What the public thinks about ethnic identity terminology

Final Report

February 2023

EXECUTIVE SUMMARY

In response to actions from Inclusive Britain, the Inclusive Data Taskforce and the GSS Harmonisation workplan, the Cabinet Office’s Equality Hub and the Office for National Statistics commissioned The Social Agency (then Basis Social) to undertake qualitative research into the use of language and terminology relating to people’s ethnic identity.

The research comprised a mixture of group discussions and depth interviews with participants identifying as members of different ethnic groups, including five one-to-one interviews with members of the public living in the UK who reported no-to-low English proficiency. Focus group participants were also invited to take part in a task to identify inclusive and stigmatizing ethnicity language and terminology.

The research sought to explore 3 themes:

How people define their ethnicity and the factors driving this
Attitudes towards current ethnicity classifications
Stigmatising language and its consequences

The evidence from this research will be used by the Equality Hub to support the development of guidelines to encourage responsible and accurate reporting on race and ethnicity issues. It will also be used by ONS to improve inclusion and the conceptual measurement of ethnicity, supporting possible improvements and changes to the question design as needed.

Main findings

The report sets out 12 main findings and the evidence behind each. Overall, these can be summarised into 3 overarching conclusions:

Ethnicity is a multifaceted concept that is deeply important to people’s self-identity but can be fluid depending on the context in which they are expressing it.
Ethnic signifiers vary between groups, but overall national identity (especially British) was considered important whereas religion was not; skin colour was divisive as an ethnic signifier.
Current labels and classifications are relatively well understood but should be used carefully and considerately in reports to minimise stigmatization.

Recommendations for government

For data collection

Ethnicity data should be reflective of how people want to self-identify and this should be made clear in the way the question is asked.
People should always have the option to complete a free-text response about their ethnicity.
Include caveats alongside the reporting of ethnicity data to explain that care should be taken when interpreting it because ethnicity is a multifaceted concept which can mean different things to different people and self-reported ethnicity may be influenced by when it is collected, by whom and for what perceived purposes.
Although not universally accepted skin colour should remain a signifier in the classification mechanism for White and Black ethnic groups, but more effort should be made to show the relevance and usefulness of collecting this information in questionnaire guidance.
Skin colour should not be incorporated as a signifier for Asian or other ethnic groups.
Standardise the use of the label ‘British’ across the classification, or include the nationalities English, Welsh, Scottish, Northern Irish, or British in each of the ethnic groups and not just the White group.
Include growing communities, particularly within the White Other and Other categories such as Eastern European and Latin American.
Re-order the ethnicity classification alphabetically, to avoid creating a sense of importance and priority which is unhelpful in reinforcing feelings of prejudice and inequality.

For communications and reporting

Ethnicity should not be used as an identifier when reporting on a given story unless there is a clear and valid reason to do so.
Be aware of the potential for ethnic terminology to be used in a way which is stigmatizing or reinforces stereotypes.
Avoid using umbrella groups when reporting ethnicity and make, references as specific as possible to the exact ethnic group of relevance to avoid confusion or misrepresentation.
Avoid using umbrella terms such as “people of colour” or "Black and Minority Ethnic (BAME)” as they are both too general to be useful and can be divisive.

BACKGROUND

In March 2022 the government published the Inclusive Britain report. This is the government’s response to the independent Commission on Race and Ethnic Disparities. The Inclusive Britain report contained 74 actions to be taken forward by a range of government departments, including the Cabinet Office's Race Disparity Unit (RDU).

Action 7 of the report states that the RDU will work with people from different ethnic groups to better understand the language and terminology that they identify with. They will also review ethnicity terminology used in the media and other public materials. This work will help the government to develop recommendations to encourage responsible and accurate reporting on race and ethnicity issues.

The National Statistician published the Inclusive Data Taskforce’s Implementation Plan at around the same time to improve the UK’s inclusive data holdings. The plan includes:

activities to build trust with people participating in data supply;
methodological work to improve inclusion and conceptual measurement of currently under-represented population groups;
improvements in the accessibility of data and analysis.

In February 2022 the Government Statistical Service (GSS) published their Harmonisation workplan, which seeks to review the current harmonised ethnicity question. The GSS will work with stakeholders and user groups to understand user needs and test possible improvements and changes to the question design, as needed.

To provide evidence for these plans, in September 2022, The RDU and the Office for National Statistics (ONS) commissioned an independent organisation, Basis Social, to undertake qualitative research into the use of terminology relating to people’s ethnic identity. This research would provide the evidence for each of these plans. The research was commissioned to understand the opinions of people from different ethnic groups about:

what is inclusive;
what is not inclusive;
how to describe their ethnic identity most appropriately.

The main objectives of the commission were to help develop best practice in the use of ethnicity terminology by:

considering how people from a range of ethnic groups identify, and the factors that affect this;
asking for feedback on the existing ethnicity classification, with particular interest in colour terminology, such as “other” and “Mixed”;
asking people what types of language are inclusive, neutral, or stigmatizing;
exploring the consequences of non-inclusive or stigmatizing terminology.

This report presents the initial findings from this research. In-depth, quality assured, findings will be presented by the GSS Harmonisation team on behalf of ONS in late spring 2023.

METHODOLOGY

Basis Social designed a programme of research which included a mixture of group discussions and depth interviews with participants identifying as members of different ethnic groups. Basis Social worked with the recruitment agency Ethnic Opinions to recruit 80 participants to participate in 16 two-hour group discussions conducted online using Zoom. Focus groups were split both by ethnicity and levels of education, and were Mixed in terms of the age, sex, socioeconomic status, and location of participants. Participants self-identified their ethnic group using the 2021 Census classifications during the screener interview. Group discussions were conducted with the following ethnic groups:

White English, Welsh, Scottish, Northern Irish, or British
White Other
Mixed
Black African, Black Caribbean, or Black Other
Indian, Pakistani, Bangladeshi [South Asian]
Chinese or Asian other [East/ South East Asian]
Other (+ Arab)
younger people (aged 18-24) from a range of ethnicities

Focus group participants were also invited to respond to a variety of news headlines. These headlines acted as a stimulus towards identifying inclusive and stigmatizing ethnicity language and terminology in public discourse.

Ten participants from the focus groups were invited to participate in a follow-up depth interview. These interviews were designed to allow deeper exploration of the questions asked during the focus groups. The sessions concentrated on participants who had a more complex Mixed ethnicity, or participants who viewed religion, non-British national identity or their cultural heritage as core to their ethnicity. Each interview lasted up to 60 minutes and was conducted using Zoom.

Finally, 5 one-to-one interviews were conducted with members of the public living in the UK who reported no-to-low English proficiency. These interviews were designed to ensure people were not excluded from the study if they spoke little English, or no English. The interviews covered similar questions to the focus groups. Each interview was conducted with a participant with a different first language and was conducted in that participant’s first language. Interviews were conducted with speakers of:

Polish
Vietnamese
Mandarin
Bengali
Arabic

The moderation team involved in conducting interviews and group discussions analysed the data to identify common themes, topics, ideas, and patterns of meaning. This thematic analysis involved a deductive approach (i.e. focusing on the main research objectives and the associated topics that formed the basis of the discussion guide). Additional themes, where identified, were discussed within whole-team analysis sessions. The team then returned to the source data, such as transcripts, to consider whether and how these themes were present across different subgroups.

This report contains provisional findings based on analysis by Basis Social. Additional analysis of raw data will be conducted in Spring 2023 by researchers within ONS and RDU to validate the presentation and interpretation of data.

KEY FINDINGS

1. How people define their ethnicity and what affects this

Within this research significant variations were identified in how different people and groups understand and define the term ‘ethnicity’. Some participants said they define their ethnicity differently depending on the context.

1.1. Ethnicity as a multi-faceted concept

Participants in this research associated a broad range of characteristics with ethnicity. These characteristics included:

skin colour
ancestral, familial or personal place of origins
nationality - for the purposes of this research, ‘nationality’ refers to a person’s legal status as a citizen of a particular country, as reflected by, for example, holding a passport
national identity - for the purposes of this research, ‘national identity’ refers to how that person chooses to classify themselves, for example in terms of the country they feel closest to
cultural heritage
personal values
culture
customs
practices
traditions

Some participants said:

“It's what I descend from, and what I'm passing on. It's what makes my culture, my beliefs, the way I was raised, the family influences, the larger societal influences that I grew up within. And my DNA, it's who I am.” White Other, Lower Education

“I think when you talk about [ethnicity] people automatically think about race and where you're from. They don't think about the other attributes that could come with it. Like, for me it's not just about where my parents were from, it's also about, for me, it's more about my moral code as well and who I am.” South Asian, Lower Education

Participants often defined ethnicity by using a wide range of different terms and phrases, including:

race
ancestry
heritage
background
where a person is from

However, across the sample there was little consistency in these definitions, either when participants were defining the abstract concept of ethnicity, or when they were defining their own ethnic identities. This indicates that ethnicity is a multifaceted concept that holds different meanings to people in different contexts.

These differences are clearest when contrasting ethnicity as a means of classifying people, with ethnicity as a component of someone’s sense of self.

1.2. How people understand ethnicity labels as a method of classification

As a term, ‘ethnicity’ was most often associated by participants with public services and employers who use forms to collect personal data to categorise people. Ethnicity can therefore be understood as something which holds ‘external relevance’. It is data that is being collected by ‘other people’ who want to use this data to group people based on shared common characteristics, in this case, their ‘ethnicity’.

Participants recognised that ethnicity is used to classify people based on certain criteria, specifically their skin colour or a location. This could relate to someone’s place of birth or the place in which their parents, grandparents or ancestors were born.

People involved in this research described their ethnicity in a way that appeared subjective, but ethnicity as a classification seemed objective because it relates to things that cannot be changed. For example, it references a skin colour or a nationality. This contrasts with, for example, religion, language, or national identity, which participants saw as changeable. This illustrated a tension between a classification definition which is objective and fixed, such as ‘I am Black and British’, and how people understand and express their ethnicity which is more nuanced, for example ’I am Black, and British, but my father is Black African, from Nigeria, and my mother is Asian but born in Ghana’.

Some participants and groups were more comfortable than others with the idea of using ethnicity to categorise people. Their opinions varied within and between groups depending on the context in which this data was understood or expected to be used. Participants who were most comfortable tended to be more aware and accepting of how data can be used to benefit different communities, for example by identifying inequalities. But this varied across the sample of participants.

There were some groups who were distrustful of the collection of ethnicity data because they were concerned about how that data could be used for racial profiling and discrimination. This feeling was most evident in participants from Black ethnic groups.

It is important to recognise that the labels used to categorise people by their ethnicity carry meaning and come with (often more negative) assumptions, biases and stereotypes attached. This is particularly true of labels that relate to Black British, Black African, Black Caribbean, or Black Other ethnicities. This makes it particularly important to explain and justify the purpose of data collection.

“Why is it relevant? I think 'Why am I being asked this question?' If I put White British, if I put black Caribbean, if I put British Asian, would it make a difference to the outcome of whatever it is I'm trying to achieve? Would it make a difference to how I'm being perceived?” Black British, Black African, Black Caribbean, or Black Other, Higher Education

“So, it's like when I say I'm South African, do people associate, you know, apartheid with my nationality and where I come from?” 18-24 years and Mixed ethnicity, Lower Education

“It makes me think about my ethnic minority background and how I identify as a young, black male living in a predominantly white society, like Britain… it does make me think ‘How do I identify within that?’… in terms of your ethnicity, it makes me feel a little bit defensive.” Black British, Black African, Black Caribbean, or Black Other, Lower Education

Most participants, aside from the Black British, Black African, Black Caribbean or Black Other participants, gave relatively little thought to providing their ethnicity for the purposes of surveys or equivalent data collection exercises. In many cases these participants would select options which were best fit or simplest to complete. This was most evident in younger people.

“I think it's just part of the application form, like, just as it is with writing your first and last name. I think it would be weird if I didn't see it. It's something that I always fill out and I don't find troubling filling it out because they give a lot of options as well.” 18-24 years and mixed ethnicity, Higher Education

"I think there's so much more behind it, that it just doesn't explain in those two words, and because it's so detailed, but for the sake of ease and peace, it's the best way.” 18-24 years and mixed ethnicity, Higher Education

Ethnicity labels were something that participants felt conditioned to providing without reading too much into why they were being asked for that information, how it would be used, or whether the information would result in particular consequences for them personally.

But there were circumstances where participants reported being more careful in providing their ethnicity. This was often in situations where they felt being affiliated with a particular ethnic group would be either beneficial or detrimental, such as deciding whether disclosing their ethnicity may be advantageous or disadvantageous when applying for a job. This was more so the case for participants of Mixed heritage who felt they were able to choose between different aspects of their ethnic identity.

1.3. Ethnicity as an important part of self-identity

“I think that when you're just filling out a form, where you're from is enough, but when you're talking to somebody…I think that's when you go deeper into [it] and talk about every aspect of your ethnicity.” South Asian, Lower Education

Another way in which participants understood ethnicity was as a way of thinking more deeply about their personal and social identities. In this respect ethnicity was seen as having ‘internal relevance’. It was part of their sense of self-identity, their heritage, and their connection to a greater collective identity, where nuance and detail is important. Ethnicity was something that people took pride in. It was seen to relate both to their inheritance, which includes their culture, history, tradition, language and beliefs, and their immediate group membership, which includes their family and community social circles. These thoughts were typically expressed through conversation with others.

The nuance in ethnic identification makes a person who they are. This was shown in how participants described themselves when asked to define their ethnicity. While some participants defined ethnicity in a way that aligned with the broader ethnicity classifications, such as White, or Black Caribbean, most people talked about a mixture of things. They defined their ethnicity by talking about their country of birth, race, skin colour, cultural or geographical heritage, values, and communities.

There were differences in the degree to which ethnicity was seen as more of a fluid or fixed concept. This was often connected to the complexity of participants’ own identities and cultural heritage. Some participants associated ethnicity with DNA, genetics, and historical facts, such as where you were born. These participants understood ethnicity as more of a fixed concept and this idea was more evident in particular participants from the White British group. Other participants felt that someone’s ethnic identity could evolve and that a person could decide how to describe their own ethnicity. In our sample these tended to be participants who had migrated to new countries or whose understanding of their own ethnicity had changed over time. For example, this could be because they had learnt more about different parts of their heritage.

“I don't think it's something you can just change like that. It has to develop over time…Also, I think it would depend on the person: maybe for someone it could be five years, for someone else it could be a whole generation, to actually change the way you identify yourself…” Other (British Cypriot), Higher Education

“Ethnicity is to do with the genetics of it and, like, maybe a bit of the generalised location of where they're from.” White British, Lower Education

“Whether you accept it, and like your culture, it's irrelevant because you can't change it, you are born that way, your genetic make-up is from that country.” White other, Higher Education

The following table summarises the key ways in which different ethnic groups reflected on their ethnicity.

While ethnicity was seen as something which was inherited it could also be complicated to define consistently. There could be many different reasons for this, such as:

the different levels of awareness, knowledge, and connection that people may have with their heritage – this means they have different starting points when defining their ethnicity;
the tendency for a person to ‘lean in’ to the cultural background which is closest to their environment and who they are surrounded by or most closely identify with;
migration to Britain, or parents or grandparents who migrated to Britain - the relevance, knowledge and importance of different national identities can be weighted differently, which can affect how people of a more Mixed heritage describe themselves;
situational factors – this includes considerations about what a person may gain or lose by describing their ethnicity in a particular way in a given situation.

“I only know, like, my ‘Nigerianess’ from my mum and that's from her history and that's what I take to represent who I am.” 18-24 years and Mixed ethnicity, Lower Education

1.4. Defining ethnicity outside of the UK

Participants in non-English language interviews were presented with the Census 2021 ethnicity question translated into their own language. Findings from these interviews suggest that the concept of ‘ethnicity’ has varying degrees of ambiguity and complexity when translated into different languages. For example, an Arabic speaking participant understood the translation for ethnicity to mean:

“a person’s roots, roots stem from the earth, and they grow, and so going back to one’s roots of where they grew up from.” Translation of an Arabic speaking participant’s interview

This suggests an emphasis on personal ancestry which is just one of the connotations that the word 'ethnicity' has in English. The Arabic Speaker later defined their ethnicity as “Arabic” in their interview.

A Vietnamese interviewee, by contrast, struggled with the translation. They said:

“you can talk about an ethnic group in terms of religion…It’s like a tradition…I don’t really understand it.” Translation of a Vietnamese speaking participant’s interview

Non-English interviews also showed that some participants were more familiar with the word ‘ethnicity’ in English, as well as the range of connotations that the English word can carry for people who have lower proficiency in English, or do not speak English at all. The Polish interviewee, for example, reported seeing the English word frequently in surveys and forms, and associated it with “roots” and “origins”.

“The first thing that comes to my mind is roots…Where do we come from? That's the first one for me. I don't know if that's the correct association or not, but that's just how I associate it. Just where we originate from.” Translation of a Polish speaking participant’s interview

For the Mandarin, Bengali and Vietnamese speaking interviewees, ‘ethnicity’ carried strong connotations of being ‘foreign’ and having minority status. They had heard the term most frequently in the context of communications around ‘ethnic minorities’, which is a phrase some people understood to mean ‘being from another country’.

“I didn’t even know what an ‘ethnic minority’ is at the beginning. Now gradually after living in this country I got to know the meaning. A lot of political incidents occur in this country. I got it gradually.…I just know that if anyone comes from another country, then it is called an ethnic minority.” Translation of a Bengali speaking participant’s interview

“[Ethnicity or ethnic group] means that he belongs to a minority in this society, a marginal group. There is a main group, and he needs to integrate into this main group.” Translation of a Mandarin speaking participant’s interview

1.5. Self-identifying ethnicity in different contexts

Most groups, except White British participants, reported that they would change the label they used to describe their ethnicity depending on the situation and context they were asked. There were several explanations about when and why some participants would change how they describe their ethnicity. These responses suggested two core themes.

Firstly, for many participants, defining ethnicity is an intersubjective process (i.e. a process which accounts for a social context and the interaction between two or more people). When participants were asked to define their ethnicity, they described making judgements about:

whether the questioner has another motive for asking this question;
what types of responses the questioner was likely to find acceptable;
whether the questioner would have knowledge and familiarity with their ethnicity.

Participants then adjust the labels they use to describe themselves to find a best fit between what is true for them, and what they believe about the purpose or intentions of those asking for their ethnicity, or where they are “from”. Participants also reported adjusting their responses when they fill in forms. This process sometimes meant participants would describe only a small part of their identity, perhaps at a more general level than they might do in other contexts. They may even deliberately leave out certain parts.

For example, participants reported suppressing their ‘Britishness’ when describing their ethnicity. This was often because they felt it was questionable because of external factors, such as accents.

“I can say I'm British because I have a British passport. I've been living here for so many years. But, I feel it's not right to say that. Because every time I say I'm originally from Poland, people can hear, and they kind of nudge.” White Other, Lower Education

“I will always say Indian, because if I were to say British, I would feel the opposite person is definitely going to be thinking, 'She's not really British. What's her ethnicity?' I hope I do get comfortable to say British at some point. … filling a form, a document, I would definitely want to choose Indian over anything else because I just feel like the other person is going to judge my answer.” South Asian, Higher Education

Participants from a Mixed background also spoke about times when they identified as Black British, Black African or Black Caribbean in general conversation. This was partially because these were the aspects of their ethnicity they most identified with, but also because the colour of their skin was clearly “not White” so they felt that defining themselves in this way would also best align with the expectations of others.

Secondly, some participants described their ethnicity differently depending on the perceived risks and benefits of identifying in a particular way and in a specific context. Participants mentioned risks such as being stereotyped or discriminated against, for example in the form of jokes, or being passed over for promotion at work.

“When you say, 'I am from Colombia,' they start with a joke. The usual joke of, you know, Pablo, Narcos or whatever… And then they start to say, like, 'Oh, do you sell this?' And you're like, you know, it's kind of a joke, but it's racist… I think I am more comfortable with ‘South American’. I love my country. When I feel comfortable to disclose my nationality that's perfect. If I am with civilised people, you know. But with the rest, I am South American, or Latino, that's it, simple.” Other (Latin American), Lower Education

These findings exclude White British participants.

2. Attitudes towards current ethnicity classifications

Participants in both the focus groups and interviews were presented with the ethnicity question used in the 2021 Census for England and Wales. This helped us understand attitudes towards current ethnicity classifications. Participants were largely comfortable in identifying themselves within the range of options available to them, and while not all-encompassing, the opportunity to self-identify in free-text form meant that most people were satisfied with the ethnicity questions as presented to them.

2.1. Skin colour as a divisive ethnic signifier

Skin colour was mentioned by most participants recruited to the White British group, and the Black African, Black Caribbean or Black British groups when defining their own ethnicity. Colour terminology is not currently used in Asian ethnicity labels and Asian and Asian British participants did not use skin colour to describe themselves. Across the participants, there was a range of views on both the acceptability and usefulness of including skin colour as an ethnic signifier.

Skin colour was an important part of self-identity for participants who identified as Black or Mixed. Discussions among participants of Mixed ethnic backgrounds opened a much more nuanced vocabulary for defining skin colour. For example, they would use terms like “coffee coloured”, “pinky”, “light-skinned”, or “dark-skinned”. ‘Black’ was an important signifier of ethnicity for those who identified as Black, integral to their self-identity and a point of pride.

“Ethnicity and race they go hand in hand to me, so it's saying your colour, kind of, establishes more of who you are.” Black British, Black African, Black Caribbean, or Black Other, Higher Education

“I am more proud of the fact that I'm black than I am of being British… I put more significance on the fact that I'm black.” Black British, Black African, Black Caribbean, or Black Other, Higher Education

Some, however, were distrustful about how the data would be used by authorities. This distrust was most widespread among lower educated participants from both Mixed and Black groups. It led some participants to choose not to report their ethnicity depending on who was asking and for what reasons

“I think 'How do I identify?' Do I identify as a Black, British male? Do I identify as Black, African, because on account of, you know, my family being from Ghana? And then, I have to ask myself the question 'Well, why is it relevant?'” Black British, Black African, Black Caribbean, or Black Other, Lower Education

On the other hand, participants of White and Asian ethnicities were largely uncomfortable with the use of skin colour and saw it as an outdated classifier. Beyond the appropriateness of classifying someone by their skin colour, many participants questioned how useful skin colour classifications were.

“I don't understand why people need to know this information. To me, it just seems to be defining, what colour is your skin. Because you're probably White. I'm not White. But you're referring to me as the group that is White. And you're referring to the next people as Black and Asian. Why do you need to know?” White Other, Lower Education

“I don't understand why complexion is even relevant to even ask, if that makes sense? I understand that the census wants to keep track of the data, and that makes sense…if you're asking about complexion, why don't you just put a colour chart…It feels like a pointless exercise.” Other (British Cypriot), Higher Education

Participants saw colour classifications as potentially misleading proxies for ethnicity or, at worst, racist. Asian and Asian British participants expressed strong personal discomfort with the prospect that they might be identified in such a way.

“When I hear describing people as White or Black, I don't like the terms. It doesn't sound nice at all; I wouldn't like to be classed as a colour. You know? And, if somebody would put a label on me as a colour, they would say yellow. And I wouldn't feel good about that at all.” South East Asian, Lower Education

While White British and White Other participants did use skin colour to describe themselves, there was a degree of discomfort in the use of skin colour, both in self-defining and in classifying people of Black British, Black African, Black Caribbean or Black Other ethnicity.

“Ethnicity is where you're from, where your origins are, my origins are in England. Regardless of what the colour of my skin is, you know, that, to me, is that ethnicity is.” White, Lower Education

Despite these reservations, there was recognition from some participants that colour terminology has a purpose in some contexts. These participants were generally more educated or worked with data themselves. But many of these participants also felt that collecting data on skin colour should be confined to specific contexts, for example health research and when monitoring Diversity, Equality, and Inclusion outcomes.

2.2. National identity as an important ethnic signifier

Most participants referred to a place to describe their ethnicity. This could be a country, region, continent, or other geographic identifier. This was typically seen as the most important signifier of ethnicity in classifications.

Participants had different reasons for choosing to identify with specific places. They may choose to identify with a specific place because:

they were born there;
they grew up or spent their formative years there;
their parents, grandparents, or ancestors were born or grew up there;
they are a citizen of that place (e.g. by birth or naturalisation, as evidenced by having a passport);
they identified with the culture, values, customs, practices, and traditions of the place – this includes identifying with these things through their current relationship with members of that 'community'.

The ethnicity classification participants chose to identify with could vary depending on the perceived purpose of the survey or who was asking the question. For example, this could lead to nationality being used instead of national identity, or to a participant choosing not to answer the question.

Participants across multiple groups highlighted inconsistencies in the way the British label is used across different categories in the Census 2021 question. These inconsistencies have led some participants to feeling as though they could not or should not identify as British, even if they consider ‘British’ to be a core part of their ethnicity. For example, the fact that only one tick box explicitly mentions British, under the White category in a sequence including ‘English, Welsh, Scottish and Northern Irish’, means that:

some White participants born outside of the British Isles feel discouraged to identify as British, even if they hold dual nationality;
British people belonging to an ‘other’ ethnic group can only identify as British by using free text;
Asian British and Black British participants can’t distinguish themselves from others in their category who do not identify as British.

“I think the only improvement for me that would make sense, is if it's like British. The first one where it's English, Welsh, Scottish and Northern Irish, British, then there is like ethnic British as a second option. Then that captures all first-generation British people and people that fall into that category that wouldn't necessarily identify themselves as any of the below. So, yes, ethnic British I think captures a lot of people.” Other (British Cypriot) Higher Education

“You have many generations of people who have lived here their entire life. And to just be, 'You are still Indian, you are still Pakistani,' but that person has, like, their family have lived here for three, four generations now and to not class them as British is pretty poor.” White, Lower Education

Some participants in the East and South-East Asian groups did not self-identify as ‘Asian’. They associated this label with the Indian sub-continent instead. These participants would prefer a separate top-level category that distinguishes South-East Asian from other Asian groups. We have insufficient data to determine exactly what top-level category people of South-East Asian descent would most prefer beyond the current ‘Asian’ category.

Both South-East Asian participants not from China and Han Chinese participants who are not from China reported that the use of the Chinese label is too broad. They felt that the label may encourage the assumption that “everyone who looks like me, will be Chinese”. These participants preferred a more nuanced set of options, for example, reflecting a wider range of South-East Asian countries including Korea, Vietnam, and the Philippines. They also preferred to have options that reflected the existence of Chinese people not from China, for example, Chinese from Hong Kong.

In all these cases participants would typically use the ‘Any other Asian background’ open text box response option, which satisfied their preference to identify in a particular manner.

2.3. Pride in ‘Mixed’, or ‘multiple ethnic group’ identities

Participants of Mixed backgrounds often expressed a high degree of pride in their heritage. They viewed the ‘Mixed’ label as a byword for multiculturalism. They identified as ‘Mixed race’ and had minimal issues with classification labels like White and Black African, despite these being relatively broad.

Some participants who identify as Mixed and participants with children who are Mixed, identified that the tick boxes all assume a mix of White and something else. This can make it difficult for Mixed children of non-White parents, two non-British White people, or any other mixes, to self-identify without using the free-text option. White Arabic and Anglo-Indian participants felt there was a lack of relevant options for them. They noted, for example, that “not all Indian people are Asian”.

Participants from a Mixed background would often qualify their ‘Mixed’ label in discussions about their ethnicity by explaining where their parents were from. Some identified more strongly with one culture than another. This could be because of the local community in which they were raised, or coming from a single-parent household. This meant they might identify as, for example, Black Caribbean, even though they had more of a Mixed background.

“I just say that I'm Mixed race overall and then I go into what. So, specifically, I'm half English. No, so, I'm half Jamaican. I'm a quarter English and I'm a quarter Irish.” Mixed, Higher Education

“When it boils down to it, I'm Mixed-race or if it's between one of the two, I will be black. But, I'm just Mixed-race and I take culture from a lot of things.” Mixed, Lower Education

2.4. Inclusivity and the ‘Other’ category

The Census 2021 question that the participants saw, uses ‘Other’ in two ways. It is used as a top-level category, as well as within each top-level category as a response option. It includes a free-text option for respondents to write in their ethnicity if the tick-boxes do not present a good fit. Participants generally considered that the ability to write in their ethnicity increased the inclusivity of surveys by ensuring that everybody had an option to self-define according to their own preferences, regardless of the tick-boxes provided.

“I like the fact that they also give you space to write if you really want to express your identity, your ethnicity, a little better for people to understand…” 18-24 years and Mixed ethnicity, Higher Education

Participants across most groups acknowledged the impracticality of providing a tick box for every possible way of defining ethnicity as this would lead to a very long list. The ‘Other’ option was therefore seen as a practical middle-ground. But some participants suggested that it is inappropriate that relatively large and growing sub-groups like Latin Americans and Eastern Europeans do not have their own option.

“I don't feel offended but I do feel a bit left out. We (Latin Americans) are a big community in the UK, and I don't see why, so often, we are not mentioned. We have to tick the Other box.” Other (Latin American) Lower Education

2.5. Religion signifiers

“It's not relevant. I can't see the link between describing someone's ethnicity with religion, not at all.” South East Asian, Lower Education

“I think getting culture involved in forms overcomplicates things because it just becomes very subjective.” Mixed, Higher Education

Ethnicity was widely regarded as being fundamentally different from religion as it is something more fixed than personal beliefs, that cannot be changed ‘at will’. There was strong consensus across all the groups that labels referring to religion should not be used as part of the ethnicity question. Reasons for this included:

perceived differences between ethnicity and religion - ethnicity was seen as being relatively fixed and not changeable at will, unlike religion as a belief which can change;
religious beliefs transcend most conventional ethnicity classifications, so would make the ethnicity question impractical;
religion is private and personal, and so should not be collected by authorities;
religion can be a sensitive and divisive topic;
data on religion is often collected through its own question, and so should not be included in the ethnicity question.

Participants also expressed little support for including more tick-boxes to capture data on culture and heritage. While relevant to personal definitions of ethnicity, they did not see how this data could be efficiently captured within the ethnicity question, or its relevance in most public data contexts. Regardless, this data could be captured through other questions that ask more directly about religious beliefs.

It should be noted that, while quotas were used to ensure representation of different faiths, some faiths were represented by only two participants. This includes the Jewish faith.

3. Stigmatising language and its consequences

Conversations in the focus groups and interviews included how ethnicity terminology is used in general discourse, demonstrated by representations in the media such as TV, film, print news and social media. These conversations helped us explore the types of language participants perceived as stigmatizing and to understand the effects of such language.

3.1. Using ethnicity terminology divisively

Aside from obvious and widely recognisable derogatory language, there was little evidence that participants viewed the use of ethnicity terminology as stigmatizing . What matters most is the context in which it is used. Participants recognised that simply referencing ethnicity sets a tone and contributes to a narrative which may be more or less positive depending on the ideological leaning of the author or channel.

The most common perception, and criticism, from participants of the way ethnicity terminology was used was that non-White skin colour or broad ethnic groups were used as identifiers in ‘negative’ stories. These identifiers were often not relevant to the story itself. In these cases, participants commonly assumed the ethnicity terminology is serving a hidden purpose to trigger a reaction. This could be to act as clickbait, or to play to the prejudices of particular audiences and ‘othering’.

“I just think sometimes the way that some in the media might not just mention the facts, but also might say ‘Muslim gentleman’ or stuff like that, or they might say ‘he's just come back from his country’, do you know what I mean? Stuff that they wouldn't say if it was just a White British person. You don't really describe people as Christian males, but if you were to say a Muslim male, I think it's got different connotations built up.” Mixed, Higher Education

Participants believed that this behaviour reinforces stereotypes in the minds of certain types of consumers. They provided numerous accounts of stories purposefully playing on and strengthening existing negative stereotypes.

Participants also referenced stories about non-White British celebrities and prominent figures as especially prone to racism and discrimination. They observed that these individuals are routinely criticised to a much greater extent and held to different standards than White British equivalents.

“He [Rishi Sunak] is British born. So, what does it matter if he's got brown skin? But he's not accepted, he's not one of them. Do you understand what I'm saying? He's not one of them because he's a different colour, and they have to highlight that.” Black British, Black African, Black Caribbean or Black Other, Higher Education

“If Lewis Hamilton was White I know that the media would have treated him differently from the way that he's treated. There's a lot of examples of that. It's subtle, it's never overt and on the headlines.” Other (British Cypriot) Higher Education

Some participants imagined a similar story but featuring a White British person or group. If ethnicity would not be reported in this imagined story, then participants suggested that it should not be reported in the case of non-White British ethnicities either. That said, participants did reflect positively on examples of where their ethnicity was presented in a positive light, such as in films like Black Panther, or where headlines were informative, such as why Tyson Fury is called ‘The Gypsy King’. These were seen as a refreshing change to the way ethnic minorities were typically represented.

3.2. Using ethnicity terminology informatively

The use of umbrella labels, such as Black African or Asian, were seen by most participants as unhelpful, especially when more specific alternatives could be used. Of all the examples shown to participants, the label ‘people of colour’ evoked the most negative emotions. Among other critiques, participants indicated that these sorts of labels encourage division by ignoring the diversity and nuance of how people actually identify, to create binary “us vs them” distinctions.

“Everybody's a person of colour, right? Some colour. So, they say white is the default colour and everybody else is different. It's like someone saying, 'You have an accent.' It’s got a certain racist connotation. Maybe it's mild but it's still racist.” South Asian, Higher Education

Participants expressed a preference for ethnicity references to be as specific as possible. This helps to avoid groups being associated with the subject of the story where this had no relevance to the story itself. This could be achieved by:

avoiding umbrella terms such as ‘people of colour’;
being more specific in references to ethnic groups if relevant to the story – for example say Black Caribbean rather than just Black;
shifting the emphasis away from an ethnic group by using other non-ethnic identifiers.

“Even the fact that they say 'Black women in the UK are four times more likely to die in pregnancy and childbirth.' So, what are you trying to signify here, that black women can't carry kids as effectively as white women…Is it talking about black women from Africa, black women from the Caribbean, Black women from where? What is the message it’s trying to convey?” Black British, Black African, Black Caribbean or Black Other, Lower Education

“I mean, for me, names make it a bit more relevant and a bit more individual, based on the individual and based on their story… You're trying to provoke by having a whole ethnicity there when the story doesn't necessarily apply to that whole ethnicity.” Other (British Cypriot), Higher Education

3.3. Reinforcing negative stereotypes and biases through ethnicity terminology

Most participants from ethnic groups other than White British reported disengaging from narratives perceived as encouraging discrimination. Participants used terms like “clickbait”, “lazy journalism” and “low quality” to refer to them. This was typically the case when ethnicity was referenced, as these were perceived to be stories which were negative. Some participants said that while they may once have felt personally attacked, they have now deliberately given up reading these stories and disengaged with them or developed a degree of resilience when encountering these types of stories.

“I don't think it affects my sense of belonging, no. I think, as I said, I think it plays a part in reinforcing stereotypes, but I don't think it affects my sense of belonging.” Other (Latin American), Lower Education

Some participants found reviewing headlines which portrayed their ethnic group in a negative light disconcerting and upsetting.

“It [headlines referencing ‘Mixed race’] makes you feel divided from the country where you were born and raised. Divide and conquer, isn't it?!” Mixed, Lower Education

Most participants recognised they were not the audience these stories were being targeted at. The main initial negative effect was therefore on those who were the intended audience, particularly people already receptive to discriminatory and divisive rhetoric. This was perceived to reinforce negative stereotypes and encourage anger towards ethnic minority groups and people born outside of the UK. Participants felt that these portrayals would increase the likelihood that a person belonging to a stigmatised group would be the victim of racism and discrimination. This created anxiety at a personal level and concern for how their ethnic community was seen by others.

“I just feel really sad, and I just feel really helpless as well because I can't deny that I'm Chinese, but in a way, that's [the COVID-19 pandemic] got nothing to do with me, but you can’t actually totally disconnect yourself from that. You just really struggle, and you don’t know how to explain to people because obviously it’s not enough for you, for one person, to say, ‘That’s nothing to do with us’.” South East Asian, Higher Education

“If your group or your ethnicity or your religion is going to be targeted, then I guess it's a sense of when is it going to be you? Or when you're going to be targeted for something? So just the basically being unsure and quite wary of what you say or who you're around. Because you just never know.” 18-24 Years and Mixed ethnicity, Higher Education

In a small number of cases participants did reflect on very specific and tangible effects of negative stories, portrayals or social media comments which were associated with their ethnic group. These ranged from abusive language directed at all Asian communities after reports of child sexual abuse scandals in places like Rotherham, Telford and Manchester, to the loss of business amongst Chinese and Asian businesses because of stories which suggested that Chinese people were spreading the Coronavirus (COVID-19 infection.

RECOMMENDATIONS FOR GOVERNMENT

This research sought to understand how people from different ethnic groups self-identify their ethnicity, and the factors affecting this. We have some recommendations based on our findings.

For data collection

Ethnicity data should be reflective of how people want to self-identify and this should be made clear in the way the question is asked.
People should always have the option to complete a free-text response about their ethnicity.
Include caveats alongside the reporting of ethnicity data to explain that care should be taken when interpreting it because ethnicity is a multifaceted concept which can mean different things to different people and self-reported ethnicity may be influenced by when it is collected, by whom and for what perceived purposes.
Although not universally accepted, skin colour should remain a signifier in the classification mechanism for White and Black ethnic groups, but more effort should be made to show the relevance and usefulness of collecting this information in questionnaire guidance.
Skin colour should not be incorporated as a signifier for Asian or other ethnic groups.
Standardise the use of the label ‘British’ across the classification, or include the nationalities English, Welsh, Scottish, Northern Irish, or British in each of the ethnic groups and not just the White group.
Include growing communities, particularly within the White Other and Other categories such as Eastern European and Latin American.
Re-order the ethnicity classification alphabetically, to avoid creating a sense of importance and priority which is unhelpful in reinforcing feelings of prejudice and inequality.

For communications and reporting

Ethnicity should not be used as an identifier when reporting on a given story unless there is a clear and valid reason to do so.
Be aware of the potential for ethnic terminology to be used in a way which is stigmatizing or reinforces stereotypes.
Avoid using umbrella groups when reporting ethnicity and make references as specific as possible to the exact ethnic group of relevance to avoid confusion or misrepresentation.
Avoid using umbrella terms such as “people of colour” or "Black and Minority Ethnic (BAME)” as they are both too general to be useful and can be divisive.