2021 Census: Ethnic Group Stakeholder Follow-up Survey – Summary of responses

Between June and August 2015, the Office for National Statistics (ONS) conducted a public consultation on topics for the 2021 Census. This consultation confirmed a need to continue to collect information on ethnic group and highlighted areas where specific needs may not be met by the current question. As a result, the ONS ran an "Ethnic Group Stakeholder Follow-Up Survey" between November 2016 and January 2017. The aim of this survey was to gain a deeper understanding of user need for ethnic group information.

The 2021 Census evaluation criteria were set out in the May 2016 census topic consultation report (PDF, 796.4KB). The Ethnic Group Follow-up Survey provided evidence that helped the ONS evaluate the strength of user need for the data, the comparability of data over time, and the public acceptability of terminology being used in this question.

Recommendations put forward in the White Paper, The 2021 Census of Population and Housing in England and Wales, are based on extensive research and assessment against the criteria.

The work aimed to:

• find out who is answering the survey or interested in ethnic group data

• find out how ethnic group data are used by organisations and individuals

• give the opportunity to provide more information to help ONS maintain or improve the relevance of ethnic group

• inform how we collect information about ethnic groups in 2011

Table 1 shows the number of responses received, including the sectors of responding organisations. A list of all the responding organisations can be found in Annex A.

The ONS received 132 responses to the follow-up survey, including 46 responses from organisations and 86 responses from individuals. Responding organisations were from a range of sectors, including local authorities, charity and voluntary organisations, government departments and public bodies (shown in Table 1). The survey gave the opportunity to provide more information to help the ONS maintain or improve the relevance of the ethnic group data outputs, which will inform how we collect information about ethnic groups in 2021.

Overall, the survey found that most data users require general information on the ethnic composition of the population, and that they are most likely to require data at a local authority level geography.

Most responders of the survey found the terms used in the 2011 Census question acceptable. Those who found them unacceptable gave reasons such as finding them confusing and terms being too broad.

The survey asked respondents what information on specific ethnic groups they require and other requirements they have for new ethnic group information. The answers included suggestions for changes such as:

• adding new response options

• changing existing response options

• producing new ethnic group outputs

• changing the design of the ethnic group question

Requests from the survey, in combination with responses from "The 2021 Census – Assessment of initial user requirements on content for England and Wales", identified requests for 55 new ethnic group tick boxes to be added to the 2021 ethnic group question.

Additional census information requested would be used for a variety of reasons. The most commonly stated reason given by organisations was for policy development, and for individuals it was most likely to be used for academic and other research.

Data requirements were also considered, for topics including national identity, language and religion. This was asked as the ONS aimed to investigate how the ethnic group question works alongside the suite of identity questions (religion, national identity and language). Responses highlighted the need for comparability and stability of response options, with changes requested to existing response options and for new subtopics.

The uses of commissioned or ad hoc tables were asked about within the survey and it was found that 20% of respondents had previously used the service. Organisations found that the tables met their requirements more than individuals. It was also noted that data users would rather integrate the data themselves. The ONS is exploring the options to fulfil this need.

The census takes place every 10 years and provides vital information on the population of England and Wales. The census is a compulsory exercise carried out on a household enumeration basis; each respondent is required to complete all relevant questions on the questionnaire except the question on religion. As such it is important that there is a clear basis for determining whether topics are included. This involves evaluation against our publicly stated criteria, which are set out in the May 2016 census topic consultation report.

Between June and August 2015, the Office for National Statistics (ONS) conducted a public consultation on topics for the 2021 Census. This consultation confirmed a need to continue to collect information on ethnic group and highlighted areas where specific needs may not be met by the current question.

Responses from the public consultation confirmed a strong need to continue to collect information on the topic of ethnic group, and confirmed the user need, which included:

• resource allocation by central and local government

• to inform policy development

• to help organisations meet and monitor their statutory obligations under the Equality Act 2010

Following this, the ONS ran an "Ethnic Group Stakeholder Follow-Up Survey" between November 2016 and January 2017 to gain a deeper understanding of user needs for ethnic group information. The responses, alongside research and other evidence, contributed to the evidence to inform the ONS recommendations on how we collect information about ethnic groups in 2021.

The ONS sent the survey to stakeholders who had responded to the "Ethnicity and national identity" topic in the 2021 Census topic consultation. Details of the survey were also published on the ONS website, so that anyone could respond. Respondents could answer online or on paper via post or email.

The survey received 132 responses, with 86 responses from individuals and 46 responses from organisations. This report aims to provide a summary of the responses received to the "Ethnic Group Stakeholder Follow-up Survey". This is referred to as "the survey" throughout this report.

Some organisations or individuals submitted more than one response, with varying content.

As the main aim of the survey was to gain a deeper understanding of user need for ethnic group information, all responses have been included in the analysis.

The survey asked about how individuals and organisations use ethnic group information.

The survey asked about seven themes relating to ethnic group:

• ethnic group data requirements

• suitability of the 2011 Census ethnic group categories for the 2021 Census

• uses of ethnic group data

• acceptability of ethnic group terminology

• suggested changes to the ethnic group question

• the impact of ethnic group question changes on data collectors

• requirements for comparability over time and harmonisation with other censuses and surveys

Ethnic group data requirements

Most data users (79%) require general information on the ethnic composition of the population, with 91% of organisations and 72% of individuals requiring this information.

Need for ethnic group information to be compared with other census data

Several respondents specified a need for ethnic group information to be compared with other census data. This need included continued comparisons with census topics such as:

-   housing

-   education

-   health

-   occupation and employment

-   industry

Several respondents also stated that they needed cross-tabulations of ethnic group information, with other census topics, at low-level geographies.

Further breakdown of ethnic group information by age groups

Several users stated that they needed ethnic group information to be broken down into more detailed age groups. For example:

"For some service-provision there is a need for more detailed data on numbers of people by ethnic identity by age and sex. One recent example has been a desire to identify the number of young females at risk of Female Genital Mutilation, which has not been possible from the detailed Census ethnic identity data (or from country of birth). Due to the nature of these needs it is not necessarily possible to identify what specific breakdowns/detail will be required/sought in future. In general the more detailed the breakdown available (more detailed ethnic groupings, age ranges, and by gender) the greater the likelihood that Census outputs will meet the service need." (Oxfordshire County Council)

Suitability of the 2011 Census ethnic group categories for the 2021 Census

The survey asked whether the 2011 Census category outputs would provide the information that users require from the 2021 Census in England and Wales, to find out whether the current combined ethnic group and single ethnic group categories meet the needs of users for the 2021 Census. The questions were asked to those respondents who indicated that they use the three main ethnic group output types:

-   combined ethnic group output – these outputs are the five high-level ethnic group categories ("White", "Mixed/multiple", "Asian/Asian British", "Black/African/Caribbean/Black British" and "Other ethnic group")

-   single ethnic group output – these outputs are the 18 low-level ethnic group categories (including ethnic group tick boxes such as "Indian", and "Any other White background")

-   write-in response outputs – these outputs provide more detail about the ethnic groups in the low-level write-in response options

For combined and single ethnic group output categories, individuals were more likely than organisations to say their needs were met.

Just over half of responders said that the higher-level category data would not meet their requirements for 2021 Census data (Table 4).

When considering single ethnic group (Table 5) and write-in response (Table 6) outputs, around one-third of responders did not think these outputs would meet their census data needs in 2021.

Taken together, this evidence suggests that the ethnic group data are not fully meeting some data users' needs.

Uses of ethnic group data

The survey asked respondents how they would use ethnic group information. Respondents were asked how they would use any additional information that they had requested (for example, requests for new ethnic group response options or outputs).

The survey found that organisations and individuals would use the new ethnic group information for multiple purposes. Among organisations, over four-fifths reported that they would use new ethnic group information for policy development, equality monitoring and service planning and delivery. Among individuals, the most common use would be for academic and other research purposes.

Common uses of ethnic group information that respondents reported in the "Other" option included:

• genealogy and family history research

• profiling the local community

• setting representative quotas and sample sizes for research

• monitoring changes over time

• advocating the needs of certain ethnic groups and applying for funding

Additional information was requested under "other purposes" needed for multiple reasons, based around the area of service design, delivery and planning purposes.

Ethnic group data are used at all geographic levels. Half of the organisations who responded used data at the lowest geographical level we asked about, which was output area.

Other geographies noted by respondents included:

• individual level (after 100 years)

• middle and lower layer super output areas

• urban or rural areas

• parish

• postcode

• workplace zones

Acceptability of ethnic group terminology

The survey asked respondents how acceptable they found the terminology used in the 2011 Census ethnic group question in England and Wales. The survey found that the majority of respondents found the terms used acceptable.

Those who found the terms in the 2011 Census ethnic group question for England and Wales unacceptable provided the following reasons:

• terminology is confusing and uses different concepts of ethnicity

• problems with the combined categories

• use of the term "Gypsy"

• ability to only identify as "Gypsy or Irish Traveller" and "White"

• some categories are too broad

• certain specific ethnic groups are not included

• location issues

• concerns about category ordering

These are discussed in more detail in this section.

Terminology is confusing and uses different concepts of ethnicity

Several respondents stated that the terminology used was confusing and combined different ethnicity concepts. For example:

"In the Irish community there is a misunderstanding of ethnicity and the lack of ability to explain this better in the Census in a problem." (Irish in Britain)

Problems with the combined categories

Several respondents stated that they found the combined categories unacceptable, with an individual specifically mentioning the terminology for the "Mixed/multiple" ethnic group:

"For the mixed/multiple ethnic groups, the term 'mixed parentage' or 'mixed ethnic heritage' would be preferable." (Individual)

Use of the term "Gypsy"

Some respondents expressed concern about the use of the term "Gypsy". For example:

"'Gypsy' can be seen as a derogatory word, see Urban Dictionary! It needs to be more specific; people tend to view a gypsy as Irish rather than Romani." (Manchester City Council)

Ability to only identify as "Gypsy or Irish Traveller" and "White"

Concerns were also raised about the ability of non-White groups to identify as "Gypsy or Irish Traveller". For example:

"It assumes all gypsy or travellers are white and or Irish and this may not be the case." (Individual)

Some categories are too broad

Several respondents stated that they found some of the existing ethnic group categories to be too broad. In particular, the "Any other White background, write in" category, "Gypsy or Irish Traveller" category and "English/Welsh/Scottish/Northern Irish/British" category. For example, respondents stated:

"Any other White background, write in. Because it does not enable us to distinguish consistently between European backgrounds for service delivery." (Broxbourne Borough Council)

Certain specific ethnic groups are not included

Some respondents stated that the 2011 Census ethnic group categories were unacceptable as they did not include a specific ethnic group for which information is required. For example:

"Unacceptable because a Cornish category or categories are not included." (Individual)

"There is no Sikh box in the ethnicity question." (Individual)

Location issues

One respondent questioned the location of the "Arab" tick box, which in the 2011 Census was included under the "Other ethnic group" combined category:

"Should Arab be in "other"?" (National Assembly for Wales)

Concerns about category ordering

One respondent raised concerns about the ordering of categories within the ethnic group question stating that:

"Although the terms are acceptable, we maintain that the ordering of the ethnic categories after White should be done in alphabetical order. So the Mixed category should appear after Black rather than as the second category. In addition to it not following alphabetical order, the position of the Mixed category does not even follow numerical order. We also believe that within the Mixed category the sub-sections should be alphabetised." (Runnymede Trust)

Suggested changes to the ethnic group question

The survey asked respondents what information on specific ethnic groups they require and other requirements they have for new ethnic group information. Answers to these questions included suggestions to change the census ethnic group question.

Adding new ethnic group response options

The survey and topic consultation identified requests for 55 new ethnic group tick boxes to be added to the 2021 Census ethnic group question. The full list of requested ethnic groups, including the number of evidenced responses for each requested tick box, is shown in Annex B.

An evaluation tool was developed in 2011, and updated for 2021, to assess the case for information on the requested ethnic groups.

Changing existing ethnic group response options

The survey identified several requests for changes to be made to the existing 2011 Census ethnic group response options. These are detailed in this section.

Breaking down the "English/Welsh/Scottish/Northern Irish/British" tick box

Several respondents suggested breaking down the "English/Welsh/Scottish/Northern Irish/British" tick box into separate categories.

"The very great gains to be had from including separate boxes on the census form for 'English', 'Scottish', 'Welsh' and 'Northern Irish' would far outweigh any slight loss of comparability." (Individual)

Breaking down the "Gypsy or Irish Traveller" tick box

Several respondents requested that the "Gypsy or Irish Traveller" tick box is broken down so that separate information can be collected on these groups.

"Gypsy and Irish Traveller are put together (whereas they are separate identities)." (Cornwall Council)

Breaking down the "Any other White background" write-in option

Several respondents stated that the "Any other White background" write-in option was too broad and should be broken down. For example:

"The category White Other has also grown to a degree that it might be worth segmenting it. French or German respondents within this category have very different outcomes compared to Polish or Romanian respondents." (Runnymede Trust)

Breaking down the "African" tick box

Several respondents felt that the "African" tick box was too broad and needed to be broken down. For example:

"The single ethnic group category Black African is increasingly less useful as an umbrella category. It merges a number of different communities together, who have very different migration trajectories and outcomes. As a result it has become less meaningful. For example educational outcomes for children of Somali heritage or Nigerian heritage are very different, however, under Black African we are not able to understand that nuance...Having a more detailed breakdown of the ethnic category 'Black African' will allow us to map the experiences of some of the most disadvantaged communities in the UK." (Runnymede Trust)

Breaking down the "Asian/Asian British" combined category

Two respondents raised concerns about the "Asian/Asian British" category terminology, stating:

"Asian/Asian British" are two different worlds." (Individual)

Changing the "Mixed/multiple ethnic groups" combined category terminology

A few respondents mentioned concerns regarding the "Mixed/multiple" category terminology. For example:

"Mixed/multiple should just be 'mixed'; being followed by Asian British and Black British may confuse respondents if 'multiple' is used, as Asian and British sounds like it is multiple even though it isn't." (Manchester City Council)

Changing the "Black/African/Caribbean/Black British" category terminology

One respondent commented on the terminology used in the "Black/African/Caribbean/Black British" category, stating the following:

"Black/African/Caribbean/Black British – why the / after Black? If you are White African you might tick here because you are African, and if you are defining the country of origin why do you need the Black before British?" (Manchester City Council)

Producing new ethnic group outputs

Respondents to the survey also requested changes to the outputs produced from the 2011 Census. These requests included:

Breaking down the "Any other White background" write-in output

Several respondents suggested breaking down the "Any other White background" output. Most of these respondents wanted this category broken down into European identities and other White identities. For example:

"Single ethnic group outputs - It would also be useful to further split the 'Any other White background' to 'White European' and 'Any Other White background' in standard outputs." (London Borough of Waltham Forest)

There were also requests for a more detailed output for specific European identities such as Eastern European, with respondents stating:

"The ethnic group classification proposed has the advantage of consistency but it should also be able to reflect the changing ethnic composition of the population - e.g. distinguishing eastern European in White Other." (Individual)

Breaking down write-in outputs into separate countries

One respondent requested more detailed outputs based on the write-in responses. This organisation requested that outputs from several write-in responses are broken down by country. For example:

"Outputs from the detailed write-in responses - would it be possible to have separate categories for all the EU member countries (e.g. Estonia, Latvia, Lithuania instead of a single category for Baltic States)." (London Borough of Waltham Forest)

Need for outputs on non-White Gypsies and Irish Travellers

There was also a requirement for information on non-White Gypsies and Irish Travellers:

"The lack of data for Welsh/ Irish/Gypsy/Irish Traveller etc. in relation to any category other than White provides significant difficulty for us. Given this available data showing Welsh/ Irish/ Gypsy/ Irish Traveller etc. other than White would be of particular use to us." (Tai Pawb)

Changing the design of the ethnic group question

Some respondents suggested changes to the design of the 2011 Census ethnic group question. These changes included:

Allowing non-White groups to identify as "English/Welsh/Scottish/Northern Irish/British"

Several respondents commented on the inability of non-White groups to identify as English, Welsh, Scottish, Northern Irish or British within the ethnicity question. For example:

"There is an issue when there is a nationality option under ethnicity for British (Welsh/Northern Irish etc.). People who are not white must have the same scope to specify that they are for example: black British is one option, but it is not possible to identify as Indian British or Arab British. If "British" nationality is included in the ethnicity questions, there at least needs to be some explanation of the inclusion of some nationality options in the ethnicity question." (National Assembly for Wales)

Incorporating drop-down menus or write-in options into existing response categories

A few respondents suggested that write-in options or drop-down menus could be added to response categories on the ethnic group question. For the 2021 Census, the ONS aims to receive 75% of responses online. One organisation stated:

"We understand, and fully appreciate, that the Census ethnicity category cannot be expanded ad infinitum as it would become unworkable. However, with the next Census being digital by default, this could perhaps offer up new opportunities to collect more detail under some of the broad headings, given the questionnaire will be online. For example, the number of ethnicity 'write-in' categories could be expanded. More specifically, a write in category alongside the Black African category could be added. Alternatively, ONS could perhaps test whether drop down boxes could be added under categories such as Black African and Black Other with a list of pre-defined sub-categories which capture the most commonly cited groups (e.g. Somali, Nigerian), while retaining the 'Other write-in' option for those residents not captured by the pre-defined categories." (London Borough of Tower Hamlets)

Impact of an ethnic group question change on data collectors

Any change of question will have an impact on data collectors and data users.

A total of 74 respondents indicated that they use census ethnic group response options for their own data collection purposes and were asked to specify what impact a change to the ethnic group question would have.

Around one-third of these "data collectors" indicated that they would need to update their own data collection process if the census ethnic group classifications changed. Respondents indicated that changes would have to be made to the classifications and forms they use, survey sample weightings and even a possible need to either re-do surveys or re-evaluate existing data. For example, respondents stated that:

"When we do our own equalities work, we need to be able to compare ourselves against a national standard. If the census altered its groups, we would have to too." (Powys County Council)

Walsall Council mentioned that a change to the ethnic group response options would be negative because it would affect comparisons across different organisations, while the update is taking place:

"We have sought to align our internal ethnicity data collection (e.g. for Equality Impact Assessments) to the 2011 census classifications to ensure consistency and benchmarking. There is always a time lag in getting all service areas/partners to adopt new categories (as there has been between 2001-2011), which means trying to reconcile all data collection results can be challenging. Making any major changes again in 2021 would mean this problem re-emerging."

Some respondents indicated that a change to the census ethnic group classifications would have a small or non-significant impact. In addition, approximately one-fifth of respondents indicated that a change to the census ethnic group classifications would be positive. The most common reasons why a change would be positive included:

• more accurate and granular data being collected allowing better provision of resources and services

• new information becoming available for ethnic groups that had been requested

Overall, data users had mixed views to changes to ethnic group classifications, with 26 respondents having a neutral view.

Requirements for comparability over time and harmonisation with other censuses and surveys

Responses indicate that a small loss of comparability is acceptable; a larger loss would have more impact on data users and collectors. Without options to consider, the benefits and costs cannot be easily communicated to, or assessed by, data users and collectors.

Small loss of comparability

The survey asked respondents about their need for comparable data over time.

When asked what impact a small loss of comparability would have on their use of ethnic group information, responses to this question can be grouped into the following themes:

• little or no impact

• small loss of comparability is acceptable if detail increases

• aggregation up to 2011 Census categories

Little or no impact

Out of all responses to the small loss of comparability question, around one-third stated little or no impact would be caused if there was a small loss of comparability on their use of ethnic group information.

"A small loss of comparability would not have a large impact on the Council's use of ethnic group information as two of Kingston's largest minority populations (Korean and Tamil) are currently categorised under any other Asian background." (The Royal Borough of Kingston upon Thames)

Small loss of comparability is acceptable if detail increases

Many respondents stated they would accept a small loss of comparability if it meant greater detail or granularity was available. For example:

"The effect will depend on whether change increases accuracy or blurs it. For example if the ethnic group category becomes more detailed, the net effect of being able to speak more accurately about the experiences of ethnic minority groups will outweigh the small loss of comparability." (Runnymede Trust)

Aggregation up to 2011 Census categories

Some respondents indicated a small loss of comparability would be acceptable if it was possible to aggregate data up to be comparable to previous censuses. For example:

"It is essential to retain the comparability to the previous Census. Further breakdowns within the broader category should be possible as they can be aggregated to compare to the previous data if needed." (London Borough of Waltham Forest)

Large loss of comparability

Stakeholders were asked what impact a large loss of comparability would have on their use of ethnic group information.

Significant or negative impact

Around one-third of respondents to the question "What would be the effect of a large loss of comparability on your use of ethnic group information?" stated that a large loss of comparability would cause a significant negative impact on their use of ethnic group information. For example:

"We need comparability for service planning, policy development etc, so we would not want any major changes." (Lancashire County Council)

Small impact

Some respondents indicated that even a large loss of comparability would not impact their use of ethnic group information. The majority of these respondents were individuals, for example:

"For the most part, our use of the ethnic group information has been solely of the most recent census year, rather than a comparison over time. While we would prefer to be able to make comparisons over time (and there are situations where this is really useful), I don't think it would have a substantial impact." (Individual)

Requirements for information on national identity, language and religion

As the ethnic group question works together with other identity questions (such as, national identity, language and religious affiliation), the Office for National Statistics (ONS) wished to gather stakeholders' views on these topics.

The survey asked respondents if they had any comments in relation to national identity, language or religion – in particular, any comments relating to response options.

Comments on national identity, language and religion can be grouped into the following themes:

• suggested changes to response options

• need for comparability and stability of response options

• confusion between ethnicity and national identity

Suggested changes to response options

Several respondents requested changes to existing response options for the national identity, language and religion questions. Suggested changes to the national identity question included expanding the response options to include broader geographical terms, such as "Europe".

One respondent also asked for the national identity question to measure strength of identification with national identity:

"On national identity, the question should also ask for the strength of identification with a nation, compared with identification with the UK." (Individual)

Requests for changes to the religion question were also received. One respondent mentioned that the "Christian" response option in the religious affiliation question was too broad and needed to be broken down. This request stated that:

"We have in the past requested the Christian category broken down at least to Catholic and protestant which in parts of Britain can be revealing in relation to equalities." (Irish in Britain)

Another respondent also required the "No religion" category to be broken down to collect information on non-religious beliefs:

"Separation of atheist and agnostic from 'no religion'." (Individual)

Need for comparability and stability of response options

Although the ONS received requests for changes to be made to the national identity, language and religion response options, there was also a need for these options to remain the same:

"Maintaining comparability with the previous census would also be beneficial in these areas." (Equality and Human Rights Commission)

One respondent specifically requested that the language response options remain the same:

"We would like the language response options to remain the same. As already mentioned, Kingston has relatively large Korean and Tamil populations. In the absence of full ethnicity data on these two groups, language data can be used to estimate the sizes of these populations. If the question relates to proficiency in English then we would also like the response options to stay the same." (The Royal Borough of Kingston upon Thames)

Another respondent required stability of the religion response options:

"For my own purposes, and I believe for broader academic and social policy purposes it is very important to maintain the religion (or perhaps what should be called "religion and belief") questions and for standard outputs to produced that cross-related to ethnicity data in particular." (Individual)

Confusion between ethnicity and national identity

Several respondents stated concerns about people confusing ethnicity and nationality. For example:

"Very important that respondents understand that nationality is about nationality and 'ethnicity' about how they define themselves, perhaps how they feel about who they are." (Individual)

Requests for new sub-topics

Some respondents used this survey to request that new sub-topics were added to the census. These requests included:

• languages spoken and languages spoken at home

• first language

• religious activity

• migration status

• inclusion of "Trans"

The 2011 Census collected information on main language. Respondents to the survey requested that information was required for first language, all languages spoken and languages spoken at home. For example:

"Knowing the general language spoken at home can be very helpful, to identify and help any small groups of residents who do not speak English fluently." (Broxbourne Borough Council)

The 2011 Census religion question is a measure of religious affiliation, not a measure of religious practice. Several respondents requested that information on religious practice should be collected in the census. For example:

"The religion response option should really include a question on religious activity or practice, as this would make the data collected on religion more accurate. At present the Census only measures religious affiliation, which: is extremely superficial and arguably misleading as many who indicate affiliation do not practice their religion very much." (Cytun – Churches Together in Wales)

Use of the ONS's commissioned table service

The survey asked respondents about their awareness and experience of using the census commissioned table service offered by the Office for National Statistics (ONS). We produce census commissioned tables, also known as ad hoc tables, in response to user requests for information. They contain combinations of data that are not available in standard publications. Once created, commissioned tables are published for all to use.

The survey found that one-fifth of respondents had used the census commissioned table service. Of those respondents who had used this service, over two-thirds stated that the service had met their data requirements. Organisations were more likely to state the service had met their requirements than individuals.

Over one-third of respondents (36%) found that the tables did not meet their requirements (32% of organisations and 44% of individuals). Reasons why the tables did not meet requirements included:

• respondents would prefer to integrate data themselves

• the capacity of the commissioned tables is too narrow

We are currently exploring methods of allowing respondents to integrate the data themselves.

This list includes organisations that responded to the "Ethnic Group Stakeholder Follow-Up Survey". If multiple responses were received from an organisation, the name only appears once.

Individuals also responded to the survey.

Government department or public body

Equality and Human Rights Commission

National Assembly for Wales

NHS Digital (formerly Health and Social Care Information Centre)

Sport England

Welsh Government

Local authority

Blackpool Council

Bournemouth Borough Council

Bristol City Council

Broxbourne Borough Council

Caerphilly County Borough Council

Chesterfield Borough Council

Cornwall Council

East Riding of Yorkshire Council

Greater London Authority

Hertfordshire County Council

Lancashire County Council

London Borough of Camden

London Borough of Redbridge

London Borough of Tower Hamlets

London Borough of Waltham Forest

Manchester City Council

North Yorkshire County Council

Oldham Council

Oxfordshire County Council

Powys County Council

Royal Borough of Kensington and Chelsea

Royal Borough of Kingston upon Thames

Sandwell Metropolitan Borough Council

Tameside Council

Walsall Council

Housing

Yarlington Housing Group

Academic or research

British Sociological Association

Charity and voluntary

Bewnans Kernow

Cytun (Churches Together in Wales)

Friends, Families and Travellers

Irish in Britain

National Federation of Gypsy Liaison Groups

Runnymede Trust

Tai Pawb

The Muslim Council of Britain

The Sikh Network

Other

Golden Tree Productions

Market Research Society and MRS Census and Geodemographics Group

Mebyon Kernow – the Party for Cornwall

The number of requests for each ethnic group, which has provided appropriate evidence to evaluate, are provided in this section and detailed in brackets.

The Office for National Statistics (ONS) does not consider the number of responses but evaluates the strength of user need to evaluate the case for adding tick boxes.

Number of requests for each ethnic group

Afghan (2)

Breakdown of Black "African" tick box (4)

Breakdown of "Any other Black/African/Caribbean" write in (2)

Breakdown of outputs for EU countries (3)

Breakdown of "Any other ethnic group" write in (1)

Breakdown of White "English/Welsh/Scottish/Northern Irish/British" tick box (5)

Breakdown of "Any other White" write in (10)

Breakdown of "White" output (1)

Bulgarian (1)

Central/South/Latin American (1)

Congolese (1)

Cornish (12)

Cypriot (1)

Eastern European (7)

English (4)

European (1)

Filipino (1)

French (1)

Gurkha (1)

Greek (1)

Greek Cypriot (1)

Gypsy (6)

Gypsy/Roma/Irish Traveller (2)

Irish Traveller (5)

Italian (1)

Jewish (4)

Kashmiri (1)

Korean (1)

Kurdish (1)

Latin American (3)

Lithuanian (1)

Mixed identities including an Irish identity (1)

Muslim (1)

Nepalese (including Gurkha) (1)

Northern Irish (1)

Non-European (1)

Orthodox Jewish (1)

Other nomadic (1)

Polish (7)

Roma (7)

Romanian (5)

Romany Gypsy (1)

Scottish (2)

Sikh (14)

Somali (7)

Sri Lankan/Tamil (1)

Syrian (2)

Tamil (2)

Turkish (3)

Turkish Cypriot (1)

Vietnamese (1)

Welsh (2)

Western European (2)

White European (2)

Yemeni (1)