Ethnic group, national identity and religion

Collecting data on ethnic group, religion and national identity can be complex because of the subjective and multifaceted nature of the concepts.

Membership to each of the concepts is something that is self-defined and subjectively meaningful to an individual.

However these concepts are measured, in particular, ethnic group and national identity, they tend to evolve in the context of social and political attitudes or developments.

This guidance replaces “Ethnic group statistics: a guide for the collection and classification of ethnicity data (2003)”.

Its content is based on the recently published GSS harmonised standards on ethnic group, national identity and religion, which recommend a harmonised approach in the data collection process to allow consistency and comparability of statistical outputs across GB and the UK.

It begins by looking at the development of the recommended harmonised country-specific questions for use on surveys in England, Wales, Scotland and Northern Ireland and highlights issues surrounding the specific requirements for constituent countries.

It provides advice on how to ask questions and present data on ethnic group, religion and national identity on social surveys in the UK and notes some of the complexities that should be considered when collecting and classifying data.

The guidance also acknowledges that in some instances not all surveys will have the resources to ask the harmonised questions in different countries; where this does occur, guidance following the recommended country-specific options should be followed.

It is hoped that this guidance will be a useful tool for those collecting data on ethnic group, national identity and religion.

The guidance will be revised and updated when necessary, and if you have any queries or wish to feedback on its content, please email: equalitiesandwellbeing@ons.gov.uk

Who is this guidance for?

This guidance is to help those who would like to collect and present ethnic group, national identity and religion data from social surveys in the UK, for example, social researchers, public bodies, etc.

Who should respond to the ethnic group, religion and national identity questions?

Ethnic group, religion and national identity are self-identification measures reflecting how people define themselves. Therefore, a response to a question should be answered by the respondent directly, particularly if the respondent is an adult. It is sometimes possible to ask another member of the household to reply on behalf of a respondent, however, this should be used only as a last resort. Where this does occur, notes should be recorded to reflect this. It is also important that interviewers do not attempt to use their own judgements.

The questions have been designed for use with adult respondents aged 16 or over. If the target population is below this age, guidance may be needed from the child’s parent, guardian or carer, particularly if the child is below the age of 12 as they may not understand what the question is asking. It is not recommended that categories are removed from the response options available to children as their choice should not be limited because of their age (or other factors).

How were the harmonised ethnic group, national identity and religious affiliation questions developed?

The ethnic group, national identity and religious affiliation questions were developed through a cross-government harmonisation project, undertaken through consultation and workshops with key stakeholders between 2008 and 2010 that included:

• Office for National Statistics

• Scottish Government and National Records of Scotland

• Welsh Government

• Office of the First Minister and Deputy First Minister/Northern Ireland Statistics and Research Agency

• Equality and Human Rights Commission

• Data Standards Working Group

• other government departments

• academics

Why use the recommended harmonised questions?

The questions should be used in order to achieve a harmonised approach in the data collection process that will allow for consistency and comparability of statistical outputs from different sources across the UK.

Why are there different requirements for asking questions on surveys in Scotland?

For Scotland, specific requirements came out of a review of the way Scottish surveys classified ethnicity. This followed recommendations made by the Race Equality Advisory Forum in 2001 and community concerns about the classification used in Scotland’s 2001 Census. The review conducted by the Scottish Government and National Reords of Scotland identified the benefits of developing a separate national identity and new ethnicity classification for use on the Scottish Census and relevant Scottish Official Statistics. Both questions used together would allow people to self-express their “Scottish-ness”, “British-ness” or any other national identity before expressing their ethnic group. The new ethnicity classification was published in July 2008, and details of the classification (including research, consultation and question-testing) can be found at the Scottish Government website. The classification was discussed by Members of the Scottish Parliament during considerations of the 2011 Census in Scotland and was subsequently amended for use in the Census. Details of the committee discussions can be seen in the Official Reports of the Economy, Energy and Tourism Committee (9th and 12th meetings).

There are also specific requirements for collecting data on religion in Scotland. This is done in order to provide data on which to measure or examine sectarianism, rather than looking at Christian and non-Christian. For further guidance on asking questions specifically for surveys in Scotland, please refer to the Scottish Government website.

Why are there different requirements for asking questions in Northern Ireland?

For Northern Ireland, specific requirements comply with the Good Friday Agreement (where it is not acceptable to ask respondents to choose between Northern Irish/British and Irish identities) and legislation under the Race Relations (NI) Order 1997, which outlaws discrimination on grounds of colour, race, nationality or ethnic or national origin. The Irish Traveller community is specifically identified in the Order as a racial group against which racial discrimination is unlawful. ‘Guidance for monitoring racial equality’ was published in July 2011 by the Office of the First Minister and Deputy First Minister and Northern Ireland Statistics and Research Agency.

There are also specific requirements for collecting data on religion in Northern Ireland. This is done in order to provide data on which to measure or examine sectarianism, rather than looking at Christian and non-Christian.

What are the different modes of administration?

The questions in this guidance are suitable for interviewer-led computer-assisted telephone interviewing, face-to-face and computer-assisted personal interviewing. The questions can also be used in self-completion modes of administration, e.g. paper-based, computer-assisted self-interviewing and internet. There are some variations in the way the questions need to be asked, depending on mode of administration.

This section provides a brief overview of the key legislation relating to ethnicity, national identity and religion.

Equality Act 2010

The new Equality Act 2010 has brought together over 116 separate pieces of legislation into one single Act. It provides a new discrimination law with a legal framework that protects the rights of individuals from unfair treatment and pushes forward equality of opportunity for all.

Equality Law applies regardless of:

• the size of the organisation

• the sector of work

• whether there is one worker or hundreds or thousands of workers

• whether or not the organisation uses formal processes or forms to help inform decisions

Public sector equality duty

The new public sector equality duty, which applies to all bodies carrying out public functions, came into force on 5 April 2011 and covers the 8 protected characteristics defined in the Equality Act, among which are race, religion and belief. The duty requires public bodies to have due regard to the need to eliminate discrimination, advance equality of opportunity and foster good relations in the course of developing policies and delivering services.

Specific duties

In addition, the Equality Act gives the government a power to impose specific duties on certain public bodies that set out steps they must undertake to enable them to better perform the public sector equality duty.

Section 75 of the Northern Ireland Act 1998

Section 75 of the Northern Ireland Act requires public authorities, in carrying out their functions relating to Northern Ireland, to have due regard to the need to promote equality of opportunity:

between persons of different religious belief, political opinion, racial group, age, marital status or sexual orientation; between men and women generally; between persons with a disability and persons without; between persons with dependants and persons without.

Without prejudice to these obligations, a public authority is also required, in carrying out its functions, to have regard to the desirability of promoting good relations between persons of different religious belief, political opinion or racial group.

Policy and service delivery

There is no express requirement under the public sector equality duty to collect monitoring information. However, in order to meet the duty, public bodies need to understand the impact of their work, and collecting monitoring information is likely to be an effective way of gathering sufficient information to enable them to do this.

Under Section 75 of the Northern Ireland Act, public bodies must monitor for adverse impact of their policies and publish the results of this monitoring.

Data Protection Acts

Laws exist to protect the confidentiality of data about living individuals (“personal data”) and give individuals rights to privacy or to access information held about them by public authorities. These laws are:

• Data Protection Act 1998

• Human Rights Act 1998

• Freedom of Information Act 2000

The Data Protection Act is concerned with the fair and lawful processing of personal data. Information about an individual’s ethnic background and religion is classified as sensitive personal data. Sensitive data is recognised under European and domestic law, and the conditions for processing such data are listed in the Data Protection Act.

What is express consent?

Legally, express consent means that respondents have authorised their data to be processed for the purposes that they have been informed about. This will assist with compliance with the first principle of the Data Protection Act.

If there are plans to share sensitive data, there is a responsibility to ensure that all parties processing the data are aware of the responsibilities and duties under the Data Protection Act, and that this is part of a formal data access agreement.

Specific requirements for asking questions on a survey in Scotland

Following the Race Equality Advisory Forum in 2001 and concerns by some communities in Scotland about the classifications used in the 2001 Census, the Scottish Government and the General Register Office for Scotland worked together to conduct a review. This review came with specific requirements for the ethnicity classification for Scotland, which was subsequently discussed by Members of the Scottish Parliament and amended for use on the Scottish 2011 Census.

Specific requirements for asking questions on a survey in Northern Ireland

For Northern Ireland, specific requirements comply with the Good Friday Agreement (where it is not acceptable to ask respondents to choose between Northern Irish/British and Irish identities) and legislation under the Race Relations (NI) Order 1997, which outlaws discrimination on grounds of colour, race, nationality or ethnic or national origin. The Irish Traveller community is specifically identified in the Order as a racial group against which racial discrimination is unlawful.

Since ethnicity is a multifaceted and changing phenomenon, various possible ways of measuring ethnic groups are available and have been used over time. These include country of birth, nationality, language spoken at home, skin colour (an aspect for consideration for some and not for others), national/geographical origin and religion. What seems to be generally accepted, however, is that ethnicity includes all these aspects, and others, in combination.

Country of birth

For many years, the only ethnicity statistics regularly available in the UK were based on people’s country of birth. This has limited reliability and has become increasingly less relevant when used on its own as the proportion of ethnic minority individuals born in the UK has increased and also because it includes children born abroad to British-born parents.

Nationality

Some host countries use nationality as their primary criterion, implying that migrants renounce their ethnicity once they have qualified for citizenship. However, it is clear that many of the disadvantages and other experiences associated with ethnic minority status continue long after naturalisation has been completed.

Language spoken at home

For some minority ethnic groups, language spoken at home may be an effective way of defining ethnicity. Such a question has been commonly asked in large national surveys of minority ethnic groups, not only to identify members of the minority but also to permit the matching of interviewer with respondent in cases where the interview is conducted in the indigenous language. But, as time goes on, this measure is becoming increasingly less useful: with the emergence of the second and third generations, young families may use English as their main language, even though they still identify with particular minority ethnic groups.

Skin colour

Skin colour is an option for considering ethnic group. However, it’s not an adequate criterion in its own right, and for some, its use is seen as unacceptable.

National/geographical origin

A question may include aspects of national or geographical origin, with the assumption that these help to identify ethnic groups. For example, the terms West Indian or Indian are taken as shorthand terms for members of ethnic groups originating in those parts of the world. A further development has been to combine national or geographical origin with a colour term such as Black, as in Black-African, to identify more precisely which group is being referred to for people originating from a part of the world which is itself multi-ethnic, such as sub-Saharan Africa.

Religion

One of the important defining characteristics for some ethnic minorities is their religion. Some commentators think that the religious dimension should be recognised more explicitly.

Collecting data on ethnic group is complex because of the subjective, multifaceted and changing nature of ethnic identification. There is no consensus on what constitutes an ethnic group and membership is something that is self-defined and subjectively meaningful to the person concerned.

The terminology used to describe ethnic groups has changed markedly over time, and however defined or measured, tends to evolve in the context of social and political attitudes or developments. Ethnic group is also very diverse, encompassing common ancestry and elements of culture, identity, religion, language and physical appearance.

This guidance highlights some of the complexities that need to be considered when collecting and classifying data on ethnic group. In the main, it provides advice on how to ask questions on ethnic group and how to present data from surveys using the recommended harmonised country-specific questions for England, Wales, Scotland and Northern Ireland. The are some differences in the questions because of the requirements of the constituent countries, and where this has implications for producing GB and UK outputs, these issues are highlighted.

It is recognised that not all GB- or UK-wide surveys will have the resources required to ask different questions in different countries. In these rare situations, it is recommended that the England and Wales question be used in Scotland. When this choice is made, a full UK and GB output can be produced.

What is the recommended harmonised ethnic group question to use on surveys?

The recommended ethnic group question for use on surveys came out of a 2-year cross-government consultation programme that wanted to harmonise data collection to enable consistency and comparability of data. It is recommended that the following question be used:

“What is your ethnic group?”

What is a "self-identification" measure of ethnic group?

In a survey, respondents are invited to select, from a list of categories, the ethnic group to which they consider they belong. There appear to be 2 factors determining the ethnic group that is recorded for each respondent:

• their own choice of how they view their own ethnicity

• the list of options are presented to them

The first consideration is the most straightforward; the ethnic group that each person chooses as his or her own is intrinsically the ethnic group of self-identity, rather than being ascribed by anyone else.

The second consideration is apparently not so clear-cut. The ethnic group options presented to the respondent are not completely ones of self-identity, since the respondent is likely to have had no say in the names or the number of the different alternative ethnic groups presented to them. Therefore, the freedom the respondent has to select their own group is constrained and influenced by the options presented to them.

Why have categories?

A category is used to assign data reported or measured for a particular situation according to shared characteristics. We use them to ensure consistent description and comparison of statistics. In practice, they are a set of “boxes” into which items can be put in order to get some kind of meaning.

Categories allow us, in an accurate and systematic way, to arrange our data according to common features, so that the resulting statistics can be easily reproduced and are able to be compared over time and between different sources.

Why are there differences in the ethnic group response categories?

When collecting GB or UK data, the format of the question will vary slightly, depending on whether data is being collected in England, Wales, Scotland or Northern Ireland. Each of the constituent countries has different requirements that have led in some instances to the use of different terminology and different questions. Therefore, if producing outputs for surveys with GB or UK coverage and using the Scotland harmonised country-specific questions, response categories can only be aggregated and presented at the main-level category (because of the variations in the questions being asked).

What are the differences in the ethnic group response categories?

The main differences in the ethnic group response categories are:

• White categories (excluding Traveller categories): Scotland separates Scottish from Other British and includes separate Irish and Polish categories; England and Wales combine UK and British categories but include a separate Irish category; Northern Ireland has 1 White category

• Traveller categories: in Northern Ireland, Irish Traveller is a main category separate from White; Scotland has a Gypsy/Traveller category and England and Wales have a Gypsy or Irish Traveller category

• Mixed Multiple ethnic groups: there is no category breakdown for the country-specific question in Scotland; there are no suggested categories as opposed to the other countries, where there are options

• there are some differences in the terminology and data collection of the country-specific Scotland question that make these categories difficult to compare: the African category in the Scottish question is presented in a separate section to the Caribbean or Black category; however, under the harmonised output, these 2 categories are output as part of “Black/African/Caribbean/Black British”– the African categories used in Scotland could potentially capture White/Asian/Other African in addition to Black identities

What are the implications for a GB and UK output when using the country-specific questions?

When collecting GB or UK data, the format of the question will vary slightly depending on whether the data is being collected in England, Wales, Scotland or Northern Ireland. Since each of the constituent countries has different requirements, the use of different terminology and different questions is unavoidable to ensure that data from surveys is comparable with that from their censuses. Therefore, if producing outputs for surveys with GB or UK coverage and using the Scotland harmonised country-specific questions, response categories should be aggregated and presented at the main-level category, and it will be necessary to explain differences in footnotes (see data presentation for details).

What instruction should be used when asking the ethnic group question in a face-to-face interviewer-led survey and self-completion survey?

• It is recommended that the ethnic group question will be asked in a way that allows the respondent to see all possible response options before making their decision. Therefore, in face-to-face interviewer-led surveys, a single show card should be used that presents all response options. The interviewer should then ask the respondent to select the option that best describes their ethnic group or background. Similarly, a self-completion survey (e.g. paper-based) should use a single question.

• The instruction “please describe” should follow the “Other” response options. This should be in non-bold font. These instructions should also be included on paper-based surveys.

What is the recommended ethnic group question for use on a survey in England?

Below is the recommended country-specific ethnic group question for use in England. This question is recommended when a show card is used in a face-to-face interview or self-completion survey (both paper and electronic).

Applies to all
Interviewer to read:

What is your ethnic group?

Choose one option that best describes your ethnic group or background

White

1. English/Welsh/Scottish/Northern Irish/British
2. Irish
3. Gypsy or Irish Traveller
4. Any other White background, please describe

Mixed/Multiple ethnic groups

5. White and Black Caribbean
6. White and Black African
7. White and Asian
8. Any other Mixed/Multiple ethnic background, please describe

Asian/Asian British

9. Indian
10. Pakistani
11. Bangladeshi
12. Chinese
13. Any other Asian background, please describe

Black/ African/Caribbean/Black British

14. African
15. Caribbean
16. Any other Black/African/Caribbean background, please describe

Other ethnic group

17. Arab
18. Any other ethnic group, please describe

What is the recommended ethnic group question for use on a survey in Wales?

Below is the recommended country-specific ethnic group question for use in Wales. This question is recommended when a show card is used in a face-to-face interview or self-completion survey (both paper and electronic).

Applies to all
Interviewer to read:

What is your ethnic group?

Choose one option that best describes your ethnic group or background

White

1. Welsh/English/Scottish/Northern Irish/British
2. Irish
3. Gypsy or Irish Traveller
4. Any other White background, please describe

Mixed/Multiple ethnic groups

5. White and Black Caribbean
6. White and Black African
7. White and Asian
8. Any other Mixed/Multiple ethnic background, please describe

Asian/Asian British

9. Indian
10. Pakistani
11. Bangladeshi
12. Chinese
13. Any other Asian background, please describe

Black/African/Caribbean/Black British

14. African
15. Caribbean
16. Any other Black/African/Caribbean background, please describe

Other ethnic group

17. Arab
18. Any other ethnic group, please describe

What is the recommended ethnic group question for use on a survey in Scotland?

Below is the recommended country-specific ethnic group question for use in Scotland. This question is recommended when a show card is used in a face-to-face interview or self-completion survey (both paper and electronic). This question has been developed to enable direct comparison with the Scottish Census and other sources in Scotland.

Applies to all
Interviewer to read:

What is your ethnic group?

Choose one option that best describes your ethnic group or background

White

1. Scottish
2. Other British
3. Irish
4. Gypsy/Traveller
5. Polish
6. Any other White ethnic group, please describe

Mixed or Multiple ethnic groups

7. Any Mixed or Multiple ethnic groups, please describe

Asian, Asian Scottish or Asian British

8. Pakistani, Pakistani Scottish or Pakistani British
9. Indian, Indian Scottish or Indian British
10. Bangladeshi, Bangladeshi Scottish or Bangladeshi British
11. Chinese, Chinese Scottish or Chinese British
12. Any other Asian, please describe

African

13. African, African Scottish or African British
14. Any other African, please describe

Caribbean or Black

15. Caribbean, Caribbean Scottish or Caribbean British
16. Black, Black Scottish or Black British
17. Any other Caribbean or Black, please describe

Other ethnic group

18. Arab, Arab Scottish or Arab British
19. Any other ethnic group, please describe

Note: When producing outputs for surveys with GB or UK coverage, it is only possible to harmonise with the rest of GB or UK at the “top-level category”. Footnotes should be included to explain the differences in the data collection.

Where it is not feasible to ask the recommended country-specific Scotland ethnic group question, the recommended question for England and Wales should be used. Ensure that the Scottish option is first in the response categories. See below:

Applies to all
Interviewer to read:

What is your ethnic group?

Choose one option that best describes your ethnic group or background

White

1. Scottish/English/Welsh/Northern Irish/British
2. Irish
3. Gypsy or Irish Traveller
4. Any other White background, please describe

Mixed/Multiple ethnic groups

5. White and Black Caribbean
6. White and Black African
7. White and Asian
8. Any other Mixed/Multiple ethnic background, please describe

Asian/Asian British

9. Indian
10. Pakistani
11. Bangladeshi
12. Chinese
13. Any other Asian background, please describe

Black/African/Caribbean/Black British

14. African
15. Caribbean
16. Any other Black/African/Caribbean background, please describe

Other ethnic group

17. Arab
18. Any other ethnic group, please describe

What is the recommended ethnic group question for use on a survey in Northern Ireland?

Below is the recommended country-specific ethnic group question for Northern Ireland. This question is recommended when a show card is used in a face-to-face interview or self-completion survey, both paper and electronic.

Note that the harmonised ethnic group question for Northern Ireland is different to that used in the 2011 Census of population in Northern Ireland. The Census question had fewer categories, although the harmonised question aggregates to the same categories. Census Office took a conscious decision not to use terminology such as British Black, as it was felt this would have needed to be counterbalanced with Irish Black in a Northern Ireland context. With limited space, Census Office omitted this terminology.

Applies to all
Interviewer to read:

What is your ethnic group?

Choose one option that best describes your ethnic group or background

1. White
2. Irish Traveller

Mixed/Multiple ethnic groups

3. White and Black Caribbean
4. White and Black African
5. White and Asian
6. Any other Mixed/Multiple ethnic background, please describe

Asian/Asian British

7. Indian
8. Pakistani
9. Bangladeshi
10. Chinese
11. Any other Asian background, please describe

Black/African/Caribbean/Black British

12. African
13. Caribbean
14. Any other Black/African/Caribbean background, please describe

Other ethnic group

15. Arab
16. Any other ethnic group, please describe

What instruction should be used when asking the ethnic group question on telephone surveys?

• The use of a show card is not possible on the telephone; therefore the question should be asked in 2 stages because of its length. The recommended stages are presented below as part 1 and part 2.

• The interviewer should use the word “or” after each response option in parts 1 and 2 of the 2-stage question. A pause in speech should be used to indicate a forward slash (/). For Scotland, “or” is used instead of a forward slash.

What are the harmonised country-specific ethnic group questions recommended for telephone surveys?

What about the "Other" category?

The list of ethnic group categories provided in the recommended questions are the result of extensive consultation and testing, but they are by no means definitive, and they do not capture all ethnic diversity in the UK. For this reason, a write-in option for each high-level category is made available under “Any other”. The write-in option is a very important category for the acceptability of the question and response rates.

What if more ethnic group categories are needed?

Categories within the recommended ethnic group questions are the result of extensive consultation and testing and by no means capture all ethnic diversity within the UK. The categories are designed to enable the majority of the population to identify themselves in a manageable way, and as a result they are restricted to those groups that make up the majority of the population. The length of this is not intended to exclude any groups of the population but rather the aim is to prevent the ethnic data collection from becoming overly complex and confusing, while ensuring the majority of the population can record themselves accurately.

While the measurement of ethnic group must take into account practical issues surrounding data collection and presentation, if categories are not sufficient, for some authorities it might be necessary to expand the “Other” category/categories to take into account local needs. This might result in a longer classification that will allow for identification of many groups; however, caution is required with analysis of such numbers (to ensure reliability and to avoid disclosure). If figures produced are too small to publish, then they can be suppressed, or it might be more suitable to aggregate to the relevant “Other” category.

National identity is a measure of self-identity, reflecting the subjective nature of national identity. A question on national identity allows a person to express a preference as to which country or countries, nation or nations that they feel most affiliated to. There has been an increasing interest in “national” consciousness and many people wanting acknowledgement of their national identity. In response, a recommended harmonised “national identity” question has been developed for use on social surveys that offers some insight into how an individual views their identity. This guidance outlines the process for asking and presenting national identity data.

It is recommended that the question should be used in addition to the ethnic group question and whenever there is a need to collect data about national identities; for example, when respondents classify themselves as English, Scottish, Welsh, Northern Irish, Irish, British or another national identity.

It is important to be aware that in Northern Ireland, to comply with the Good Friday Agreement, the national identity question must be asked in such a way that no one is forced to choose between being British, Irish and Northern Irish. This requirement is satisfied through the question being multi-response.

What is the recommended national identity question to use on government surveys?

The recommended national identity question for use on surveys came out of a 2-year cross government consultation programme that wanted to harmonise data collection to enable consistency and comparability of data. The recommended question to be used is:

“How would you describe your national identity?”

Where on a survey should the national identity question be asked?

The national identity question should be asked as a separate question but in addition to the ethnic group question. This is because research has shown that classifying ethnic group is best achieved separately from national identity.

Research also suggests that people were happier when asked about their national identity first. It is therefore recommended that the national identity question is asked immediately before the ethnic group question in surveys.

Why have categories?

A category is used to assign data reported or measured for a particular situation according to shared characteristics. We use them to ensure consistent description and comparison of statistics. In practice, they are a set of “boxes” into which items can be put in order to get some kind of meaning.

Categories allow us, in an accurate and systematic way, to arrange our data according to common features, so that the resulting statistics can be easily reproduced and are able to be compared over time and between different sources.

What order should the categories be?

It is recommended that the order of response categories should be changed depending on the country in which the question is being asked. The recommended orders of categories are given below in national identity question (a).

Where cross-country comparability is a priority above comparability with respective census questions, categories for a single UK question have been recommended in national identity question (b). It is still recommended, however, that the category orders follow those presented in national identity question (a). Note that there is a slight variation of response categories for Northern Ireland; in both (a) and (b), an Irish category is located between the British and Northern Irish categories. To comply with the Good Friday Agreement, the national identity question must be asked in such a way that no one is forced to choose between being British, Irish and Northern Irish. If they wish, respondents may report more than 1 national identity.

Multi-response

The national identity question is a multi-response question; this allows for a proportion of respondents who wish to identify with 2 or more identities. Multiple-response categories for the presentation of results are primarily data-driven. Categories will need to be agreed before presentation.

It is important to be aware that in Northern Ireland, to comply with the Good Friday Agreement, the national identity question must be asked in such a way that no one is forced to choose between being British, Irish and Northern Irish. This requirement is satisfied through the question being multi-response.

Where there are multiple responses, presentation of results should include a footnote to explain why the total number of responses is greater than the sample population.

What about the ‘Other’ category?

The list of national identity categories provided in the recommended question is by no means definitive and does not capture all national identities in the UK. In order to capture nationalities outside of the UK, a write-in option under “Other” national identity is available. This category is very important for the acceptability of the question and response rates.

What if more national identity categories are needed?

Categories of national identity in the recommendation questions by no means capture all identities within the UK, and in certain cases, it might be necessary to explicitly list additional categories. Categories must take into account practical issues surrounding data collection and presentation. A longer classification might allow for identification of many smaller groups of identities; however, caution is required with analysis of such numbers (to ensure reliability and to avoid disclosure). If figures produced are too low to publish separately, then they may need to be aggregated with the “Other” group for publication. For some cases, it might be necessary to expand the “Other” category.

What instructions should be used when asking the national identity question?

It is recommended that a show card is used in interviewer-led surveys. Where this is not possible (e.g. telephone interviews), the response categories should be read out by the interviewer in the same order as they appear in the box below.

The question is a multi-response question and the show card must include the instruction “Please choose all that apply” in addition to being read out by the interviewer. This is particularly important in Northern Ireland, where respondents must not be asked to choose between British, Irish and Northern Irish (see above).

The instruction “please describe” should also be included on the show card following the “Other” response option. This should be in non-bold font. These instructions should also be included on paper-based surveys.

What is the recommended national identity question for a survey in England, Wales, Scotland and Northern Ireland?

This is the recommended national identity question and layout for use on a survey in England, Wales, Scotland and Northern Ireland. The order of response categories are changed depending on where the question is being asked.

What is the recommended national identity question for asking a single UK question?

This is the recommended national identity question and layout for use on a survey when using the single UK question.

For many people, religion is one of the most important defining characteristics of an identity.

While religious identity can be objective, this guidance focuses on measuring the subjective nature of religious identification and how people define themselves.

Surveys can measure different concepts of religion including religious “affiliation”, “practice”, “belief” and “belonging”. This guidance begins by outlining those concepts but then specifically details the process of asking about “religious affiliation”, which is the recommended concept for the Government Statistical Service (GSS) harmonised question for use on surveys.

The question was developed to enable consistency and comparability of data across GB and the UK. The guidance provides advice on how to ask, present and discuss religion data at a country-specific level and a GB and UK level, while also highlighting how to avoid data being misinterpreted by users.

A question on religion can be asked by itself or along with other questions, such as ethnic group, as they are often closely interrelated concepts. When used together, they can complement each other to allow for a more distinct picture of a population.

What are the concepts of religion that can be measured on social surveys?

There are different concepts of religion that can be measured on social surveys, and these can include affiliation, belonging, belief and practice. It is important to be clear about what concept is being measured:

• Religious affiliation is the connection or identification with a religion irrespective of actual practice or belief.

• Religious belonging can be interpreted as both loose self-identification and active or formal belonging to a religious group. This can produce problems of ambiguity as some people may respond that they have a religious affiliation but not belong to a religion.

• Religious belief includes beliefs typically expected to be held by followers of a religion and how important those beliefs are to a person’s life.

• Practice includes specific religious activities expected of believers.

Are UK comparisons possible with the census questions on religion?

The census is the most commonly used source for statistics on religion, and a voluntary question was asked for the first time in GB in 2001 (a religion question has been asked in Northern Ireland since 1861). In England and Wales, the question asked was “What is your religion?” For Scotland and Northern Ireland, there were 2 questions asked: “What religion, religious denomination or body do you belong to?” and “What religion, religious denomination or body were you brought up in?” The same religion questions have been asked on the 2011 Census of the respective countries; however, in Scotland, only 1 question has been asked: “What religion, religious denomination or body do you belong to?” The question on upbringing wasn’t asked. The religion question in Northern Ireland is used to derive community background (that is, Catholic, Protestant or Other). The additional question is included in Northern Ireland partly to enhance response rates. Legislation in Northern Ireland was framed in terms of “perceived religious identity”.

Due to the differences in the questions it is important to be aware that when trying to make UK comparisons with census data on religion:

• The 2011 question on the England and Wales census asks about ‘affiliation’, whereas, the 2011 Scotland and Northern Ireland censuses ask about ‘belonging’.

• The way people respond to questions on religion is sensitive to what question is being asked and how it is asked.

As a result, comparing or combining data relating to different concepts is not recommended.

Should comparisons on religion be made with different data sources?

Caution should be used when comparing religion data from and with different sources. The census of England and Wales asks a different religion question to that of Scotland and Northern Ireland.

There is also variation in the religion questions asked on different surveys; some ask about belonging and attendance whereas others ask about importance or belief. Note that on the LFS in Northern Ireland, the religion question asked is about denomination. It is important to note that the way people respond to questions on religion is sensitive to what question is asked and how it is asked.

What is the recommended harmonised religion question to use on government surveys?

A religion question on affiliation was recommended for use on surveys after a 2-year cross-government consultation programme that wanted to harmonise data collection across GB and the UK to enable consistency and comparability of data. It is therefore recommended that where a single question on religion is required for data collection in the UK, the GSS harmonised religious question that measures affiliation should be used, and that is:

“What is your religion?”

If further information on religious practice is required, a follow-on question about practice can be used, such as (this question is not GSS harmonised):

“Do you consider that you are actively practising your religion?”

What instruction should be used when asking the religion question?

It is not recommended that a single Christian category is used across GB or the UK. The recommended breakdown for the Christian denominations that should be used in Scotland and Northern Ireland are provided below.

It is recommended that a show card is used in interviewer-led surveys in GB. Where this is not possible (e.g. telephone interviews), the response categories should be read out by the interviewer in the same order as the appropriate show card.

The instruction “please describe” should be included on the show card following the “Any other religion” response option. This should be in non-bold font. This instruction should also be included on paper-based surveys.

Use of a show card is not recommended in Northern Ireland. The interviewer should read the question and wait for a spontaneous response. If a response is not forthcoming, the interviewer may prompt using the categories in the box below. (A prompt is often required to follow up a response such as those responding as “Christian” or “Protestant” with the question “And what type of ‘Christian’ (or ‘Protestant’) is that?”) The use of show cards and prompts should be acknowledged when comparing data.

What is the recommended religion question for a survey in England?

This is the recommended religion question and layout for use on a survey in England.

Recommended religious affiliation question for England

Interviewer to read:
What is your religion?

Interviewer to read options
No religion
Christian (including Church of England, Catholic, Protestant and all other Christian denominations)
Buddhist

Hindu
Jewish
Muslim
Sikh
Any other religion, please describe

What is the recommended religion question for a survey in Wales?

This is the recommended religion question and layout for use on a survey in Wales.

Recommended religious affiliation question for Wales

Interviewer to read:
What is your religion?

Interviewer to read options
No religion
Christian (all denominations)
Buddhist
Hindu
Jewish
Muslim
Sikh
Any other religion, please describe

What is the recommended religion question for a survey in Scotland?

This is the recommended religion question and layout for use on a survey in Scotland when wanting to harmonise with the rest of GB and the UK. The recommended breakdown for Christian denominations for Scotland (like Northern Ireland) is different to that of England and Wales. This is in order to provide data on which to examine the differences between the main Christian groups. However, categories can be aggregated at the main level of category for a full GB or UK output.

If it is more important to compare data sources within Scotland, it is recommended to use Scotland’s 2011 Census question on religion (which asks “What religion, religious denomination or body do you belong to?”) or Scotland’s recommended question for social surveys, which asks the same question but includes a “Pagan” response option.

Recommended religious affiliation question for Scotland

Interviewer to read:
What is your religion?

Interviewer to read options
No religion
Church of Scotland
Roman Catholic
Other Christian
Buddhist
Hindu
Jewish
Muslim
Sikh
Any other religion, please describe

What is the recommended religion question for a survey in Northern Ireland?

This is the recommended religion question and layout for use on a survey in Northern Ireland. The recommended breakdown for Christian denominations (like Scotland) differs slightly to that of England and Wales. This is different in order to provide data on which to measure or examine community background (that is, “Catholic”, “Protestant” or “Other”) and for other purposes rather than looking at Christian and non-Christian. However, categories can be aggregated at the main level of category for a full GB or UK output (see below).

If an interviewer is asking a question, the use of a show card is not recommended. The interviewer should read the question and wait for a spontaneous response (note that “interviewer to read options” is omitted from display). If a response is not forthcoming, the interviewer may prompt using the categories in the box below. A prompt is often required to follow up a response, such as those responding as “Protestant” or “Christian”, with the question, “And what type of ‘Protestant’ (or ‘Christian’) is that?” The use of show cards and prompts should be acknowledged when comparing data.

Recommended religious affiliation question for Northern Ireland

Interviewer to read:
What is your religion?

No religion
Catholic
Presbyterian
Church of Ireland
Methodist
Baptist
Free Presbyterian
Brethren
Protestant – Other, including not specified
Christian – Other, including not specified
Buddhist
Hindu
Jewish
Muslim
Sikh
Any other religion, please describe

What about the ‘Other’ category?

The list of religious categories provided in the recommended question is by no means definitive and does not capture all religious diversity in the UK. For this reason, a write-in option for the “Any other religion” category is available. This category is very important for the acceptability of the question and response rates.

What if more religious categories are needed?

Measurement of religious diversity must take into account practical issues surrounding data collection and presentation. If categories are not sufficient, a longer classification could be produced (provided they can be aggregated to the main category), or it might be necessary for some authorities to expand the “Other” religion category. However, this could lead to identification of smaller religious groups. If figures produced are too small to publish, then they can be suppressed or aggregated to the relevant category or included within “Other” religion (footnotes should be included to explain what is included).

A longer classification is often used to allow greater variety of aggregations; however, this can also allow for identification of many smaller religious groups. Therefore, caution is required with analysis of such numbers (to ensure reliability and to avoid disclosure).

How to avoid misinterpretation of religion data by users?

Because there are several dimensions to religion, talking and reading about religion or specific religions without qualification can be confusing and lead to misinterpretation by data users. Therefore, if presenting data from the harmonised religion question, it is important to be explicit and refer to the specific “concept” being measured, and that is “affiliation”. It is also recommended that when presenting data on religious affiliation, it should be accompanied by a short note, such as:

Respondents were asked the question, “What is your religion?” which measures affiliation – that is, the identification with a religion, irrespective of actual practice or belief. There should also be caution when using inappropriate terminology that can lead to confusion, for example:

The term faith should not be used as an alternative to religion when referring to statistics as this may imply a stronger connection to a religion than is actually meant. In this instance, it would be misleading to say that a “percentage of people have a faith”.

Writing about people “being religious” rather than “having a religious affiliation” may be inferred as a reference to levels of practice or belief. For example, it may be misleading to state that “Chinese people are the least religious; in 2001 more than half said that they had no religion” when what is meant is “Chinese people are the least likely to say they have a religious affiliation”.