorate, which is led by Pete Benton and includes the Census Delivery area, which is responsible for the successful delivery of the 2021 Census

  • Public Policy Analysis, which is led by Liz McKeown and is responsible for developing analytical outputs for the 2021 Census

In 2023, we will present recommendations to government as to the future of census arrangements. This will be based on the progress made using alternative sources of data, new methodologies and how far new systems will be fit for purpose.

Success criteria for the 2021 Census and beyond

The strategic aims for the 2021 Census are based around the following success criteria. The census supports decision-makers throughout the country, with:

  • results that reflect the country we live in today by meeting quality targets, which are, as in 2011:

    • nationally accurate as measured by a confidence interval of plus or minus 0.2%, with bias less than 0.5% for the population estimate for England and Wales
    • high-quality locally with 95% confidence intervals for the population estimate for each local authority of plus or minus 3%
    • minimal variation within local authority area
    • response rate targets of 94% nationally and 80% locally in all local authorities, to support these quality levels
  • outputs that are timely and easy to use (with first results being released within a year of census day)

The census operations are designed with respondents at the heart to enable delivery of high-quality data for decision-makers. This is achieved by ensuring that:

  • the census is easy and convenient to complete, so 70% of respondents provide data without follow-up

  • we protect respondents’ data, ensuring the data are used for statistical purposes only, and we are seen to protect respondents’ data in everything we do

  • census data reflect the needs of today’s society

  • the census will be predominantly online with 75% of responses provided online, and assistance provided to those who need it, to make this the most inclusive census ever

The census is part of CDCTP, which will deliver value for money for the census and beyond by:

  • delivering benefits of at least five times the cost of the business case

  • delivering the widest possible transformation of the capabilities of the ONS through transformation of its digital and data capabilities, use of integrated data and modernised collection of business and social survey data where it is needed

  • developing new skills and capabilities to support new ways of working, ensuring the ONS can, through surveys, collect the data that surveys are best suited for

Consultation and wider engagement

The design and content of the 2021 Census has been shaped by:

  • requirements from the user community about the types, quality, frequency and detail of outputs required

  • changes in technology, in particular the opportunities offered by the online census, and the willingness and ability of the public to use government online systems

  • improvements in the quality of, and access to, data already held by government and other sources, giving potential for increased use in the production of statistics

  • the continued and ongoing need to make the most effective use of public money

  • successes and lessons learned from the 2011 Census

  • developments and lessons learned in international census taking

The White Paper Help Shape Our Future: The 2021 Census of Population and Housing in England and Wales (PDF, 967.5KB) was published on 14 December 2018. This report:

  • recommends on the content and conduct of the 2021 Census for England and Wales, including the proposal that, for the first time, the census will be predominantly online

  • gives the proposed date for the 2021 Census in England and Wales

  • outlines the topics recommended for inclusion in the 2021 Census and the consultation process that informed these recommendations

  • gives confidentiality provisions for personal data collection and digital security

  • describes plans for the publication of, and access to, 2021 Census data

To develop the proposed census design, we have undertaken extensive engagement with interested parties, users and members of the public. This has taken place through international liaison, formal advisory committees, topic-related working groups, outputs-related working groups and via channels such as formal ONS consultations and the census pages of the ONS website.

We meet quarterly with NRS, NISRA and the WG at the UKCC (Appendix 5) meetings. This enables each agency to keep abreast of plans and progress, to discuss ideas and risks, and to support work on harmonising censuses in the UK. These meetings are complemented by regular discussions at an operational level.

We have held a number of formal public consultations, supported by roadshows and open meetings. These consultations have been open to all and have welcomed views and submissions on the census from all sources. More can be read about our consultations in Public consultations, roadshows and meetings in Section 6.

The CDCTP news and events webpage is used to update our users following notable publications. Events are also publicised such as conferences, roadshows and more targeted community engagement, for example, the Sikh community open meeting.

Our Twitter channel tweets occasionally on behalf of census. We posted tweets announcing the census White Paper being published and explaining why the census matters. We also tweeted when we announced the rehearsal areas. In summer 2019, a 2021 Census Twitter channel and Facebook page will be launched.

Internal and external review

We undergo various independent reviews to evaluate the effectiveness of our statistical processes. The National Statistician’s Quality Reviews (NSQRs), established in 2017, cover thematic topics of national importance, conducted on behalf of, and for, the Government Statistical Service (GSS). These reviews are future facing, ensuring methods used by the GSS are keeping pace with changing data sources and technologies. NSQRs complement existing quality assurance practices, providing an additional tool to make sure methods are, and remain, fit for purpose and among the best in the world.

The latest NSQR covers privacy and data confidentiality methods. The approach and methods being developed to protect confidentiality in the outputs from the 2021 UK censuses are consistent with the findings from the NSQR on privacy and data confidentiality. In particular, they reinforce and encourage a balanced disclosure risk compared with data utility approach (next step 3 in latest NSQR). The use of a new dissemination strategy, a table builder that applies automated disclosure control, has exploited the potential for practical applications of new software and technologies in protecting privacy and confidentiality. Designing and developing this has required us to explore new methods and support the traditional census anonymisation approach with those of the differential privacy paradigm, and highlighted their applications and limitations (next step 6 in latest NSQR).

The Infrastructure and Projects Authority (IPA) provides external assurance to support CDCTP. In July 2018, the IPA carried out a project assessment review on CDCTP. The review ensures we are working efficiently and effectively, and helps us to deliver benefits for society and provide value for money.

More can be read about internal and external audit and peer review in in Developing sound methods in Section 5.

4.5 Principle T5: Professional capability

People producing statistics should be appropriately skilled, trained and supported in their roles and professional development. For more information, please see Principle T5 of the Code of Practice for Statistics.

Recruitment

As part of the Civil Service, we must adhere to the recruitment principles, updated in 2018, which explain the legal requirement that selection for appointment to the Civil Service must be on merit on the basis of fair and open competition.

The Civil Service is changing the way it recruits – making it modern, fairer and more inclusive. The Success Profile Framework is being introduced to attract and retain people of talent and experience from a range of sectors and all walks of life, in line with the commitment in the Civil Service Workforce Plan.

The Success Profile Framework moves recruitment away from using a purely competency-based system of assessment. It introduces a more flexible framework that assesses candidates against a range of elements using a variety of selection methods. This gives the best possible chance of finding the right person for the job, driving up performance and improving diversity and inclusivity.

In the new framework, many of the actions and activities previously included in the Competency Framework are now called “behaviours”. Behaviours are one of five elements within the success profiles, the others being: strengths, ability, experience and technical. We have been piloting the use of some of these elements in recruitment since November 2017 and fully transitioned to this method of recruitment in January 2019.

The roles and responsibilities of those involved in the production of census statistics are clearly defined from the outset. Job advertisements on the Civil Service Jobs website have a job description, with the sections:

  • about you

  • about us

  • the role

  • responsibilities and person specification

  • qualifications needed (if applicable)

  • behaviours

Performance and training

Once in post, staff are required to adhere to the performance management policy, which includes setting goals that are then reviewed on a regular basis with line managers.

We provide the necessary space that people need to learn to allow staff to develop their talent and capability. This includes granting time that is set aside specifically for learning. A standard approach for each learner is five days of learning each year.

Dependent on their profession, staff are encouraged to spend an average of 100 hours learning per year over a three-year period.

All staff are required to complete six mandatory e-learning courses as soon as they are employed and ensure they are completed on an annual basis. One course is “Responsible for Information”. By completing the course, staff will be able to:

  • explain how security classifications help protect information

  • describe their responsibilities when protecting and sharing information both inside and outside of government

  • manage the risks of handling and disposing of information

  • manage the risks when working with information outside the office

  • guard against email and other online fraud

  • explain their responsibilities in recognising and preventing fraud and bribery

  • describe how to manage any risks to a specific information asset they may have

All staff who are new starters in CDCTP attend an induction. The session introduces staff to the purpose, methods and transformation journey of ONS surveys and the census.

Included in the induction course content is the CDCTP Roadmap, which lays out our goals and principles. A copy of it is also included in the new starter pack. Staff are encouraged to think about how their role and the work of their team fits with the roadmap.

Staff working on the census are expected to attend the “Improving Quality in CDCTP” course. The course is available to staff once they have been in post for six weeks. The course introduces the main principles of quality management, quality assurance and quality control.

A further course specifically run for staff working on the census is “Best Practice for Quality Assurance”. The course is available to staff once they have been in post at least a month and for those who are new to working in a statistical role. The course provides guidance on quality assuring work produced in census-related statistical roles. It includes:

  • an understanding of quality assurance processes

  • some of the quality checks that can be carried out on statistical outputs, code, reports and administrative sources

  • tools that can be used to guide the quality assurance process

4.6 Principle T6: Data governance

Organisations should look after people’s information securely and manage data in ways that are consistent with relevant legislation and serve the public good. For more information, please see Principle T6 of the Code of Practice for Statistics.

Confidentiality principles

We fully recognise the importance of privacy, confidentiality and security for the public. There is an obligation on the government to ensure that personal information provided by the public for statistical purposes is treated in the strictest confidence. The statutory requirement to provide personal census information is prescribed by the Census Act 1920 and secondary legislation in the form of the Order and Regulations made under the Act. The Act – alongside the confidentiality provisions of the Statistics and Registration Service Act 2007 – imposes strict requirements on us to protect the confidentiality of the information collected.

We recognise that the public needs to be confident that personal information will be held securely and used only for statistical purposes. As in previous censuses, all the information provided will be treated in strictest confidence and these protections will be made clear to respondents. There is a well-established reputation of maintaining census confidentiality, which we will uphold.

The information collected in the 2021 Census will only be used for the production of statistics and statistical research. In all cases access to, or use of, census data must comply with requirements set out in the Statistics and Registration Service Act 2007 and data protection legislation (the General Data Protection Regulation (GDPR) and the Data Protection Act 2018), as well as the standards set out in the Code of Practice for Statistics.

We are in the process of writing a full data protection impact assessment (DPIA) for both the 2019 Rehearsal and the 2021 Census. We produced a privacy impact assessment (a precursor to DPIAs) at the initial research stage of the census in 2015 and for the 2017 Test. The DPIA meets our statutory obligation within GDPR concerning the processing of personal data.

We have carried out a privacy impact assessment for ServiceNow. This is the incident management system we will be using to log any incidents that occur during the 2019 Rehearsal and the 2021 Census operations, and their impact. This system has been developed by ONS’s Digital Services and Technology department and the census Operations team and allows for the logging, triaging and resolution of any incidents that are raised.

Privacy information

Under GDPR, it is a statutory requirement to provide privacy information to data subjects prior to processing their personal data. We will achieve this by publishing privacy information on our website, including a link to the information clearly within the electronic questionnaire (EQ) system and sending the information alongside paper questionnaires.

We will also provide the necessary privacy information to our staff, whether that be permanent staff or temporary members of the field force. Information about how we process staff personal data will be available on the ONS intranet. If field staff are unable to access this information, privacy information will be provided by other means and will be available on request.

We will ensure that government-wide standards relating to information risk management and data security are met. Census data will be treated in accordance with the principles described in the White Paper (PDF, 968KB) (paragraph 6.5). Our data and security policies that underpin our data principles were updated and published in January 2019.

Statistical confidentiality

A main strength of the census is the ability to generate statistics about small areas and groups of people, so that public policies and plans can take account of the needs of local communities. However, it also poses a risk of inadvertent disclosure of identity if results are published that show small values in particular population counts in small areas. As a result, particular care must be taken to meet both the need to safeguard statistical confidentiality and minimise the damage to the utility of the data.

This is not a new risk, and strong precautions will be taken so that any data made available to the public in any form comply with the confidentiality provisions in the Statistics and Registration Service Act 2007 and the Code of Practice for Statistics. Measures will be taken to prevent the release of statistical information that identifies characteristics about an individual person, household or group of people. Approaches to security are developed in close co-operation with the Registrars General in Scotland and Northern Ireland, who are responsible for the census in those jurisdictions.

In a census context, where thousands of cross-tabulations can be generated from the underlying data, the protection of statistical confidentiality will be best addressed by introducing uncertainty about the true value of small cells. The preferred option is combining a pre-tabular method such as record swapping as the main form of protection (similar to 2011) with a post-tabular cell key perturbation (adding random noise to cell counts) to protect against disclosure through differencing. These methods will enable users to produce outputs in a more flexible and timely manner than for previous censuses.

The three UK census offices have agreed that small counts may be included in publicly disseminated census data if:

  • uncertainty has been created systematically about whether or not the small cell is a true value

  • creating uncertainty does not significantly damage the quality of information

The final method and parameters will be agreed by the National Statistician and the Registrars General in consultation with users and other experts, including the Information Commissioner. Depending on the need, procedures will include:

  • modifying some of the data before the statistics are released, using record swapping with cell key perturbation

  • limiting outputs by simplifying or grouping data, for example, providing information for aggregated age groups

  • amalgamating information to ensure that data are only available in a non-disclosive way, for example, by grouping categories of data or only providing information for larger geographical areas

A joint agreement (PDF, 164KB) has been published by the National Statistician and the Registrars General for Scotland and Northern Ireland covering the statistical procedures to be applied to census data including methods for disclosure control. The procedures to be used will be tested, evaluated and agreed in advance of the census, and co-ordinated across the UK.

Engagement on statistical confidentiality

In April and May 2018, we held a series of roadshows in support of our outputs consultation. At the event, we updated interested parties with information about our plans for statistical disclosure control for the 2021 Census, which is based on a similar approach used in the Australian census. The event included information on cell key perturbation, which introduces a sufficient level of ambiguity or uncertainty by altering some numbers, and record swapping, which swaps a sub-set of records or households.

Overall, the feedback about our proposal on statistical disclosure control methodology was positive, as it would speed up the release of data and allow users to create their own tables. Some users were concerned about the effect of applying the cell key perturbation method and the impact on table totals leading to different estimates depending on how the table were defined.

Controls of further use of census data

We will permit controlled access to census data, while keeping them secure at all times. This will allow approved researchers to realise the potential public benefits of these data, while being able to assure respondents that their details will never become public. Access is permitted to approved researchers using a framework that is commonly called the “Five Safes”: safe people, safe projects, safe settings, safe outputs, and safe data. The Approved Researcher Scheme requires researchers to be accredited by the ONS, and their specific research proposals to be scrutinised to ensure there is a clear public benefit. The criteria for accreditation require that all researchers be fit and proper.

We require each accredited person to sign a declaration to show that they understand the confidentiality obligations placed upon them and we publish a record of all such researchers and the data accessed. Access to census information will only be allowed within a secure environment.

Information asset

Census data are an information asset; a body of information defined and managed as a single unit so it can be understood, shared, protected and exploited efficiently. It is vital that the confidentiality, integrity and availability of ONS information are appropriately protected. To protect information assets from unauthorised access, modification or destruction, it is necessary to ensure that only individuals authorised by the Information Asset Owner (IAO) have access to them and the systems on which they are processed.

An IAO is a senior individual involved in running the relevant business. Their role is to understand:

  • what information is held

  • what is added and what is removed

  • how information is moved

  • who has access and why

As a result, they are able to understand and address risks to the information, and ensure that information is fully used within the law for the public good, and provide written input to the Senior Information Risk Owner (SIRO) annually on the security and use of their asset.  

Back to table of contents

5. Quality

Data and methods that produce assured statistics

Quality means that statistics fit their intended uses, are based on appropriate data and methods, and are not materially misleading.

Quality requires skilled professional judgement about collecting, preparing, analysing and publishing statistics and data in ways that meet the needs of people who want to use the statistics.

5.1 Principle Q1: Suitable data sources

Statistics should be based on the most appropriate data to meet intended uses. The impact of any data limitations for use should be assessed, minimised and explained. For more information, please see Principle Q1 of the Code of Practice for Statistics.

What population information is needed?

Every decade there has been a review to establish what population information is needed, and whether a 10-yearly census is the best way to meet those requirements. In 2014, the then National Statistician recommended (PDF, 236KB) that these needs be met through the conduct of a predominantly online census in 2021, and alongside this, the use of administrative data from across government be fully explored to produce more timely estimates.

The UK government welcomed these recommendations in a letter from the minister for the Cabinet Office (PDF, 99.3KB) on 18 July 2014. The Welsh Government confirmed its support (Word, 183KB) for the proposed census approach in a statement published on 27 March 2014 by the minister for finance.

Recommendations

The recommended date for the census in England and Wales is 21 March 2021. We will enumerate all people in England and Wales who are staying, or expecting to stay, in the UK for six months or more. This is the same enumeration base as in the 2011 Census for England and Wales.

The primary output base for the 2021 Census will be all usual residents. The “usually resident” population refers to anyone who, on the census date, is in the UK and has stayed or intends to stay in the UK for a period of 12 months or more. Anyone temporarily away from their usual residence in the UK for less than 12 months will be counted as a usual resident. This is consistent with the primary output base for the 2011 Census.

Understanding the needs of census data users

The census is the largest peacetime operation in England and Wales, and it is mandatory for all to take part. Therefore, it is important that the topics and subsequent questions collect appropriate data, are of relevance and meet intended user need.

To understand the needs of census data users, in June 2015, we launched a 12-week public consultation. The consultation asked census users for their views and evidence on the topics they would like to be included in the questionnaire in England and Wales for the 2021 Census. Anyone could take part in this consultation and the resulting feedback informed further research that contributed to the development and testing of proposed census questions.

We published our full response to the consultation in May 2016, setting out the evaluation criteria used and updating the initial view (of ONS) on content for England and Wales. We also published individual topic reports, each setting out the evaluation of that topic against the set criteria, the updated view for that topic, equality implications and the next steps to be taken.

In June 2015 and July 2017, we held regional roadshows to share the latest news on the proposed content for the 2021 Census questionnaire.

In addition to the main consultation, we also worked with a number of interested parties to understand detailed needs for specific questions and definitions to be included, for example, on types of communal establishment.

Development of proposed new questions

Testing and development of proposed new census questions, or amendment of existing census questions, follows a similar path for each question:

  1. identify the strength of user need and the specific uses to which the information will be put; this is to guide the type and development of question wording and appropriate response options

  2. if it is a new question, we may start with a survey to assess public acceptability of the question before proceeding with further development

  3. we may undertake cognitive testing through focus groups and smaller, often individual, interviews to help understand how the respondent understands the question

  4. this step is likely to require a number of iterations and is used to test, and refine, question wording (and often associated guidance) and to ensure that response options work, are acceptable and understood amongst these small groups of respondents

  5. the questions that have changed since the 2011 Census, or are being newly included, have undergone cognitive testing in the Welsh language

  6. finally, if a question works in cognitive testing, the question is then tested on a larger sample of the population as a self-response survey, to explore whether the question will work in census-type conditions

This evidence is then considered alongside overall strength of user need, overall respondent burden, space constraints in the paper form and legal duties, to make a recommendation on whether to include a question in the 2021 Census. We published two topic reports in December 2017 and December 2018 that provide an update on our research and testing of 2021 Census questions and topics.

Use of administrative data

Another aspect of the recommendation made in March 2014 included the increased use of administrative data and surveys to enhance statistics from the 2021 Census and improve statistics between censuses.

We are developing our use of administrative data to enhance the 2021 Census by increasing its use in data collection, processing and outputs. This work is in the early stages and no decisions on the use of administrative data have been made. All work is being conducted in accordance with our policy on data security, disclosure and ethics.

The areas where we are investigating the use of administrative data are:

  • collection

  • processing

  • quality assurance

  • outputs

  • geography

These are discussed in more detail in this section.

Collection

For use in supporting or supplementing field collection of “dummy data” – a placeholder questionnaire for basic household information completed by field staff for households where they have been unable to collect a return. This research has identified that Valuation Office Agency (VOA) data may be of use and further research will be undertaken on Council Tax data in the future.

To improve and enhance the quality of the address register. For example, attributes of administrative data, such as flags to denote an address as a care home or unoccupied, may help us deliver an efficient census enumeration process.

To develop a Hard-to-Count (HtC) index to identify sub-populations or geographical (lower layer super output) areas at risk of census non-response. This will be used as a tool in the 2021 Census to support pre-planning including work allocation of field follow-up. Data from Driver and Vehicle Licensing Agency (administrative) and Ofcom (commercial) will be used to develop the methodology.

Processing

For use as auxiliary information during imputation of age (PDF, 590KB). The Census Research Advisory Group (CRAG) has signed off a proof of concept but further implementation still needs to be done and tested before we are able to commit to using the administrative data.

To estimate the under response in large communal establishments and feed into the imputation of missing records.

To provide information that will help us to place donor households (and people within them) to adjust the unit level census database for under coverage.

Quality assurance

We recognise the importance of building on the experience and lessons learned from the 2011 Census to ensure that the quality of the census is maintained and improved, even as introducing new content poses new challenges. As a result, we will use administrative data to improve the quality of the data collected by the census and to increase the value of our outputs. The initial proposal for quality assurance of census data states administrative data will be used to quality assure the 2021 Census population results, to ensure the estimates are plausible when compared with data from administrative sources.

Outputs

To complement the release of census data, we intend to create outputs where census data will be supplemented by integration with other data sources. We are working with tax and benefits data from the Department for Work and Pensions (DWP) and HM Revenue and Customs (HMRC) with an aim to develop administrative data-based income data that can be integrated with the data collected on the 2021 Census. We have already produced some research outputs to demonstrate the potential of this approach.

In 2017, we looked at VOA data as a potential alternative to estimating the number of rooms and bedrooms on the 2021 Census. We held a public consultation to invite users to respond to this proposal.

Given the findings of the consultation and our view that it is only appropriate to ask one question to meet user needs, we recommend (PDF, 968KB) keeping the number of bedrooms question, and removing the number of rooms question. We are developing an approach for edit and imputation for VOA number of rooms.

We will continue to work with the VOA to explore the feasibility of supplementing the census questions with administrative data on total floor space and property type.

We are working with the Department for Education (DfE) to understand whether administrative data could be used to derive highest qualifications obtained. However, it is unlikely the required population coverage can be achieved.

Geography

If administrative data can be used to support statistical geographies, they will support the improved efficiency of census outputs.

For disseminating 2021 Census outputs there is a need to maintain the core statistical building block geography – output areas (OAs) – which change only when there is significant population or household change, as recorded in the census. They are the lowest level at which census estimates are released so need to be in place before census outputs are created.

In 2011, the process needed to maintain OAs and super output areas (SOAs) could only be done once the census data was processed and adjusted for coverage, and is one of the reasons why 2011 Census outputs took longer to be released. To address this, we are investigating whether maintaining the OAs can be done in advance of 2021 Census processing using administrative data to produce indicative revised OAs for some areas. These planned revisions can then be further enhanced as necessary when the final 2021 Census processing has been completed.

For maintaining OAs following the 2021 Census, administrative data will therefore be used to initially identify the population and household changes that will necessitate OA changes (splits or mergers). This will allow the minimum and maximum population and household thresholds to be maintained, then any final adjustments can be made when 2021 Census processing has been completed.

Improving the quality of the census

The data collection operation – the public face of the census – covers the activities carried out to support respondents in completing the census in the way that best meets their needs, while ensuring the relevant quality targets are met. These activities include:

  • publicity and marketing

  • identifying all addresses where someone may be resident

  • writing to each household and communal establishment informing them of the need to take part

  • public support to assist respondents with completing their questionnaire

  • recruiting field staff to provide face-to-face support for non-responding households

Quality is at the heart of planning the census. Many of the planned innovations that were trialled in the census test in 2017 are about improving the quality of census collection processes and therefore the data. A census, by its nature, is designed to cover the whole population, but there are inevitably difficulties with achieving this. A good census design will recognise the limitations. For example:

  • the address register – no matter how up to date – may fail to record all residential accommodation in an area at the time of the census

  • field staff may fail to identify all households within a multi-occupied dwelling, not all members of a recorded household may be included

  • there may be system or processing errors

As outlined in the census White Paper (PDF, 968KB) (paragraph 4.89), the improvements being introduced for the census in 2021 have been designed to minimise the impact of any possible errors. For example:

  • the creation of an up-to-date single address register covering households and communal establishments, together with a response management system functioning in real time; this should ensure that unique access codes (UACs) or questionnaires are delivered to, and returns made or collected from, all residential addresses and that monitoring of return rates can be done in real time

  • the collection of up-to-date, electronic fieldwork information to assist in managing the operation from headquarters, and to support area managers and co-ordinators in the management of work for their teams

  • the use of address co-ordinates to help field staff locate addresses for follow-up visits

  • the use of an electronic questionnaire to make it easier for respondents to complete the questionnaire, and the increased volume of responses received from this channel to make it easier for us to process the data

  • the use of an electronic questionnaire should lead to far lower levels of non-response to particular questions than on paper returns (as demonstrated in the 2011 Census and 2017 Test)

  • focusing on the most cost-effective method of chasing non-response to improve the overall coverage – through targeted marketing, community liaison, reminders and the improved allocation of field staff in areas or in groups of lower response

Census Coverage Survey

Even with these measures to ensure response, there is still a risk of non-response. The Census Coverage Survey (CCS) will be the main source of information on the extent and distribution of a census undercount. The CCS sample will be approximately 1.5% of the postcodes in England and Wales. The sample may have unequal sampling fractions, with higher sampling fractions in the areas with lower census response.

It will be a separate sample survey planned to be carried out over a four-week period, starting six weeks after census day. It will take the form of a short interview, which in 2021 we plan to collect digitally. The field force of approximately 4,600 CCS field staff will conduct an independent enumeration of households and individuals within households in the CCS sampled areas to collect basic demographic characteristics (such as age, sex, marital status, ethnic group and economic activity).

The information obtained from the survey will be matched (or linked) to the census data, and potentially to administrative data. This will help produce a consistent set of census-based estimates, which will form the new base for the ongoing series of population and migration statistics for local and health authorities after 2021. It will also produce census estimates for the whole population on all characteristics. The results of the CCS will also be used to adjust the census database to include households and persons estimated to have been missed, so that the census outputs reflect the whole population.

Confidence intervals for the census-based population estimates will also be computed, allowing assessment against the success criteria described in Success criteria for the 2021 Census and beyond in Section 4.

Census Quality Survey

The Census Quality Survey (CQS) will be a small, separate voluntary survey that will be carried out after the census to estimate the level of respondent error as an indicator of its quality. It aims to measure the accuracy of answers given to census questions by asking a sample of the census questions again. By comparing the responses given in the CQS to those given in the census, agreement rates will be calculated, which will provide an indication of how accurately the census questionnaire has been completed by the general public.

The CQS will be carried out independently of the CCS. It differs from the CCS in that it does not aim to measure how many people were not counted by the census but rather the quality of their responses.

Along with other quality information published about the 2021 Census, the CQS results are intended to help users understand the strengths and limitations of census data and how to use those data appropriately.

5.2 Principle Q2: Sound methods

Producers of statistics and data should use the best available methods and recognised standards, and be open about their decisions. For more information, please see Principle Q2 of the Code of Practice for Statistics.

Developing and assuring the methods

A vital part of developing and assuring the methods is the comprehensive programme of research and testing that is part of the preparation for the 2021 Census. Early planning, testing and evaluation are essential for a successful census and a continuous testing, integration and delivery approach has been adopted for the census (statistical design and operations). Evidence from the 2017 Test has helped shape developments for the upcoming rehearsal.

We aim to use the best available methods through innovative development, sharing best practice and putting in place quality assurance measures. This includes reviewing previous methods and developing and implementing new methods. Specifically, ensuring methods are robust in the context of design changes to the 2021 Census, for example, the “online-first” approach to collection, the increased use of administrative data and changes to the dissemination approach.

Building on the experience and lessons learned from the 2011 Census, we will put in place good practice measures including:

  • setting design goals and standards across main processes to embed quality in the design

  • frequent and open engagement with users to help improve methods and processes to build user trust in results

  • building on the statistical methods used in the 2011 Census and exploring new methods that could improve the statistical accuracy of the census data

  • maintenance of a high level of transparency and openness in quality assurance methods and processes, by involving experts from ONS and externally to provide assurance that methods and processes will produce robust results

  • building on the use of the quality assurance panels introduced as part of the 2001 Census to challenge, scrutinise and endorse census estimates

  • identifying new methods that might improve the quality assurance of census results such as data science techniques and data visualisation tools

  • considering potential ways of using previous census information to look at changes over time that highlight significant and implausible outliers, particularly for small areas

  • the appropriate use of administrative data to improve the quality of the data and to increase the value of our outputs

We are currently in the design phase for the 2021 Census, so many of the methods we will use to conduct the 2021 Census are still under development. The main stages of the method used to produce 2011 Census estimates are described in the 2011 Census Statistics for England and Wales Quality and Methodology Information (QMI) report. Building on this, the main stages planned for the 2021 Census can be found in the White Paper (PDF, 968KB) (Chapter 5: Processing, analysis and output production). For ease, they are described later in this section under the subheading Stages of processing and analysis.

The end-to-end statistical design involves using sound methods for:

  • questionnaire design

  • census operations collection of information via online returns and paper questionnaires

  • processing

  • quality assurance of the data

  • delivery of census outputs

Development of the census questionnaire is well advanced. Requirements gathering through a topic-based consultation and ongoing engagement with users has been followed by rigorous testing carried out on both paper and electronic questionnaires. More information on testing and development of proposed new census questions can be found earlier in this section under the subheading Development of proposed new questions.

We have been working closely with the National Records of Scotland (NRS) and Northern Ireland Statistics and Research Agency (NISRA) to harmonise questions where possible, explaining where there are accepted differences. Harmonisation issues are regularly covered in the quarterly UK Census Committee (UKCC) (Appendix 5). This is chaired by the National Statistician and includes the Head of Profession (HoP) from each devolved administration and the Director General for Population and Public Policy of the ONS.

The collection stage involves paper questionnaires being receipted and returned direct to the questionnaire capture facility where the information provided by respondents will be digitised before being transferred to the ONS for further processing.

Stages of processing and analysis

Data load

Online responses will be merged with data scanned and captured from the paper questionnaires and other information needed for processing. During the field operation, these data will be used to track household completion.

Edit and validation

These processes are designed to deal with inadvertent systematic errors or inconsistencies in the answers provided by respondents, for example, resolving multiple responses from the same household. This process flags any missing, invalid or inconsistent responses (for example, when a respondent does not answer a question on their form, or reports that they are four years old and married). As mentioned previously, an assessment of lessons learned from 2011 was done.

Validation includes resolving duplicate responses. Evaluation from 2011 found some of these were not resolved at the cleaning stage and had to be allowed for later in processing with an overcount adjustment. Research is ongoing to improve this process, for example, by expanding the search radius. An assessment has been made of edit rules applied in 2011, and whether there are any improvements needed for 2021.

Item imputation

This stage resolves the items flagged in the previous stage through the imputation process, which makes appropriate adjustments to the census data leading to complete and consistent census records. It has been accepted since the 1981 Census that it is better for this to be done by us rather than leaving “not answered” or “not stated” categories in the outputs. Leaving these categories in the outputs leads to a variety of assumptions being made by end users as to how to analyse the results.

Our imputation methods will be consistent with internationally recognised statistical standards for large-scale imputation applications such as a census. As part of our research on making better use of administrative data, we are exploring the potential of using such information to support imputation processes.

Coverage assessment and adjustment

The success of the census relies heavily on its accuracy and how well it reflects the population and its characteristics. The census will never count every single household and individual. A necessary process of coverage assessment and adjustment will be undertaken to provide accurate national and subnational estimates of the population.

Post-enumeration survey

The Census Coverage Survey (CCS) will be conducted to measure under- and over-count in the census by undertaking an independent enumeration of a representative sample of the population. The CCS will enable the census population counts to be adjusted for the number of people and households missed in the census at the England and Wales and local authority levels. The CCS records will be matched at the individual and household level with those from the census to produce a consistent set of coverage error-adjusted census outputs. The strategy for the CCS in 2021 is to develop an improved methodology built on the framework of the CCS in 2011 and associated coverage adjustment methodology.

Matching

Census and CCS data are matched to a very high level of quality (minimising false positives and false negatives), using exact automatching, probalistic matching, clerical checking and clerical searching. Matching methods used in the ONS and elsewhere have improved since 2011, and we are taking these developments into account in the development of the 2021 system.

Coverage estimation uses the results of the census and CCS matching to model under-coverage. We are building on the success of the methods used in 2001 and 2011, developing a more robust modelling-based dual-system estimation approach.

We are researching how administrative data can be used to improve the process of adjustment, helping to identify valid non-responding addresses to impute fully imputed households into.

Statistical disclosure control

Statistical disclosure control methods will be applied during processing and immediately prior to dissemination. These will add sufficient uncertainty so that no individual or characteristics of groups of individuals are inadvertently identified during publication of the results. More information on the methods can be found under the subheading Statistical confidentiality in Section 4.

The end-to-end processing of the census data needs to enable us to produce first outputs within a year. This is an ambitious challenge but recognises the need to balance the quality dimensions of timeliness, relevance and accuracy. All of the previously defined methods and checking mechanisms are being built with this in mind, and we will rehearse the live operation in 2020.

Archiving

Paper questionnaires will be destroyed in line with government security guidelines. All systems and storage media will be securely erased in accordance with government security standards. We plan to keep a dataset of copies of responses, from both the scanning process and from the online service, for the National Archives. These data will be held securely and controlled under the policy of keeping England and Wales census returns closed for 100 years.

Developing output products to meet user needs

We have already begun working with users to develop output products that meet user need and to disseminate them effectively. We have publicly stated the aim of providing more timely, flexible and accessible outputs, while safeguarding confidentiality in our consultation. This is discussed in more detail under the subheading Public consultations, roadshows and meetings in Section 6.

To deliver improved outputs, we are introducing innovative statistical disclosure control (SDC) methods and planning to implement new technical dissemination solutions.

We have been working to explain to users the benefits and trade-offs of this approach through working groups, consultation, roadshows and other stakeholder engagement meetings and events. Delivering greater flexibility and timeliness addresses users’ concerns from the 2011 lessons learnt. However, there is a trade-off in the effect of the perturbation on the tabulated results. Our plan is that the main form of protection will be record swapping. Therefore, we are planning only “light touch” perturbation, which should adequately protect against disclosure through differencing and be within the confidence levels of the final estimates.

2021 Census results and findings

As for the 2011 Census, the first results from the 2021 Census for England and Wales will be laid before Parliament, presented to the National Assembly for Wales and published. This is a requirement of the Census Act 1920.

In addition to dissemination of the census results, we will publish metadata in the form of reports on main findings and topic-based analysis, including supporting information from administrative data where possible. We will also supply definitions, classifications and quality indicators, as was done for the 2011 Census. These will include confidence intervals and other quality measures and information from post-census coverage and quality surveys – such as local coverage rates, item non-response and imputation rates. We are beginning discussions with users to better understand their metadata requirements.

Harmonised census outputs across the UK

We are working with the NRS and NISRA to deliver harmonised census outputs for the UK where possible. There is a project looking at providing a single point of access for harmonised results for areas across Great Britain and the UK where the data collected are consistent across the census offices of the UK.

In recognition of the importance of providing consistent outputs across the UK, the National Statistician and the Registrars General for Scotland and Northern Ireland have signed an agreement (PDF, 165KB) on the conduct of the census in 2021. This includes the production of harmonised statistics with specific attention being given to common questions, population bases, definitions, disclosure control methodologies and output formats.

Developing sound methods

To support the development of sound methods, we are seeking assurance through various formal and informal channels. These are detailed in this section.

Working groups

One of the functions of the working groups is to provide assurance to relevant census governance and delivery boards in the ONS. These boards make recommendations related to the design and development of the strategies and methods are based on sound research and principles and meet our methodological standards.

CRAG

High-level methodology assurance group whose purpose is to provide advice on methodology; provide assurance on methodology and evaluate items for the external review. For more information on CRAG, please see Appendix 6.

Census Statistical Outputs Design (CSOD) project board

CSOD oversees the end-to-end statistical quality of processing, focusing on statistical design of methods and how methods are linked together and provide sign off.

External methodology review panel

The purpose of the external methodology review panel is to:

  • provide external, independent assurance and guidance on the statistical methodology underpinning 2021 Census estimates and those based on administrative sources

  • identify significant gaps and risks in methods and make suggestions for mitigation

Census Advisory Groups

Census Advisory Groups (CAGs) represent a variety of interests of the main user communities including central and local government, academia, commercial, health and diversity groups. These advisory groups meet twice a year and provide a forum for discussion on aspects of the census. Advisory groups form an important component of user advice and consultation and will continue to meet up to, and after, the census date.

Working with other national statistics institutes

We have a long history of collaborating with national statistics institutes and academia both nationally (see Consultation and wider engagement in Section 4) and internationally. The design and content of the 2021 Census has been shaped by developments and lessons learned in other census-taking countries. Some examples of our international liaison are detailed in this section.

Annual United Nations Economic Commission for Europe (UNECE) Group of Experts on Population and Housing Census

The main purpose of the UNECE Group of Experts on Population and Housing Census (PDF, 81.2KB) meeting is to:

  • facilitate the exchange of experiences among countries

  • increase knowledge on various issues related to the planning and conduction of population and housing censuses

  • promote the development and implementation of the Conference of European Statisticians (CES) Recommendations for the Censuses of Population and Housing

This results in increased capacity by countries to plan and conduct the population and housing censuses and produce data comparable across countries.

Teams within census have been represented at, and provided papers for, all meetings since 2007. The topics that have been covered include quality assurance, development of the 2021 Census, 2021 Census outputs and dissemination and progress towards an administrative data census.

As discussed in Section 4, we are planning to adopt an innovative statistical disclosure control (SDC) methodology that will underpin the flexible dissemination system. This approach will enable the data to be protected dynamically and speed up the process of data release. This has been discussed previously at the UNECE meeting of the Group of Experts on Population and Housing Censuses 2017 and the UNECE Work Session on Statistical Data Confidentiality 2017, and agreed as the default method for protecting hypercubes (PDF, 629KB) across the European Union in the 2020 census round.

Joint UNECE and Eurostat Work Session on Statistical Data Confidentiality

This meeting takes place every two years. Our Statistical Disclosure Control (SDC) methodology team have been represented at, and provided papers for, all meetings since 2009. UNECE holds workshops on statistical data editing every 18 months to two years. Our edit and imputation team regularly attends and present papers.

International Census Forum

The International Census Forum (ICF) (PDF, 467.2KB) is an annual meeting of English-speaking census-taking countries (Australia, Canada, USA, New Zealand, Ireland, Scotland, Northern Ireland and England and Wales). It is a great opportunity to hear first-hand in a safe and honest environment about the experiences and challenges faced by colleagues across the world.

As part of the ICF, community of practice meetings were established in May 2014. There are approximately 6 to 12 active technical working groups that start up as needed, then dissolve as decided by the group. The groups meet via teleconference and the frequency of meetings differs depending on the group. The same census-taking countries are represented as stated earlier in this section.

5.3 Principle Q3: Assured quality

Producers of statistics and data should explain clearly how they assure themselves that statistics and data are accurate, reliable, coherent and timely. For more information, please see Principle Q3 in the Code of Practice for Statistics.

Processing and quality assurance of data

A critical part of producing census statistics is the quality assurance process. This ensures the methods have been correctly applied and that the statistics are robust and fit for purpose.

We recognise the importance of building on the experience and lessons learned from the 2011 Census to ensure the quality of the census is maintained and improved. Introducing new content also poses new challenges. The measures we will put in place can be found earlier in this section under the subheading Developing and assuring the methods.

Following the collection of information via online returns and paper questionnaires, the responses will be processed and quality assured. This will enable us to gather a set of records that can be used to produce census outputs.

As part of the processing operation, the quality assurance checks that will be carried out will ensure:

  • changes made within each of the processes are robust and do not introduce systematic error

  • the national and subnational (in particular, local authority) estimates are plausible when compared with data from administrative sources and demographic comparators and with demographic analysis (such comparisons may call for additional reviews to be carried out)

We will publish information on the quality of census results to accompany and support the release of the main statistical outputs, as we did for the 2011 Census.

Census Quality Survey

Further quality assurance for the 2021 Census will involve the Census Quality Survey (CQS), which assesses the accuracy of responses. The results are intended to help users understand the strengths and limitations of census data and how to use those data appropriately.

We will publish a CQS results report, which will provide an indication of how accurately the 2021 Census questionnaire was completed by the public. More information on CQS can be found earlier in Section 5.1 Principle Q1 under the subheading Census Quality Survey.

Quality assurance of 2021 Census

We published a Quality Assurance Evaluation report (PDF, 1.0MB) for the 2011 Census. This report describes the methods and processes used to quality assure the 2011 Census estimates. The report:

  • summarises how the 2011 Census quality assurance methodology was designed and applied

  • evaluates the outcomes

  • outlines the lessons learned

  • provides suggestions to improve quality assurance processes for future censuses and other large-scale population estimation projects

This report is an invaluable reference source when considering the design and development of a quality assurance methodology for the 2021 Census. We intend to publish a similar report following the 2021 Census.

To assure the estimates, there is a strand of work to compare the census results with other sources. We will assess the census results using demographic analysis and in the context of data sources such as the Longitudinal Study and Personal Demographic Service data. This work will be carried out with support from CRAG, the Methodology Data Science Expert group (for the development of innovative methods proposed by the University of Southampton), and demographic and topic experts from across the office.

As with the 2011 Census, in the event of a correction being required to census data after publication, we will follow the revisions policy. This will outline the process for amending data and notifying users of the correction.

Back to table of contents

6. Value

Statistics that support society’s needs for information

Value means that the statistics and data are useful, easy to access, remain relevant, and support understanding of important issues.

Value includes improving existing statistics and creating new ones through discussion and collaboration with stakeholders, and being responsible and efficient in the collection, sharing and use of statistical information.

6.1 Principle V1: Relevance to users

Users of statistics and data should be at the centre of statistical production; their needs should be understood, their views sought and acted on, and their use of statistics supported. For more information, please see Principle V1: Relevance to users.

Design and content of the 2021 Census

The design and content of the 2021 Census has been shaped by requirements from the user community about the types, quality, frequency and detail of outputs required.

To develop the proposed census design, we have undertaken extensive engagement with interested parties, users and members of the public. This has taken place through international liaison, formal advisory committees, topic-related working groups and via channels such as formal Office for National Statistics (ONS) consultations and the census pages of our website.

Public consultations, roadshows and meetings

We have held a number of formal public consultations, supported by roadshows and open meetings. These consultations have been open to all and we have welcomed views and submissions on the census from all sources.

The first consultation was on the census and the future provision of population statistics in England and Wales, which ran from 23 September to 13 December 2013. It proposed two future approaches to census taking in England and Wales. These were:

  • once a decade, like that conducted in 2011, but primarily online

  • using existing government data and compulsory annual surveys

The findings from this consultation, as well as evidence from international comparisons, cost-benefit analysis and an independent review led by Professor Chris Skinner in March 2014, resulted in the then National Statistician recommending (PDF, 236KB):

“An online census of all households and communal establishments in England and Wales in 2021, as a modern successor to the traditional, paper-based decennial census…”

More information on the recommendation can be found under the subheading What population information is needed? in Section 5.

In June 2015, we launched a 12-week public consultation asking census users for their views and evidence on the topics they would like to be included in the 2021 Census questionnaire in England and Wales. Anyone could take part in this consultation and the resulting feedback informed further research, which contributed to the development and testing of proposed census questions.

Our full response to the consultation was published in May 2016, setting out the evaluation criteria used and updating the initial ONS view on content for England and Wales. We also published individual topic reports, each setting out the evaluation of that topic against the set criteria, the updated view for that topic, equality implications and the next steps to be taken.

In June 2015 and July 2017, we held regional roadshows to share the latest news on the proposed content for the 2021 Census questionnaire.

In addition to the main consultation, we also worked with a number of interested parties to understand detailed needs for specific questions and definitions (for example, on types of communal establishment) to be included. Parties consulted included government departments, members of the long-standing Census Advisory Groups, topic experts and methodologists within the Government Statistical Service (GSS), who were regularly consulted for their in-depth, specialist knowledge.

In June 2017, we launched a 12-week public consultation asking census users for their views on how changing to an administrative data source as part of the census to estimate number of rooms and bedrooms would affect their work. Anyone could take part in this consultation and the resulting feedback helped us to understand the impact on their work, and will further our ability to meet user’s need.

This followed a commitment made in the 2021 Census topic consultation, where it was decided it was not appropriate to continue to ask two questions designed to meet a single information need. In response, we are exploring the use of administrative data from the Valuation Office Agency (VOA) to produce estimates for number of rooms while keeping a survey question on number of bedrooms.

We published our full response to the consultation (PDF, 487KB) in October 2018. This set out the responses received, and gave a recommendation on whether to continue asking the two questions.

In February 2018, we launched a 12-week public consultation asking census users for their views on our initial proposed design of 2021 Census outputs and the dissemination channels for England and Wales. Anyone could take part in this consultation and the resulting feedback is informing further research and development of 2021 Census outputs. It will also inform the order in which we release our products.

In April and May 2018, we held regional roadshows in support of the consultation.

We published our full response to the consultation in December 2018. This set out the evaluation criteria used, updating the initial proposed design of 2021 Census outputs for England and Wales and the next steps to be taken. The next steps are:

  • continue to develop user-defined outputs with single point of access and access via an API (with wider benefits to other datasets)

  • develop integrated outputs using administrative data to provide data on number of rooms and, if possible income

  • continue developing output products including specialist products, such as origin-destination and microdata

  • develop an analytical work programme, in particular to meet commitments set out in the White Paper (PDF, 968KB) to support communities, for example, Sikh, Cornish

Research and testing

A comprehensive programme of research and testing has underpinned our proposals for the 2021 Census. Early planning, testing and evaluation are essential for a successful census and a continuous testing, integration and delivery approach will be adopted.

The programme includes both large- and small-scale tests and quantitative and qualitative research. We conduct these voluntary tests and research prior to a census to:

  • ensure all questions included meet user requirements and are publicly acceptable

  • trial new technology

  • test procedures and processes

  • plan dissemination

  • test different facets of the whole census operation

Public participation in the census tests is voluntary and we are grateful to the members of the public and the organisations who have taken part in these essential trials.

2017 Test

We conducted a large-scale census test in England and Wales in March, April and May 2017. The aims included:

  • evaluating ways of maximising “self-response” (the number of responses before field follow-up commences)

  • assessing online response and assistance for the digitally excluded

  • assessing the feasibility of some major innovations in operational procedures, such as the use of technology in the field operation

  • evaluation of the effect on the level and quality of response of the inclusion of a question on sexual orientation

The test involved a total of 208,000 households, split across two components. The two components were testing questionnaire and field follow-up and testing materials.

The overall response to the 2017 Test enabled us to draw conclusions from the statistical and operational processes tests. More can be read about the test and its results in the White Paper (PDF, 968KB) (see paragraphs 2.25 to 2.44).

Census Test Evaluation Survey

We carried out a Census Test Evaluation Survey (CTES) after the 2017 Test, as an interviewer-led, follow-up doorstep survey. The CTES garnered views on the new and modified questions from respondents and importantly sought out non-responding households to understand their views on the census and why they did not respond.

We also asked for people's views about the materials sent out and engaged with them to:

  • try to understand better their intentions on using online to complete the census

  • establish whether those that filled in the paper questionnaire could be converted to using online services

  • determine how they would like those services to be supported for them to be useful

The knowledge gained from the CTES methodology provides useful insight for future household social surveys and evaluation.

6.2 Principle V2: Accessibility

Statistics and data should be equally available to all, not given to some people before others. They should be published at a sufficient level of detail and remain publicly available. For more information, please see Principle V2 of the Code of Practice for Statistics.

Benefits of the census

The main benefits of the census will be realised by all users having access to high-quality statistics, released free at the point of use, to inform services such as decision-making, resource allocation, service planning and delivery.

To maximise these benefits, we will work with users to develop appropriate products and to deliver them effectively. Statistics will meet the highest standards set out in the Code of Practice for Statistics and our emphasis will always be on safeguarding confidentiality.

Accessible outputs

Our vision is for flexible, timely and accessible census outputs. We want to make as many data available at the lowest geographic levels as possible and as easily as possible so users can access the data they need, while protecting confidentiality.

The publication of a web-based interactive dissemination system where users can specify the data they require will make the results of the census more accessible to users. We will also be able to deliver these results faster and more flexibly.

When developing our outputs and dissemination systems, we will consider the needs of people with disabilities. We will adhere to the Style.ONS guide, which aims to make statistical content more open and understandable, based on editorial research and best practice.

The responses to the outputs consultation clearly indicated the importance of providing metadata that are easily accessible to all users.

Further user research will be carried out to consider the needs of different user types and inform our planning and design of metadata. This includes how we will provide metadata through the web-based interactive dissemination system and when using an API and Geographical Information System (GIS) software. We will also commit to providing a Welsh translation of metadata, where required.

A large volume of 2011 Census metadata on our website are supplied as PDFs. PDFs are less accessible than two-dimensional web pages and therefore metadata for the 2021 Census will be published on our website as two-dimensional web pages. This makes it easier for users of all abilities to access the metadata, as screen readers and other assistive technologies can struggle if PDFs are not correctly formatted. This is in line with the Style.ONS guide and ties in with the Government Digital Strategy.

We are working towards making our census data and analysis as engaging and accessible to as wide a range of audiences as possible. We will outline our plans for this closer to the release of 2021 Census data.

We offer a commissioned output service that provides users with additional tables not included in census standard publications. The person or organisation requesting an ad hoc table is charged for the service and, once the tables are created, they are published for all to use under Open Government Licence. Details of the pricing policy are printed in the census commissioned table request form (PDF, 3.75MB), showing the different costs according to the length of time it takes to produce the requested table(s).

6.3 Principle V3: Clarity and insight

Statistics and data should be presented clearly, explained meaningfully and provide authoritative insights that serve the public good. For more information, please see Principle V3 of the Code of Practice for Statistics.

Providing insight

We understand the variety of uses and potential uses of census data and appreciate that, to fully aid their interpretation, we need to provide insight. The two main factors to achieving this are provision of analysis, and provision of metadata and supporting information.

We are currently developing our analytical programme, which will support the release of 2021 Census data. As outlined in the White Paper (PDF, 968KB) (see paragraph 1.35), we will provide analysis – including topic-based reports and visualisations – rather than just data tables, to support communities in understanding their needs.

The aims of these analyses are:

  • to provide the most relevant and insightful commentary for these communities

  • to better understand their needs

  • to inform policy and service delivery

This will be done through analysing the full range of census topics alongside other relevant data sources to comment on particular themes, such as health.

The Census Analysis Working Group (see Appendix 7) will ensure a collaborative and coherent approach to the production and dissemination of 2021 Census analysis and related outputs. Specifically, it will:

  • agree and implement a stakeholder engagement plan to ensure that activities and messages are aligned

  • agree a flexible analysis plan that is responsive to stakeholder priorities and user needs

  • design and publish analysis outputs that aim to achieve public and policy impact

We are engaging with users to understand their requirements and discuss plans for making results available, as described earlier in this section under the subheading Public consultations, roadshows and meetings. Users will be able to use the new web-based interactive dissemination system to specify data they require and create their own tables. This will enable us to provide an increased analytical offering to support communities in understanding their needs.

Consistent outputs across the UK

In recognition of the importance of providing consistent outputs across the UK, the National Statistician and the Registrars General for Scotland and Northern Ireland have signed an agreement (PDF, 164KB) on the conduct of the census in 2021. This includes the production of harmonised statistics, with specific attention being given to common questions, population bases, definitions, disclosure control methodologies and output formats.

We are working with the National Records of Scotland (NRS) and Northern Ireland Statistics and Research Agency (NISRA) to deliver harmonised census outputs for the UK where possible. There is a project looking at providing a single point of access for harmonised results for areas across Great Britain and the UK where the data collected are consistent across the census offices of the UK.

We have already started work with our colleagues in Scotland and Northern Ireland to enable the harmonisation of outputs and dissemination of UK data.

International commitments to supply census data

A large part of the UK-level data produced in 2011 was required by European Union (EU) census legislation. We will need to review our exact approach as the outcome of the EU withdrawal negotiations become clearer, but there are some clear advantages of producing the UK-level census statistics required by the EU.

We will continue to meet our international commitments to supply census data to the United Nations Statistical Commission following the 2021 Census.

Quality and accuracy of census data

In order for the benefits of the census to be realised, decision-makers need to be assured of the quality and accuracy of these data when making decisions.

Outputs from the 2021 Census will be accompanied by comprehensive metadata. This will facilitate understanding of the outputs by describing features such as definitions, classifications and quality indicators, similar to the 2011 Census. The metadata will also include confidence intervals and other quality measures. Information from post-census coverage and quality surveys such as local coverage rates, item non-response and imputation rates will be published.

We will continue to publish supporting information that helps aid interpretation of the statistics. This includes explaining the difference between the census and other datasets, comparability over time and comparability across the UK, similar to the 2011 Census.

At the census outputs consultation roadshows, several attendees volunteered to help develop relevant and accessible metadata.

6.4 Principle V4: Innovation and improvement

Statistics producers should be creative and motivated to improve statistics and data, recognising the potential to harness technological advances for the development of all parts of the production and dissemination process. For more information, please see Principle V4 of the Code of Practice for Statistics.

Developments in technology and census data

The ever-changing nature of society determines that user needs for census data are regularly refined. This necessitates change to each census operation and, for 2021, we will use developments in technology to support this.

The technology must be thoroughly tested before being adopted, allowing time to make changes and then to retest where necessary. In particular, the 2021 Census aims to adopt:

  • primarily online-first design

  • new technologies

  • flexible dissemination of statistical outputs

These are discussed in more detail in this section.

Primarily online-first design

A natural development from the traditional paper-based census, a response to the rapid changes in society and technology, and in line with the approach implemented by the main census-taking countries in the recent census round.

New technologies

To improve the efficiency and timeliness of the data collection and processing operations: online data collection offers a number of advantages in terms of improving data quality and therefore reducing the extent of data cleaning required during initial processing. The online questionnaire will incorporate question validation, which will assist users in correctly responding to questions. User-optimised indexes are likely to be employed, which present potential question responses to users as they provide their responses online.

Flexible dissemination of statistical outputs

To maximise benefit: using developments in statistical disclosure control (SDC) methodology and technology to provide more accessible results earlier, while protecting confidentiality.

2017 Test

We conducted a large-scale, voluntary census test in England and Wales in March, April and May 2017. The aims included:

  • assessing online response and assistance for the digitally excluded

  • assessing the feasibility of some major innovations in operational procedures, such as the use of technology in the field operation

More information on the results from the 2017 Test can be found in the White Paper (see paragraphs 2.34 to 2.44). These innovations will be refined via further testing and intelligence gained from the 2019 Rehearsal.

Improvements for the 2021 Census

The improvements that are being introduced for the census in 2021 have been designed to minimise the impact of any possible errors during data collection.

For example, the development of a field operations simulation model (FOS) and response chasing algorithm (RCA) to optimise field collection activities to best respond to the challenges of meeting agreed programme quality standards for coverage (94% response rate nationally; 80% minimum response rate at local authority level). Meeting these standards ensures that the population estimation procedure is fit for purpose and underpins the success of the census overall, providing users with sufficient assurance in the outputs being delivered.

FOS delivers estimates of the necessary field force effort required to undertake field follow up to encourage households and communal establishments to respond. RCA provides suggested intervention actions should responses fall behind expected response profiles during the operational period of the census. Both systems underpin the collection development activities that support the ONS’s push to a digital-first census.

More information on FOS and RCA can be found in the slide deck presented at the Integrated Data for Population Statistics Conference (slides 49 to 80 and 97 to 117 respectively).

In processing stages, improvements are being introduced, too. These include appropriate use of administrative data and improved imputation matching, estimation and disclosure control techniques. These also make use of better paradata available from the improved data collection technology.

Advances in the availability of administrative data and linkage techniques also mean we are improving the census data through the research into integrated outputs for the 2021 data, and the research ongoing for transforming the approach going forward.

User feedback on the development of the 2021 Census

We have a strong commitment to involving users in the development of the 2021 Census, particularly regarding questionnaire content and outputs dissemination. We carried out a public consultation to gather user feedback on both these elements. This has informed subsequent planning and development. Each of the consultations held are examples of involving users in the ongoing development of statistics and data.

At our outputs roadshows, held to promote the consultation, we led a hands-on session enabling users to use the prototype flexible dissemination system. We used this as an opportunity to gather user feedback, which could then be used to further develop the system.

More information on the consultations held can be found earlier in this section under the subheading Public consultations, roadshows and meetings.

More information on the SDC methodology that will enable the use of a flexible dissemination system can be found under the subheading Statistical confidentiality in Section 4.

Working with other census-taking countries

We keep up to date with developments to improve census statistics and data via engagement and collaboration with other census-taking countries, both within the UK and internationally. More information is given in Consultation and wider engagement in Section 4, Working with other national statistics institutes in Section 5 and Consistent outputs across the UK earlier in this section.

We have sought to draw on good practice by collating a report on census quality assurance (QA) in other national statistics institutes. This is being used to develop our initial proposal for QA. We have circulated the initial QA proposal to other census-taking organisations in the UK and shared it more widely (for example, with the Census Advisory Groups) in January 2019.

We have also carried out a review of international census websites and online table builder designs. The aim of the research is to provide recommendations of best practice to inform the design and development of the flexible dissemination systems and the census website and to inform user research. These results will help us to visualise user requirements, enabling us to feed into the design of our website and flexible dissemination system.

6.5 Principle V5: Efficiency and proportionality

Statistics and data should be published in forms that enable their reuse. Producers should use existing data wherever possible and only ask for more where justified. For more information, please see Principle V5 of the Code of Practice for Statistics.

Using the right data sources

The topics proposed for the census are those:

  • that have a strong and clearly-defined user need

  • where questions have been devised that can be expected to produce reliable and accurate data

  • where there is no other comparable and accessible source of the information able to meet the need

The 2021 Census will only collect information if there are no other means of obtaining the required data. When evaluating user need for topics, consideration was given to whether or not suitable information is available from other sources, for example, administrative records or sample surveys.

Information about gender identity, except in relation to sex being male or female, and sexual orientation has not previously been collected in the census. Following feedback from the 2021 Census topic consultation, we committed to taking forward work on exploring whether and how to collect information on gender identity and sexual orientation.

As stated in the White Paper (PDF, 968KB) (see paragraph 3.23), it is proposed that questions on gender identity and sexual orientation will be included for the 2021 Census. The work undertaken resulting in this proposal is described in the White Paper (see pages 35 to 42) and can also be found on our question development web pages for gender identity and sexual orientation. This covers reviewing existing approaches to the collection of data about gender identity and sexual orientation in the UK and internationally.

Reducing the burden on the public

The results of the 2015 topic consultation indicated an ongoing need for the vast majority of topics covered in the 2011 Census as well as a range of requirements for additional topics. These additional topics included:

  • armed forces community

  • gender identity

  • health conditions

  • place of birth

  • reason for migration

  • sexual orientation

  • income

  • volunteering

The differing needs were balanced against increasing the burden on the public. A stated aim of the consultation was not to increase the burden from the previous census and the length of time to complete the census (not more than 20 minutes for a four-person household).

In 2017, we looked at Valuation Office Agency (VOA) data as a potential alternative to estimating the number of rooms and bedrooms on the 2021 Census. We held a public consultation to invite users to respond to this. The consultation found that:

  • more respondents used number of bedrooms than number of rooms

  • the requirement for number of rooms is for under- and over-occupancy, which can be calculated from number of bedrooms

  • the data quality (as measured by the Census Quality Survey) for a question on number of bedrooms is greater than for a question on the number of rooms

  • in addition, VOA data include information on floor space, which could also be valuable to understanding cramped living conditions

Given these points and our view that it is only appropriate to ask one question, to minimise respondent burden, we recommend keeping the number of bedrooms question, and removing the number of rooms question. We will continue to work with VOA to explore the feasibility of supplementing the census questions with administrative data on total number of rooms, floor space and property type.

In the 2011 Census, respondents were asked if they had ever worked, and if so, were then required to write in the year they last worked. The information collected was used primarily by local authorities to profile and understand the needs of the long-term unemployed. However, this has a high respondent burden, and the evidence to support the ongoing need for information on the actual year someone last worked was limited and not sufficient to justify its inclusion in the 2021 Census.

We will explore whether there are alternative ways to provide this information, for example, through the use of administrative data from the Department for Work and Pensions (DWP).

More information on changes to the questionnaire, including which topics and questions are proposed to be added, amended or removed can be found in the White Paper (PDF, 968KB) (see Chapter 3).

Participation in the census

The fact that the census aims to cover the entire resident population of England and Wales, without exception, gives it particular value when compared with voluntary social sample surveys. The statutory requirement to provide personal census information is prescribed by the Census Act 1920 and in the Order and Regulations made under the Act.

The Census Quality Survey (CQS) that will be carried out after the census to estimate the level of respondent error as an indicator of its quality will be a small, separate voluntary survey.

The Census Coverage Survey (CCS) that will be the main source of information on the extent and distribution of the census undercount is a large-scale voluntary survey.  

Back to table of contents

7. Next steps

The Office for Statistics Regulation (OSR) will carry out the phase 1 assessment using this report as evidence, as well as other documentation in the public domain. As part of the assessment, the OSR will also meet with:

  • census teams, to gather a deeper understanding of how we are working

  • census data users and stakeholders, to garner their views and experiences of the 2021 Census for England and Wales

Once OSR has completed its assessment, it will be published on the UK Statistics Authority website. The report is likely to include recommendations that we will need to take forward and address.

If any of the recommendations require us to respond within a short time frame, we will publish an update on how we have responded to those recommendations on our National Statistics Accreditation web page, to keep users notified of our progress.

Recommendations that do not require an immediate response will be addressed in the phase 2 assessment report.

Back to table of contents

8. Appendix 1: Terms of Reference for the 2021 Census Outputs and Dissemination User Group

Formation date: November 2016

Purpose and scope

The purpose of the group is to:

  • provide ongoing user input to development work relating to dissemination of the 2021 Census

  • provide guidance on user engagement

The scope covers the development of statistical disclosure control (SDC) and dissemination methods, output content and user engagement activities.

Membership

  • ONS 2021 Census Outputs and Dissemination

  • Greater London Authority

  • Home Office

  • UK Data Service

  • Welsh Government

  • University of Southampton

  • Bristol City Council

Operations

The group will predominantly work via email but will possibly meet by phone, video or face-to-face if deemed necessary.

The Office for National Statistics (ONS) will share progress on development with specific questions for the user group to address.

The group agree to:

  • work collaboratively

  • provide constructive challenge

  • treat information the ONS shares as confidential so that wider user engagement can be managed

  • provide advice on how the ONS engages more widely with users

A review of the group will be scheduled at six-monthly intervals to ensure that its objectives are being met.

Back to table of contents

9. Appendix 2: Terms of Reference for the 2021 Census Origin Destination Working Group

Purpose

The main purpose of the working group will be to advise the Office for National Statistics (ONS) on the design, creation, storage and dissemination of 2021 Census origin-destination products by bringing together the ONS Census and the user community. It will ensure that origin-destination products meet user needs and are made available in a timely manner.

Objectives and scope

The working group has the following core objectives:

  • to provide an appropriate forum to discuss origin-destination products

  • to ensure the interests and views of users are represented

  • to advise on the content of census origin-destination products (specifically to include advising on the consideration of content trade-offs), including: statistical disclosure control methods, including any potential effects on user access; and UK harmonisation of origin-destination products

  • to advise on content for supporting metadata for origin-destination products

  • to advise on dissemination of origin-destination products including storage and terms and conditions of access to outputs

  • to advise on plans (including creation of timelines) for the production and publication of origin-destination products

  • to advise on plans for stakeholder engagement focused on origin-destination products

UK position

The working group will aim to ensure 2021 outputs are harmonised where possible. Therefore, representation from the National Records of Scotland (NRS) and the Northern Ireland Statistics and Research Agency (NISRA) will be sought. As the NRS and NISRA are responsible for their own data and outputs, membership of this group does not prevent the NRS and NISRA taking alternative actions to meet the needs of their own national users or carrying out their own user engagement.

Exclusions

The responsibilities of the working group do not include:

  • detailed planning of specifications

  • day-to-day production activities

  • the IT used in production

Table 2 provides a full list of the membership of the Origin Destination Working Group and the various roles and responsibilities of the organisations and departments.

At each meeting, a decision will be made on whether the agreed actions require input from or liaison with other teams or boards.

Meetings

  • Meetings will initially be held quarterly.

  • The meeting requirements will be reviewed quarterly, and updated if necessary.

  • The Census Outputs and Dissemination team will be responsible for organising dates, venues and distributing agendas and minutes to members.

  • Agendas will be circulated one week before the date of the meeting.

  • Members who are unable to attend should send their apologies to the Outputs and Dissemination team and are encouraged to delegate the meeting to another member of their team where appropriate.

  • We will update all users of our progress and decisions via appropriate communication channels and the ONS website.

Reporting

The working group will report to the Census Statistical Outputs and Design (CSOD) project board, UK Harmonised Working Group (UKHWG) and the UK Census Committee (UKCC) as required.

  • The sign-off process is still to be confirmed.

  • The group will consider sharing work with stakeholders and the public if deemed appropriate.

  • The outputs and dissemination team will submit a final recommendation for the design of origin-destination products to the CSOD Project Board for final sign-off by December 2020.

Back to table of contents

10. Appendix 3: Terms of Reference for the 2021 Census Microdata Working Group

Purpose

The main purpose of this working group is to advise the Office for National Statistics (ONS) and other Census offices on the design, creation, storage and dissemination of microdata products following the 2021 Census. This will be achieved by bringing together members from the ONS Census directorate, the census outputs teams of both the Northern Ireland Statistics and Research Agency (NISRA) and the National Records of Scotland (NRS), and the microdata user community. This group will seek to ensure that 2021 Census microdata products meet user needs as far as possible and are made available in a timely manner.

Objectives and scope

The working group has the following core objectives:

  • to provide an appropriate forum to discuss microdata products

  • to attain an insight into the interests and views of users

  • to advise on the content of census microdata products (specifically to include advising on the consideration of content trade-offs), including:

    • statistical disclosure control methods, including any potential effects on user access
    • sample size and design
    • microdata derived variable specifications
    • use of administrative data in microdata products
    • UK harmonisation of microdata products
  • to advise on content for supporting metadata for each microdata product

  • to advise on dissemination of microdata products, including storage requirements and terms and conditions of access to these products

  • to advise on plans for release of microdata products

  • to advise on plans for any stakeholder engagement focused on microdata products

Exclusions

The responsibilities of this working group do not include:

  • the production of longitudinal or historical Census microdata products

  • the technical implementation (including rules and parameters) of statistical disclosure control (SDC) methodology

  • detailed planning and creation of microdata specifications, including those of variables

  • day-to-day production activities

  • specification of IT used in the production of microdata products and SDC processes

UK position

The working group will aim to ensure that 2021 Census microdata products are harmonised across the UK Census offices where possible. As a result, representation from the NRS and NISRA will be required. Where the NRS and NISRA are responsible for their own data and outputs, membership of this group does not prevent the NRS and NISRA taking alternative actions to meet the needs of their own national users or to carry out their own user engagement.

Membership

Additional attendees may be invited on an ad-hoc basis to respond to specific requirements. Table 3 provides a list of the membership of the 2021 Census Microdata Working Group and the various roles and responsibilities of the organisations and departments.

Reporting and authority

The working group will report to the Census Statistical Outputs and Design (CSOD) project board, UK Harmonised Working Group (UKHWG) and the UK Census Committee (UKCC) as required.

  • The group will consider sharing work and reports with stakeholders if deemed appropriate.

  • A final recommendation will be made for the design of microdata products to the Census Statistical Outputs and Design Project Board for final sign-off by December 2020.

  • A more detailed recommendation will be included in a future consultation.

Meeting schedule

  • Meeting requirements will be reviewed quarterly and updated if necessary.

  • Meetings will initially be held as and when required, with an aim to meet at least twice in 2018.

  • These meetings will be organised by the Census Outputs and Dissemination team, who will be responsible for the dates, venues, agendas, minutes and actions of meetings.

  • Agendas will be circulated one week before the date of the meeting, and minutes and actions will be circulated shortly after each meeting.

  • Members unable to attend should inform the Outputs and Dissemination team in advance, and are encouraged to delegate attendance to another member of their team where appropriate.

  • We will update all users of our progress and decisions via appropriate communication channels and the ONS website.

Back to table of contents

11. Appendix 4: Terms of Reference for the 2021 Census Outputs Data Distributor Group

Formation date: September 2018

Purpose and scope

The purpose of the group is to provide ongoing user input to development work relating to dissemination of 2021 Census outputs; in particular, use of Application Programming Interfaces.

The scope covers the development of dissemination methods and formats, technologies and associated user engagement activities.

Membership

  • ONS 2021 Census Outputs and Dissemination

  • Pattern Recognition and Image Analysis

  • UK Data Service

  • University of Leeds

  • Nomis

  • Greater London Authority

  • University of Southampton

  • CACI Ltd

Operations

The group will predominantly work via email but will possibly meet by phone, video and face-to-face if deemed necessary.

The Office for National Statistics (ONS) will share progress on development, with specific questions for the user group to address.

The group agree to:

  • work collaboratively

  • provide constructive challenge

  • treat information the ONS shares as confidential so that wider user engagement can be managed

  • provide advice on how the ONS engages more widely with users

A review of the group will be scheduled at six-monthly intervals to ensure that its objectives are being met.

Back to table of contents

12. Appendix 5: Terms of Reference for the UK Census Committee (UKCC)

Purpose

The UK Census Committee (UKCC) has several purposes:

  • to co-ordinate UK-wide work and review alternative approaches to meeting future user needs for population and small area socio-demographic statistics

  • to ensure that censuses are run in England and Wales, Scotland and Northern Ireland with a consistency of approach, so as to meet requirements from users and as required by International and European obligations

  • to agree the census delivery approaches and plans in a way that takes full advantage of harmonisation of working practices and sharing of research and resources, especially in relation to commercial and procurement activities

  • to share plans and approaches to the use of administrative data for the 2021 Census to share research and learning and maximise consistency of outcome for 2021 Census and later

Role

  • To agree the scope of cross-UK working and co-ordination and to provide strategic input to this work, taking into account:

    • the independent authority of the ONS and the Devolved Administrations (DAs)
    • the wider initiatives underway in the DAs and the ONS
    • the potential benefits and cost savings from adopting a collaborative approach
    • the need to meet users’ needs and international obligations through the harmonisation of UK outputs
    • United Nations Economic Commission for Europe (UNECE) and Eurostat recommendations on international censuses and the legal requirements for the provision of UK aggregate statistical data to Eurostat
  • To review risks and issues and implications arising from the 2021 Census programme of work.

  • To advise on the methodological approaches to be undertaken including the approach to statistical disclosure control.

  • To review and agree the coherence of long-term strategic plans giving due consideration to the legislative implications, to user requirements, and to public acceptability.

  • To co-ordinate any legislative changes that have a direct impact on the running of the censuses.

  • To provide assurance that plans for the 2021 censuses comply with EU requirements and UK legislation.

Membership

  • National Statistician and Chair

  • Deputy National Statistician for Population and Public Policy

  • Population and Public Policy Transformation Director, Office for National Statistics (ONS)

  • Deputy Director, Statistical Design and Research, ONS

  • Head of Statistical Design, ONS

  • Head of Question and Questionnaire Design

  • Chief Executive and Registrar General, National Records of Scotland (NRS)

  • Chief Executive and Registrar General, Northern Ireland Statistics and Research Agency (NISRA)

  • Head of Census, NISRA

  • Director of Sources, NISRA

  • Census Director Scotland

  • Chief Statistician, Scottish Government

  • Chief Statistician, Welsh Government

  • Secretariat, Central Policy Secretariat, UK Statistics Authority

Meetings

All meetings are to be held on the same day as the Inter-Administration Committee (IAC) rotating around the four countries.

All papers will be circulated to full members and those who attend periodically.

Back to table of contents

13. Appendix 6: Terms of Reference for the Census Research Advisory Group (CRAG)

Purpose and scope

There will be three purposes of the Census Research Advisory Group (CRAG):

  • to provide advice on methodology

  • to provide assurance on methodology

  • to evaluate items for the external review

CRAG will be a supportive group whereby teams from across the office can bring their work to receive methodological scrutiny. Items brought to CRAG must either be in the initial planning stages or be at the end of the project to approve analysis, findings and conclusions (work can be overseen by a sponsor or working group throughout).

Any specific pieces of work that affect the statistical design of the Admin Data Census or 2021 Census are in scope, including:

  • coverage assessment and adjustment

  • matching

  • edit and imputation

  • statistical disclosure control

  • enhanced outputs

  • use of administrative data

  • processing rules

  • resolving multiple responses

  • quality assurance

  • address register

  • field methodology

  • test design

  • behavioural testing

  • Data Collection Methodology

  • statistical design

  • Census Coverage Survey

  • Census Quality Survey

All items that are brought to CRAG must either have been through a working group or have a sponsor from the list of CRAG specialists. The work must have been reviewed by the sponsor prior to presentation at CRAG. This is to ensure that the items have already been reviewed internally and when they come to CRAG they are refined.

Working groups include:

  • Question and Questionnaire Design (QQD) topic specialist groups

  • International migration working group

  • Income working group

  • Edit and imputation working group

  • Census Coverage Survey (CCS) working group

  • Maximising response working group

  • Coverage working group

  • Design Assurance group

  • Design working group

Areas that do not have a working group:

  • Data linkage

Although CRAG will not provide “final sign off” on items, items will receive approval when they are complete so that across the office they can be recognised as having come under adequate methodological critique. Due to this, teams who bring items to the group should take on board the recommendations CRAG make before approval is given.

Membership

There will be five core Office for National Statistics (ONS) members who will be required to attend each session. Each of the devolved administrations can also have one core representative.

These core members will be:

  • Methodology Census lead

  • Southampton University

  • Head of Statistical Design (including integration of alternative data across 2021 CSOD and coordinating 2021 input into the independent methodological review)

  • Admin data census in methodology

  • Admin Data Census Division

  • Welsh Government

  • Northern Ireland Statistics and Research Agency (NISRA)

  • National Records of Scotland (NRS)

There is also a chair, facilitator and secretariat.

The facilitator will hold a list of topic specialists who will be invited depending on the items being brought to the session. These will include:

  • Head of Census Statistical Outputs and Design

  • Statistical disclosure control (SDC)

  • Demographer, population statistics division

  • Collecting and editing methods and statistical computing

  • Head of migration delivery, migration statistics division

  • Head of Population Statistics Division

  • Geography definitions and coding

  • coverage, Census Coverage Survey (CCS) fieldwork, sample design, estimation, adjustment, census processing, data science

  • Demographic methods for stats relating to births, deaths, internal migration, international migration, special populations, population models, projections, ethnicity, use of administrative data for population statistics and measuring statistical uncertainty

  • Census edit and imputation

  • DD collecting and editing methods and statistical computing

  • Head of address register

  • Data linkage

  • Address register

  • Small area estimation

  • Census coverage assessment, sample design and estimation, population size estimation and record linkage)

  • Questionnaire design (NISRA)

  • Data processing (NISRA)

  • SDC, dissemination and outputs (NISRA)

  • Admin Data linkage and sources (NISRA)

  • Address Register (NISRA)

  • Questionnaire design (NRS)

  • Statistical processing and related methods (NRS)

  • SDC and dissemination and outputs (NRS)

  • Maximising response and field modelling (NRS)

  • Admin Data linkage and census alternatives (NRS)

  • Enumeration (NRS)

Functions of members

The Chair will:

  • provide clear leadership and overall direction for the meeting

  • escalate unresolved issues to the Design Assurance Group, when necessary

The facilitator will:

  • make all necessary advance preparations for meetings and be the point of contact for the group

  • agree the agenda with the Chair and circulate it to members one week prior to each meeting (along with any associated papers)

  • circulate the minutes

The secretary will:

  • take notes of the main points of the meeting, recording clearly any actions and the name of the person responsible for each

  • produce a draft of the minutes and get sign-off from the Chair

The core members will:

  • provide focused and informed recommendations for work

  • critique methodology of items brought to CRAG

  • work collaboratively with other core members and specialists to provide useful advice and assurance

The CRAG specialists will:

  • attend meetings where specialist knowledge is required

  • working collaboratively with other core members and specialists to provide useful advice and assurance

  • contribute papers to CRAG

Operations

The group will meet face-to-face on a monthly basis for two hours. Three methodology topics will be discussed at most in each meeting through a paper and a presentation. Every item brought to CRAG must be accompanied by a paper and a presentation, which will be circulated to members one week in advance of each monthly meeting.

Back to table of contents

14. Appendix 7: Terms of Reference for the Census Analysis Working Group

Purpose

The Census Analysis Working Group will ensure a collaborative and coherent approach to the production and dissemination of 2021 Census analysis and related outputs, in line with wider activity across the Census and Data Collection Transformation Programme (CDCTP). Specifically, it will:

  • agree and implement a stakeholder engagement plan to ensure that activities and messages are aligned

  • agree a flexible analysis plan that is responsive to stakeholder priorities and user needs

  • design and publish analysis outputs that aim to achieve public and policy impact

Scope

The scope of the Census Analysis Working Group covers:

  • production and dissemination of 2021 Census analysis outputs for England and Wales and for the UK

  • identification of opportunities for collaborating with external partners (including the devolved administrations) and working with them to deliver 2021 Census analysis outputs

Responsibilities

The Census Analysis Working Group is a forum for co-ordinating activity to ensure that 2021 Census and related analysis outputs meet our strategic aims.

The Census Analysis Working Group will:

  • agree the resourcing requirements and structure for delivering cross-cutting analysis across topics and sites

  • develop a stakeholder engagement plan and keep a record of all stakeholder activity and outcomes

  • agree a flexible analysis plan based on stakeholder requirements and priorities, and commitments made in the Census White Paper

  • provide topic expertise for the quality assurance of census data and estimates

  • agree design and dissemination principles to be applied across the 2021 Census and related analysis outputs

  • agree quality assurance processes for ensuring accuracy and consistency across the 2021 Census and related analysis outputs

  • agree mechanisms for measuring and recording the impact of the 2021 Census and related analysis outputs for benefits realisation

  • report progress and escalate issues to the Public Policy Impact Board

Membership

The Census Analysis Working Group will have the following standing members:

  • Head of Census Analysis Coordination (Chair)

  • Public Policy Analysis Directorate (PPA) Census Topic Leads

  • Migration Statistics Transformation

  • Population Statistics Transformation

  • Ageing

  • Equalities and Inclusion

  • Sustainable Development Goals

  • Question and Questionnaire Design

  • Data Processing

  • 2021 Census Outputs & Dissemination

  • Quality Assurance of Census Data

  • Statistical Design and Research Division (SDR) Topic Leads (as required)

  • SDR Stakeholders and Communications

  • Population and Public Policy (PPP) Strategic Stakeholder and Communications Coordinator

  • Stakeholder Engagement and Communications

  • ONS Publishing

  • Harmonisation

  • Programme Delivery Division

  • National Records of Scotland

  • Northern Ireland Statistics and Research Agency

  • Welsh Government

  • Census Analysis Coordination (Secretariat)

Population and Public Policy governance structure

Figure 1 shows the governance hierarchy for the Census and Data Collection Transformation Programme (CDCTP) (as at May 2019). The Census Analysis Working Group sits below the Public Policy Impact Board.

Operational protocols

The Census Analysis Working Group will be held every two months.

The Census Analysis Co-ordination Team will provide secretariat support. An agenda and papers will be distributed by email in advance, no less than three working days beforehand unless there are exceptional circumstances. Actions will be circulated within one working week of the board unless there are exceptional circumstances.

Members unable to attend a meeting will send a suitable deputy who is empowered to make decisions on their behalf or will contribute by correspondence.

In place of a meeting, some decisions may be made by correspondence, particularly if a discussion or debate is not required.

Briefing papers may occasionally be circulated by the Census Analysis Co-ordination Team to the group in between meetings to keep members up to date.

Standing agenda items include:

  • minutes and actions from previous meetings

  • update from Census Analysis Co-ordination Team

  • update from Outputs and Dissemination Team

  • updates from PPA Census Analysis topic leads

  • any other business and future agenda items

Back to table of contents