General
What is a census?
A census provides an estimate of the population and housing of Northern Ireland and is held every 10 years.
The most recent census took place on Sunday 21 March 2021. Everyone was asked the same questions to ensure a full snapshot of the population. This information is used to estimate the number of people and households in each area, and their characteristics.
Who runs the census?
The census is organised by Census Office within the Northern Ireland Statistics and Research Agency (NISRA).
Why do we have a census?
By knowing how many people live in Northern Ireland, it helps to determine what services are needed now and in the future.
The government uses census statistics to allocate funds, decide future policy and plan important services such as:
• population – knowing how many people live in an area helps central government allocate funding. Local government also use census data to plan who needs what in their local areas;
• health and disability – health services and policies are planned around the census data so that the necessary services are provided to those who need them;
• housing – housing needs can be much better planned if both the current and future demand is known;
• employment – by establishing how many people work in different occupations and industries, census information can be used to help plan jobs and training policies;
• equality characteristics – census information can be used to help allocate resources and monitor policies to ensure that all groups are treated equally;
• transport – identifying how and where people travel to work and study will help identify the pressures on transport systems and improve planning for roads and public transport.
Is the census compulsory?
As the results are used to inform the allocation of public money and developing public policy, completing the census is required by law.
How successful was Census 2021?
Census 2021 was delivered on time and on budget.
The quality of the outputs has been judged against a quality framework. This is detailed in the Census 2021 quality assurance report and the statement about data quality.
Outputs from Census 2021 have been designated as National Statistics by the Office for Statistics Regulation.
What was the response rate?
In Census 2021, responses were received from 97% of households in Northern Ireland – this is the highest response rate of the last three censuses (2001, 2011 and 2021). In addition, the high level of online responses has ensured that the results are of a higher quality than ever before.
In line with censuses around the world, Census Office has used detailed statistical methods to ensure all census outputs reflect the full population.
How did you decide which questions to ask?
Prior to the census, cases were made for specific topics by government departments, key census users and the general public. These cases were balanced against public acceptability, value for money and whether or not the questions could be asked in a way that gives reliable information.
The questions asked in Census 2021 were approved by the Northern Ireland Assembly in the Census Order (Northern Ireland) 2020.
Census 2021 included questions about the household as a whole and about each person usually resident in the household.
How many questions were there on the questionnaire?
There were up to 42 questions for each individual member of the household to complete.
Typically the average respondent needed to answer less than 42 questions and most could be answered with a simple tick.
What is the future for the next census?
Census Office will soon start to develop proposals on the next census. Any proposals will be informed by statistical viability, and active engagement with organisations, user groups and the general public.
The final decision on the next census will be taken by the Northern Ireland Assembly and Executive. It is not expected that this decision would be taken until 2024 at the earliest.
Were the census questions in Northern Ireland, Scotland and England and Wales exactly the same?
The majority of census questions in Northern Ireland, Scotland and England and Wales were exactly the same.
Can the statistics from the 2011 and 2021 Censuses be compared?
The questionnaires from the 2011 and 2021 Censuses were largely unchanged, and accordingly most outputs can be compared over time. More details on comparability can be found on the NISRA website .
How precise are the census estimates?
The census aims to measure the size of the population at a particular point in time - 21 March 2021. While the detailed statistical tables from the census provide population figures that are published as exact numbers, they should be regarded as estimates of the true population. As an example, it is acknowledged that some people may be missed or counted in the wrong place . The key demographic characteristics of those who are missed are determined through a coverage assessment and adjustment process, which provides estimates by area, age and sex of this group.
Further details can be found in the statement about data quality.
If the census figures are estimates, what about comparing figures from different censuses?
The census is subject to both sampling and non-sampling error. In the census, a coverage survey is used to estimate the number of people who are missed through the enumeration process along with their age and sex characteristics. Quantifying measurement error in the census in a routine manner, for example scanning, coding and capture errors, is difficult because of the self-assessment nature of the census and the various complex and inter-related methodological processes involved.
While Census Office has provided population figures that are published as exact numbers of people, all figures in the Statistics Press Notices and Statistics Bulletins have been presented in a rounded form to (a) stress the estimated nature of the census figures and (b) ease readability.
The Statistics Bulletins and Statistics Press Notices provide comparisons across time and across geographic units. Given the difficulty in assessing the non-sampling measurement error associated with such comparisons, the concept of using statistical significance (for example whether a difference is significant at the 5% level) to quantify and help interpret any differences has not been used in these documents. Accordingly, the Statistics Bulletins and Statistics Press Notices use words like “increases” and “less than” without regard to statistical significance. Clearly, the importance of any differences will be proportionate to their size.
Where can I find definitions and variables for Census 2021 tables?
The Census 2021 outputs definitions document provides definitions for terms used in the preparation and publication of statistics from Census 2021 in Northern Ireland. The Census 2021 variables index details each census variable and contains information such as the variable definition and categories. Both documents can be found on the NISRA website.
Where can I find footnotes for Census 2021 tables?
All table footnotes are contained within the downloadable spreadsheets for each census output table.
Collecting the information and quality assurance
What has NISRA done to improve accuracy this time around?
The complete census operation was designed with a key objective of ensuring the quality of the census results. Some of the ways this was achieved include:
• development of a comprehensive address register – this is a key input to ensuring that everyone had the opportunity to take part;
• running a detailed public awareness campaign to ensure the public were aware of the census and the importance attached to filling out the questionnaire;
• issuing a comprehensive series of reminder letters and conducting field visits to ensure that those who had not filled out and returned their questionnaire were reminded to do so;
• extensive community and voluntary sector engagement – Census Office engaged with communities and the voluntary sector to promote the value of the census. This included practical support to help facilitate responses among voluntary and community groups;
• providing online completion – the ability to fill in the census questionnaire online made it easier for people to respond and also improved the accuracy of the data collected;
• a larger and more focused Census Coverage Survey (CCS) sample - Census Office ran a larger CCS in 2021 than in 2011. This improved the accuracy of the resultant population and household estimates; and
• improved quality assurance methods – NISRA improved the quality assurance methods to take account of new data sources (for example the address register), engaged with local authorities ahead of the census specifically to enhance and validate the address register, and involved the engagement of external independent experts to review the application of these methods.
What is an address register?
An address register for a census is a list of all occupied (or occupiable) households and managed accommodation in the country.
The address register for Census 2021 was created to make sure that every household and managed accommodation in Northern Ireland received a census questionnaire.
NISRA developed the register primarily from the POINTER address database held by the Land and Property Services (LPS).
The register also included additional address information or intelligence from local Planning Offices, Northern Ireland Electricity and a host of other administrative data sources.
What is a Census Coverage Survey (CCS)?
The Census Coverage Survey (CCS) is an independent face to face interview carried out after the census, which asks some of the same questions included in the census. This survey is used to assess coverage and make sure the final census estimates are based on as complete a population as possible. NISRA carried out doorstep interviews from 12 May to 29 June 2021.
The people counted in the CCS are compared with the census so that estimates can be made of the people not included on a census questionnaire.
What percentage of people filled in their questionnaire online and was this percentage in line with NISRA expectations?
In total, 80.6% of household returns were made online. This was in line with NISRA expectations.
Why should I believe the census numbers?
There is no other data source that can provide the level of information that the census does. Census outputs are subjected to rigorous quality assurance checks as described in the Census 2021 quality assurance report.
In Census 2021, responses were received from 97% of households in Northern Ireland - this is the highest response rate of the last three censuses (2001, 2011 and 2021). In addition the high level of online responses has ensured that the results are of a higher quality than ever before.
In line with censuses around the world, Census Office has used detailed statistical methods to ensure all census outputs reflect the full population.
Surely the census figures are already over a year out of date?
It takes time to analyse and quality assure the data and Census Office needs to strike the proper balance between timeliness and quality. Key users are content that plans in the Census 2021 output prospectus are adequate to feed into decisions on resource allocation.
How did you deal with diverse communities?
Prior to Census 2021, NISRA carried out a number of community liaison events to increase awareness of the census among community groups and the voluntary sector.
These events focused on the benefits of the census. They also highlighted where and how people could get information and assistance with completing their census questionnaire.
What new questions were introduced in Census 2021?
For Census 2021 there was a new household question on renewable energy systems and, for those aged 16 and over, a new individual question on sexual orientation was included. (This question did not have a penalty if the respondent did not wish to answer and a tick box was included for those who preferred not to say).
There were also a number of amendments to existing questions, these were made to help respondents answer, and these changes included improved question layouts and updated response categories based upon administrative data or advice from topic experts.
Changes were tested in a large-scale census rehearsal in 2019. Further details can be found in the 2021-2011 Census questionnaire comparability report.
What questions were removed in Census 2021?
A number of questions were removed, these were:
- the household questions on number of rooms and whether the accommodation was self-contained; and
- individual questions on short-term migration and voluntary work.
In all cases there are alternative sources for this data.
Further details can be found in the 2021-2011 Census questionnaire comparability report.
What is a Census Quality Survey (CQS)?
The Census Quality Survey (CQS) is conducted independently of the 2021 Census and involves a random sample of households that have previously responded to the census. In October 2021, trained interviewers from NISRA visited households who had returned census forms and re-asked a reduced set of census questions. The responses were then matched to, and compared with, those provided in Census 2021. The Census Quality Survey (CQS) report provides the results of this comparison along with insight into whether people fully understood the questions on their census form. This then helps Census Office to better understand the quality of the statistics produced.
Processing the information
How do you deal with any under- or over-enumeration?
As with any census, there will always be a small percentage of people who will have been missed or counted twice (for example children of separated or divorced parents). In addition some enumerated people may have missed out some questions.
NISRA has many procedures in place to deal with this, including the Census Coverage Survey and quality assurance of the results. So while a census collection provides the basis for the statistics, the final Census 2021 results include small adjustments to provide a final overall best population estimate.
Were there any processing delays with Census 2021?
None, that weren’t overcome. There was a successful field operation for Census 2021 and return rates for both the census and the Census Coverage Survey were excellent.
The data was captured, coded and loaded into census systems for statistical processing. Statistical processing and its systems and methods worked effectively and as planned.
Publishing the results and accessing the data
When will the results be published?
The first results from Census 2021 were released on 24 May 2022. Please see the Census 2021 outputs prospectus for more information about later releases.
What is the timetable for census outputs?
The first census results were issued on 24 May 2022. Please see the Census 2021 outputs prospectus for more information about later releases.
When did you announce the first results?
NISRA issued the first release on 24 May 2022. Please see the Census 2021 outputs prospectus for more information about later releases.
When will dates become more specific?
The first census results were published on 24 May 2022. NISRA is planning for several releases of data and have published details of time periods for each release.
Please see the Census 2021 outputs prospectus for more information about later releases.
Why do you release the data in batches? Can’t you produce all the census data in one go, and why do there have to be phased releases?
Census statistics are used in a variety of areas and work is required to code and process the data to make it suitable for dissemination. Work is also required to ensure that the statistics produced are anonymous. This takes time.
When completed there will be a vast range of data to be released from the census. As an example, in the 2011 Census, there were one million tables produced.
More details of the Census 2021 release schedule are available in the Census 2021 outputs prospectus.
When will you publish cross-tabulations of the census variables?
Further releases will include cross-tabulations of census variables. Please see the Census 2021 outputs prospectus for more information about later releases.
If you find a problem, will you republish them?
Census Office is very confident that the published statistics are accurate.
Over the next year Census Office will publish a large number of statistics from Census 2021. In line with the Code of Practice, NISRA has a revision policy that will be used if any statistics are published in error.
How close were the intercensal mid-year population estimates to the 2021 Census population estimate?
Census Office commissioned the NISRA population statistics team to produce an ad-hoc estimate of the population for March 2021 by rolling forward the 2020 mid-year population estimate for Northern Ireland. As of March 2021, it was estimated that the population of Northern Ireland was 1,898,600. This estimate fell within 0.2% (4,500 people) of the 2021 Census population estimate, 1,903,100.
Note: you can find our full mid-year population estimates series on the NISRA website.
When were mid-year population estimates (MYEs) for 2021 published?
Mid-year population estimates for 2021, which are rolled forward from census day (21 March 2021) and relate to the population as of 30 June 2021, were published on 29 November 2022.
When will mid-year population estimates (MYEs) for 2022 be published?
MYEs are usually published around June each year and are one year in arrears. It is planned that MYEs for 2022 will be published in mid-2023.Further information on timing can be found on the NISRA website.
Why does the census include a question on religion brought up in? Why do census outputs include separate outputs for religion and religion or religion brought up in?
Census 2021 contained two religion questions, the first on current religion and a follow-up question on religion of upbringing.
There are robust justifications for the religion questions related to equality obligations under the Northern Ireland Act (1998) and the Fair Employment (Northern Ireland) Order 1998 and the need to provide information on religion to help plan public services. The 2021 and earlier census religion statistics are used to provide data to support employment monitoring and have the support of the Equality Commission and key users of census data. This support was reflected in the public consultation.
A more detailed guidance note on the religion and religion brought up in questions in Census 2021 is available on the NISRA website.
As the Local Government District (LGD) boundaries changed in 2014, can we compare LGD statistics from the 2011 and 2021 Censuses?
NISRA published Census 2011 Key Statistics for the new LGDs (LGD2014) on 30 January 2014 therefore census outputs for 2011 and 2021 can be compared for these geographic units.
What are the arrangements to produce a combined population figure for the UK?
NISRA is responsible for population figures for Northern Ireland only. The Office for National Statistics (ONS) and National Records of Scotland (NRS) are responsible for delivering population figures for England and Wales, and Scotland respectively.
Overall UK figures (i.e. data for England and Wales, Scotland and Northern Ireland) will be released once all the countries have completed their processing and are confident that their data is of the best possible quality. More information can be found on the NISRA website.
As a member of the general public, can I get information from the census?
Yes. NISRA is keen to make the data as accessible as possible to a wide range of users, from those in local government districts through to those in business, commerce, the voluntary sector and the general public. However, no one has access to individual census returns.
NISRA will be providing access to census statistics via its website to:
• maximise the use of census statistics; and
• provide enhanced tools to explore census statistics more effectively.
Flexible Table Builder
Why can I not get the data I want?
The Flexible Table Builder provides a function that queries the entire Census 2021 database. Therefore it is important to protect the individual records from disclosure, and this is done by evaluating each query on an area by area basis against a set of pre-defined rules. If your query fails one or more of these rules the data will be withheld and a message will be shown in the builder. For more information on how the Statistical Disclosure Control works, see the “Improving your results” help video.
Why are some variables not available?
Variables for all census topics are available in the Flexible Table Builder. If a variable you are interested in does not appear in the builder, you can email Census Office for further information.
Where can I access 2011 Census information? Is it available in the Flexible Table Builder?
Information on where to find 2011 Census data is available on the NISRA website.
What is the difference between a custom table and a pre-built table?
Custom tables are created by the user of the Flexible Table Builder by selecting datasets, geographic levels and variables of interest. Pre-built tables have already been designed for the user – either because they mirror tables that have already been released from Census 2021, or they are commonly downloaded table specifications from previous censuses. They can be altered and customised further within the Flexible Table Builder should users wish to do so.
Why do I only get one column of data, when I wanted a table?
It is possible to turn the column of data into a table by pivoting the results. More information is available in the “Creating a basic table” help video.
My table has codes that I don’t want – how do I remove them?
Once the data has been downloaded, the columns of codes can simply be removed by deleting the column within your data editing software. Alternatively, if you pivot the data within the Builder, the codes (except geography codes) are automatically removed. For information on how to pivot the data see the “Creating a basic table” help video.
Why do you not provide totals or percentages in tables? Can I get charts and graphs in this system?
The Flexible Table Builder is not designed to calculate sub-totals/rates/proportions etc. or chart the data. It provides counts for combinations of variable classifications that can be viewed or downloaded. Users can manipulate the data once it is downloaded by calculating their own percentages or building charts and graphs. The Census Area Explorer provides some charts and graphs, including area maps, for some pre-defined census outputs.
Not all Settlements (2015) are available in the Flexible Table Builder, which settlements are available? When will the other settlements be available?
The list of settlements currently available in the Flexible Table Builder can be viewed in the Settlement 2015 metadata. These settlements are those from the 2015 Review of the Statistical Classification and Delineation of Settlements that can be accurately approximated by aggregating the Census 2021 Data Zones. More information on Census 2021 Data Zones is available at Census 2021 output geography information papers.
For the remaining smaller settlements an alternative method to produce statistics is being considered this is likely to derive from a grid square aggregation method, and is being assessed as part of Census 2021 Grid Square product development.
How should I reference results or figures from the Flexible Table Builder?
If you wish to reference figures from the Flexible Table Builder we suggest that you cite ‘Census 2021’ as the source, and include a hyperlink to the table created in the FTB.
Confidentiality and security
Is it safe, secure and confidential?
Yes. Census Office has made it a top priority to keep your census data secure and confidential.
Census data is protected by law and will not leave the UK. It is actually a criminal offence to disclose personal census data, punishable by a fine and/or up to two years in prison. Everyone working with personal census data is security checked and must sign an undertaking that they will protect the privacy of the information.
Census Office asked independent experts to conduct an information assurance review of the 2021 Census. The review team concluded that “both ONS and NISRA have comprehensive security programmes in place designed to reduce the risk of compromise to the delivery of the census and citizen data”.
Will individuals or households be identifiable from published statistics?
Census Office has taken steps to ensure that the confidentiality of respondents is fully protected. All published results from the census have been subject to statistical processes to ensure that individuals cannot be identified. For more information, please refer to the statistical disclosure control methodology. These processes may result in very marginally different results between tables for the same statistic.
What is statistical disclosure control (SDC)?
Statistical disclosure control (SDC) refers to a range of methods that aim to protect individuals, households, businesses, and their attributes from being identified in published information.
For Census 2021, NISRA is applying two strategies - targeted record swapping (TRS) and cell key perturbation (CKP), to ensure individuals are protected from identification while minimising the impact on the quality of results.
For more information, please refer to the statistical disclosure control methodology.
National Statistics
What does it mean that census outputs are designated as National Statistics?
Outputs from Census 2021 have been designated as National Statistics by the Office for Statistics Regulation. The Regulator has assessed Census Office procedures against the Code of Practice for Statistics and confirmed that Census Office is adhering to the Code. The Regulator has published two assessment reports (numbers 345 and 362) about Census 2021.
The outputs prospectus provides dates on which new census statistics are to be published. Can anyone see the statistics prior to publication?
Census outputs are subject to The Pre-release Access to Official Statistics (Northern Ireland) 2009 Order . In line with this legislation, pre-release access to Census 2021 statistics has not been granted to anyone. Census results are produced impartially and free from political influence.
Geography
Do 2021 Data Zones (DZ) match 2011 Small Areas (SA)?
Data Zones and Super Data Zones are a newly created geographies to support the dissemination of Census 2021 statistics. The new areas are different from the Small Areas and Super Output Areas used to disseminate 2011 Census statistics. Further information on the rationale for this change and how these new Data Zones/Super Data Zones were created can be found on the NISRA website.
Usual resident & household counts have been released for the new statistical output geographies, when will the other data be released at that level?
Future releases will include tables for the new Census 2021 statistical output geographies – the full set of Census 2021 output geographies will also be included in the new flexible table builder. Please see the Census 2021 outputs prospectus for more information about later releases.
When I add up data for lower geographies I don’t get the same results as the higher geographies – why is that?
Census information released by NISRA is subjected to disclosure control. This process adds ‘noise’ to the data to protect against the disclosure of information on individuals – this is described in more detail in our guidance note. The ‘noise’ introduced can be enhanced by aggregation so users should limit the extent of aggregation and source the data from the largest geographic area available.
If the required data is unavailable and users must aggregate census data to fulfil their needs, the guidance is to aggregate from the largest base geography. For example, if a user wanted to create an area that was equivalent to the Ards & North Down Local Government District (LGD) plus a small area of East Belfast, it would be more accurate to add the figures for Ards & North Down to the relevant figures for the Data Zones that approximate the part of East Belfast required, rather than aggregate the figures for all Data Zones in the defined area.
Will Census 2021 data be available for older geographies (e.g. Electoral Wards, Super Output Areas or Small Areas)?
The Census 2021 output geography hierarchy beneath Local Government District (LGD) is District Electoral Areas (80 areas), new Census 2021 Super Data Zones (850 areas) and Data Zones (3780 areas). Census 2021 statistics will also be released for further geographic levels, these will be limited to geographies that can be accurately approximated using Data Zones (as the smallest level building block). Further information on these geographies is available on the NISRA website.
Census Office plans to produce a limited 100m and 1km grid square dataset for Census 2021. As part of this Census Office will make available geographic lookup tables that will support users who wish to aggregate grid square data for other geographies (e.g. 2011 Census Super Output Areas) outside of those that can be approximated from Data Zones. NISRA will keep users informed of developments in this area via the release plans pages on our website.
How can I compare 2021 statistics for Super Data Zone and Data Zone geographies with statistics from previous censuses?
In line with the 2011 and earlier Censuses, NISRA plans to produce a limited 100m and 1km grid square dataset for Census 2021. As part of this, Census Office will make available geographic lookup tables that will support users who wish to aggregate grid square data from the 2021 or earlier Censuses.
NISRA will keep users informed of developments in this area via the release plans pages on our website.
When will the settlements be reviewed?
The 2015 Review of the Statistical Classification and Delineation of Settlements was commissioned by the Statistics Co-ordinating Group (SCG) of the Northern Ireland Civil Service (NICS), and was undertaken by a representative cross-Departmental sub-group. While Census 2021 data would likely be used in any future review, Census Office have not received any request for input or any indication when another review of the settlement classifications will take place; on that basis, we continue to use the 2015-based classification.