Seminar presentation by BMRB
Ethnic identity and classifications of ethnicity
Vanessa Stone and Susie MacDonald, BMRB Social Research
Rowena Arshad, Centre for Education for Racial Equality in Scotland
Philomena de Lima, Inverness College
April 2005
Agenda
- Background and methodology
- Stage 1 - Data providers: Community groups and other stakeholders
- Stage 2 - Data users
- Stage 3 - Data providers: the general public
- Suggestions
Background
Aim
- To inform the development of a classification of ethnic identity which would be acceptable to individuals whilst providing data users with an approach that meets their needs.
Methodology
- Qualitative methodology
- Staged approach
- Stage 1 - representatives from community organisations
- Interviews
- Paired depth interviews
- One-to-one in-depth interviews
- Group discussions
- Fieldwork undertaken early 2005
Stage 1 - Representative organisations, stakeholders and community groups
Purpose of data collection
- Concern about lack of clarity of the purpose of ethnic monitoring - by both information providers and users
Reasons for ethnic monitoring
- As a fact finding tool
- Policy and planning tool
- To monitor progress
- As an awareness raising mechanism
Usage of data
- Little evidence of data being analysed or acted upon
- Lack of visible demonstration of how data is used (tangible outcomes) acts as a disincentive to further engage with any monitoring process
- Ethnic monitoring is more than profiling, it is about improvement of services and quality of life
Levels of data usage
- Distinction between national (Census) and local level use of ethnic monitoring data
- Recognition that Census provided large scale data which is not necessarily up to date or accurate but is nevertheless a broad planning tool
Local use of ethnicity data
- Local agencies were not drilling down from the national data for local relevance
- Local data needed to go beyond quantitative data to gathering really useful detail about needs and experiences
Dissemination of data
- None of the interviewees were able to provide examples of good practice in dissemination of either the analysis of use of data
Classifications
- Conceptually flawed and confused
- Not a single respondent was content with the current format of Q15
- Data from current categories seen as difficult to use in any meaningful or appropriate way
- Categories are divisive and exclusive- imposed configuration does not aid diversity or integration
- Particular concern for non-visible minorities
- (e.g. Gypsy/Travellers) as well as visible minorities who are small in numbers
Visible yet invisible…
- Original purpose for including an ethnicity question was to enable minority ethnic groups to be counted and made visible
- However, this is a qualified success as the current framework does not capture the real diversities within Scotland ( e.g. contributions of the Italian/Polish communities continue to be unrecognised in their own right within the Census data, where are people with a Middle Eastern background)
The 'other' category
- Concern that this category is expected to encompass increasing range of diversities
- 'Other' box seen as psychologically devaluing
- Concern about how data in this category was being analysed e.g. if someone classified themselves in the 'Other' category as Jewish, would those coding the data automatically assume that person was white?
- How is this category assisting to inform the ethnic categorisation question for the 2011 Census?
Colour and ethnicity
- Generally respondents found the colour coding of some groups and not others to be unacceptable
- Some felt that racism was complex, affecting visible as well as non-visible communities and was therefore not a matter of colour
- Overall respondents felt that questions on ethnicity and colour should be separated
Decoupling colour from ethnicity
- The significance of colour in addressing issues of discrimination was understood
- Generally respondents saw a need to retain some form of colour coding though some were concerned about a colour category being included
- Overall respondents wanted to see the 'black/white' application to some groups and not others removed
Respondent suggestions for improvement
- DECOUPLE COLOUR FROM ETHNICITY
- AVOID OPEN QUESTIONS /PROVIDE OPEN QUESTIONS
- USE BROAD REGIONAL HEADINGS AS DONE BY THE UNITED NATIONS e.g. European, Asian, African, Middle Eastern, North American, South American. Respondents could be asked to further self- define e.g. Middle Eastern and Turkish or African and Ethiopian. This allows people to have multiple identities
- Crude colour categories such as black and white were not perceived as helpful
- It was felt that a generic question about experiences of discrimination or a specific question about colour and discrimination was more helpful
Option 1
Do you consider yourself to experience discrimination on any of the following grounds?
Age Accent Colour Disability Ethnicity Faith/Belief Gender Sexual Orientation Social Class etc…
Option 2
Do you consider yourself to experience discrimination because of your colour?
Yes No
If yes, please say how…?
Improvements?
- Clarifying and communicating the purpose of gathering data on ethnicity
- Formal/official structures of the Census form do not assist those who are illiterate or have literacy issues to come forward easily. Accessibility issues might also apply to ethnic minorities who may not speak English as their first language
Other issues…
- Absence of a section on 'languages' other than Gaelic viewed as short-sighted. Being unable to speak English can result in discrimination. Why are community languages not valued?
- Need to expand the religious question - to include different sects of the other major faiths already cited on the form
- Generational differences might affect the way forms are filled- might affect data
- Future proofing required- the youth perspective
- Who decides on the categories in the Census form?
- Why are local data users using Census categorisations as the definitive framework - advice from the CRE, GROS?
- Unless change occurs nationally, local users are likely to continue to use the Census framework as their template
- Needs of small communities must not be ignored by the Census
- Gypsy/Travellers
- Literacy Issues
- Collection of Census forms from local households
- Undercounting
- Several respondents felt strongly that Scotland needed a Scottish solution
Key conclusions from Stage 1
- Purpose of ethnic monitoring needs to be better understood by data providers and users (at both national and local levels)
- There has to be tangible outcomes of a monitoring exercise and this information needs to be disseminated
- There is a need to decouple colour from ethnicity.
- Overall, it was felt there remains a need for a question that can extract issues of colour discrimination
- Overall, the current attachment of colour to some groups and not others was seen as divisive and unhelpful
- Ethnicity categories need to be broadened to include other white ethnic groups, other visible minority groups
Stage 2 - Data users
The sample
- Scottish Executive : • Office of the Chief Statistician • HR
- Education
- NHS
- Local authority
- Police
- Six paired depth interviews
- One "stats", one "policy" representative, as far as possible
Reasons for collecting data on ethnicity
- To monitor for discrimination
- To develop policies / look at impact of policies
- To set targets and inform outreach work
- To tailor services to needs of population
- staff, students, general public, prisoners and so on
- minority groups (for example, need for interpreters)
- identifying and responding to special needs
- To comply with requests for official statistics
- To raise awareness (eg stop and search forms)
How ethnicity data is used
- To see if extra resources are needed or to target resources more appropriately
- Comparing groups, large-scale versus localised
- large-scale - eg is our staff profile broadly representative of population as a whole?
• with Scottish population
• with UK population
- localised - eg how does the organisation operate
• staff across different sectors
- monitoring recruitment
- monitoring employee progress
• staff behaviour towards public
• to investigate allegations of racism
Local use of ethnicity data
- General view that data on ethnicity is more useful when can analyse significantly sized groups
frustration that numbers of people from minority groups is small in percentage terms and therefore need to aggregate statistics to carry out a meaningful analysis of the data
major concern that confidentiality is upheld and is of particular pertinence to Scotland
- desire to publish only aggregate statistics for fear of identifying individuals
- 'white', 'black', 'Asian', 'other' - four aggregate categories referred to
- however, comments indicated that confidential data was studied in more detail by data users, particularly those with an equal opportunities remit
Current classifications in use
- 2001 Census question (or very similar)
- because need to compare to census data
- enables joint initiatives to tackle problems
- Tailored to specific needs
- hence may ask series of questions to identify an individual's need for services / support
- developed own approach and have maintained it to allow for trend analysis
Views on current classifications they use
- fine, no complaints from data providers
- not perfect, could be fine-tuned
- would welcome a radical overhaul
- … but generally unsure if categories reflected best practice through lack of research among data providers
Views on current classifications they use
- Fine, no complaints from data providers
- view that minorities form small percentage of population and so no need to have more categories - be hard to analyse 'meaningfully'
- Not perfect, could be fine-tuned
- better understanding of 'white' minorities
• sense that 'white mass' hides minorities with needs eg Portuguese, eastern Europeans
- debate over how 'other' categories could be improved
• awareness that some people dislike and sometimes avoid the 'other' category and therefore can be mis-classified
• mixed views on whether a longer list of categories would be helpful or desirable (might create confusion) - 'it will never be perfect'
• but a belief that some people could benefit from being counted differently by widening choice of categories
• debate over how religion is handled - what constitutes an ethnic group?
Views on current classifications they use
- debate over nationality and / or colour
• view that some young people want to identify as primarily Scottish regardless of ethnicity so classifications without reference to nationality are 'flawed'
• comment that wrong for nationality to be used synonymously with colour - but also mentioned political concepts of colour
• Would welcome a radical overhaul
- encourage people to think about their identity
- suspicion that people respond to existing categories rather than reflecting how they really feel because they haven't thought about it before
The ideal approach
- perhaps a preference for not changing too much but a desire for 'getting it right' too
- Range of views on whether single or multiple questions
- Single question for simplicity and speed
- Multiple questions to capture detail and (possibly) overcome issue of imposing inadequate labels
• nationality and ethnicity
• main headings, sub-headings
• also mention of religion and language
- Generally a desire for a tiered or nested approach
- main headings, perhaps comparable with current approach to aggregation
- sub headings to provide more 'accurate' answers
- An approach that allows for comparability with previous records, even if data had to be collapsed
- Having an approach that better identifies changes in the population
- although concerns over capturing trends data
- Some desire for long list of pre-coded options so long as
- data can be easily aggregated
- confidentiality/anonymity protected
- not a deterrent to answering question
- Suggestion that the census focuses on main headings allowing organisations to 'drill down' as fits their situation/needs of local population
The ideal approach - categories
- Inclusion of all British nationalities, plus British and Republic of Ireland
- just having Scottish "screams inequality"
- also suggestion to have White European as separate code
- Some suggestions for adjusting existing categories and new categories, for example, Chinese, Travellers
- although debate over how some groups should be presented/defined and to what extent distinctions should be made
- … but generally looking to research with data providers to determine any new categories
Impact of change
- Concern over anticipated need to go back to many thousands of data providers to update records to allow for comparability
- Worries over disruption tempered if given plenty of lead-in time, concerns included
- changing expensive computer systems
• ideally any changes to ethnicity classification could be fitted into normal review periods
- allowing time for training and communication, including consultation exercise
- having budget to implement and manage the change
- … but general view that organisations would 'just get on with it'
- Be helpful if 'main headings' remain the same to aid trend analysis
- Comment that the stress should be on what works for data providers not administrative needs
Key conclusions from Stage 2
- Data users flexible about change but desire for something that could build on existing trends data
- A 'nested' approach to questioning
- Protecting anonymity a major concern
- General view that ethnicity question on 2001 census could be fine-tuned to provide a more sensitive approach
- Changes need to be sensitive to findings from research with data providers
Stage 3 - Data providers: the general public
Sample
- Diversity of ethnic backgrounds represented
• E.g.: White Scottish, Jewish, Mixed Scottish/South African, White South African, Black South African, Irish, Mixed Irish/Indian, Indian, Jamaican, Filipina, Gypsy Traveller, Showground Traveller, Bangladeshi, Persian, Chinese, Pakistani, Senegalese
- Diversity of ages represented
- Both sexes represented
- Two mini focus groups with 18-24-year-olds
- One with "ethnic majority" - White Scottish
- One with "White Other" people (e.g. Polish, Spanish)
What is ethnicity?
- Ethnic minorities
- Two main factors
- Nationality
- Skin colour
- Mixture of both these
- Religion
- On probing, factors such as culture, religion, language, attitudes also felt to be important
- Not all factors equally important to all people
- People vary in how specific they will be
- Can change according to context
Single, dual and multiple identities
- Some originating from outside Scotland with a single heritage feel they now have a dual or multiple ethnic identity
- Can cause confusion when attempting to classify
• "I'm black African but at the same time I can call myself black British "
- Some do not, even if now hold a British passport
- Especially if view ethnicity as based on nationality
- Conversely, some of those with a dual or multiple heritage see themselves as having a single ethnic identity
- Ethnic background of friends, relative influence of mother and father, accent, place of birth all important
Attitudes to being asked about ethnicity
- Happy in principle
- Happier for some organisations than others to ask
- Medical profession needs to know - genetic predispositions
- Some happy to give information
- Some less happy with more detailed classifications
- Less happy about being asked some things, e.g. religion
- Some unhappy with being asked at all
• Could encourage/facilitate discrimination
Reasons for being asked question and how data is used
- Not really sure why question is asked
- Not told - can guess but fairly comfortable
• Government needs it for statistics, service provision and tailoring
• Guarding against discrimination
- Although some suspect it is in order to discriminate
Recall of different classifications
- Overall, difficult to recall times have been asked about ethnicity
- Employment, school (parents, not YP)
- Even more difficult to recall details of classifications
- Detail of classifications vary
- Some feeling that shorter classifications better (White/Black/Asian) as less complicated
- Others would like more specificity
- Some recognition that ethnic frameworks have changed over time
- Black/Asian, British/Scottish
- Mixed
• "My sister who just noticed, and she was like oh, we can actually fit into something now, do you know what I mean? Because it was just so hard before and you just end up leaving it and things"
Strengths of different classifications
- "Mixed" - write-in option preferred, otherwise break down code
- "Other" - write-in option preferred, otherwise be more specific
- Two-stage question preferred by those who understood reason behind it
- Would help avoid confusion
• However, those who tended to confuse saw less need
- Allows people to state national identity and allegiance without compromising data on ethnicity
Weaknesses of different classifications
- Where nationality question is used, at a minimum should have all UK nationalities listed
- Important not just to have "British"
• "I am not British, I'm Scottish"
- Some also like option of "British"
• Some tendency to confuse "British" with "English" - language-based
- Include Ireland (ROI)
• Numerical, cultural, geographical arguments for doing so
- Issue of identifying geographical area with colour
- E.g. "Black African" - where would a white African go
• "The other one I put, and it said Black African, and I crossed it off, and I said White African, I remember doing that twice now."
- Sri Lankan under Asian
- African countries
• "Africa is very general, very broad"
• "Full list might be too long; could have N, E, S, W Africa as codes instead
The "Other" code
Different views:
- Not represented - you don't fit in anywhere, get angry or leave it
Vs.
- Assume categories are based on numbers, do not mind being unrepresented
"I just assumed that not very many Filipinos are in the community, or in this country, that's why I classified as other."
- Prefer longer classifications to enable greater specificity
Vs.
- Prefer concision at risk of not being included
"It's really a few in the country, like other nationalities, so if you're collecting data its better to have just 4 options, and then just specify the others, so I think that's fine."
Key conclusions from Stage 3
- Ethnicity means different things to different people
- Ethnicity can imply minority
- Lack of understanding about why ethnicity data is collected
- Advantage to know if one group in society is having problems so they can be addressed
- Not overly concerned about whether or how asked
- Those who did have an opinion saw benefits as well as drawbacks
- Division between those who wanted long, specific list vs. those who wanted concision
- Nationality is clear, colour more debatable