Scottish Government

EXECUTIVE SUMMARY

Background

1.1 The Census currently collects data on ethnicity by asking respondents to choose (from a list) the ethnic background they feel best describes them. However, there has been much debate about the purpose of gathering data on ethnicity, and in particular the use of terms such as 'black' and 'white'; inconsistencies in category descriptions; and the representation of different communities, particularly small communities. Margaret Curran, former Minister for Communities, committed to re-evaluate ethnicity classifications to ensure that they reflected the diversity of Scotland's communities, and gather more meaningful information to better promote race equality. In 2004 a collaborative team from BMRB Social Research, CERES and UHI PolicyWeb was commissioned to carry out research into ethnicity classifications with a view to exploring these issues among both data providers and data users.

Research Aims

1.2 The overall aim of the research was to inform the development of a classification of ethnic identity; ideally one that would meet a variety of needs. Consequently, the research needed to explore how individuals would wish to classify their ethnic identity ('data providers'), whilst looking at the informational needs of those using such data ('data users'). It should be noted that one option was for the 2001 census question on ethnicity not to be changed at all.

Research Design

1.3 The research was designed in three stages. The first stage involved interviewing stakeholders with an interest in the topic and looked more broadly at the issues to help inform subsequent stages of the research. The second stage was with data users. This explored ways in which data users were collecting ethnicity data, why they did so and what use they made of the statistics they collected. This phase also looked at alternative methods of classification and the implications of changing methods. The third stage was with data providers and involved exploring how people define themselves (exploring different contexts and reasons), the extent to which definitions may change according to the context the data provider finds themselves in, and reactions to different kinds of framework.

1.4 The methodology was wholly qualitative for all three stages. A mixture of face-to-face and telephone depth interviews was undertaken for the first phase (11 interviews); face-to-face paired depth interviews for the second phase (6 interviews, 12 respondents); and face-to-face depth interviews and mini focus groups for the third phase (28 interviews and 2 focus groups; a total of 39 respondents).

1.5 Fieldwork was carried out in both urban and rural areas across Scotland. Respondents from a mixture of ethnic backgrounds were interviewed, with interpreters being used where necessary. A systematic and comprehensive analysis of the qualitative data was then undertaken.

PHASE ONE - The views of stakeholders

Purpose of Ethnic Monitoring

1.6 Respondents identified four reasons for ethnic monitoring:

• as a fact finding tool;
• to enable the delivery of equality of opportunity;
• to monitor progress; and
• as an awareness raising mechanism to give visibility to the diversity of ethnicities currently in Scotland/ UK.

Usage and dissemination of data

1.7 Stakeholders perceived that data gathered on ethnicity could be used to better effect than appeared to be the case. Concerns were expressed that public bodies and local authorities were not utilising the statistics available to them to improve services and equal opportunities/access.

1.8 Respondents wanted to see more detailed information being collected on ethnicity/identity to enhance service provision for different ethnic groups, particularly at the local level to supplement Census data, which could be quite out of date. There were worries that certain communities being excluded because they did not constitute a large enough percentage of the population, either locally or at the national level.

1.9 Some stated that ethnic monitoring data collected by larger bodies should be made available to smaller agencies, particularly in the voluntary sector. There was recognition that issues of data protection and confidentiality were aspects to be considered, but it was felt that data could still be disseminated whilst adhering to data protection legislation.

1.10 It was thought to be critical to cross-reference data, to take into account the needs of language or religion, for example. It was also suggested that there needed to be connectivity between Census data and data gained from other surveys.

Classification categories

1.11 There was a broad consensus among this qualitative sample of stakeholders with regard to the current classifications used on a number of issues:

• The current classifications were felt to be both confusing and inconsistent (particularly in relation to colour and ethnicity) hiding the real diversities within Scotland and not allowing people to make their 'Scottish-ness' explicit if they so wished.
• The current ethnicity framework was thought to do little to promote community cohesion by marginalising the identities of non-visible minorities as well as small visible minorities.
• Concerns over the "other" category. The range of ethnicities that this category is expected to encompass was felt to be very large, relegating new communities to 'other'. For some groups constantly ticking the 'other box' was seen as ' psychologically devaluing'. It was also felt that information in the 'other' category was not being used to address needs or to inform future categorisation systems.

The absence of a section on 'languages'

1.12 It was felt to be important to include a question on languages spoken, partly because language was very much part of a person's identity but also to tailor service design and delivery such that people would not be discriminated against due to a lack of English skills.

Generational differences

1.13 Better account needed to be made of differences between older and younger people (from minority groups) when describing their ethnic identity as there could be wide variation between the generations with reference to 'Scottish-ness' / 'British-ness' and country of origin.

Suggestions for improvement

1.14 Colour to be removed from ethnicity classifications, and perhaps asked as a separate question, with the proviso that a crude colour categorisation of 'White/Black' would not be acceptable.

1.15 For ethnic categories to be inclusive of all visible and non-visible communities, irrespective of their size - enabling everyone to have a chance to be identified in a way they felt comfortable with.

1.16 Using a framework that categorises people under broad regional headings, such as 'European', 'Asian', 'African', 'Middle Eastern', 'North American', 'South American'. Coupled with additional questions, this would allow for multiple identities to be expressed and would unpack diversities in the 'white' category.

1.17 Extending the choice of categories within the ethnicity framework to allow people to better describe their identity, but in a way that allowed for data to be aggregated up for national comparisons. However, some respondents did prefer an 'open' style of question leaving people to describe their identity in their own words (others had reservations about how such data could be collated).

1.18 The purpose of collecting ethnicity data had to be better communicated in order to ensure co-operation from both those who are suspicious and those who do not see the need for collecting such data.

1.19 It was suggested that the religion question could be modified to include the different sects/denominations of the other major faiths already cited on the form.

1.20 Respondents felt strongly that Scotland needed to have its own national ethnicity statistics, which could be broad and diverse but could also be aggregated to enable UK wide comparability. It was thought that the new framework could usefully draw on international ethnic monitoring models. There was also thought to be a need to learn how other countries utilised data gained nationally for local relevance and consequence.

PHASE TWO: data users

Reasons for collecting data on ethnicity

1.21 Respondents gave a number of reasons for collecting data on ethnicity: compliance with legislation; monitoring for discrimination; devising policies to promote equal opportunities; tailoring services; raising awareness; targeting resources and responding to requests from other organisations.

How ethnicity data was being analysed

1.22 Organisations described using standard statistical techniques to compare groups within the data-sets they were using. This could occur at two distinct levels: large-scale and localised. Large-scale comparisons meant comparing their population with that of the Scottish and/or UK population as a whole. Localised analysis meant respondents were looking more closely at how their organisation operated. Another approach was to look at staff behaviour towards the general public by collecting data on ethnicity during interactions.

1.23 It was commented that when using current classifications the numbers of people from minority ethnic groups was very small when compared to the Scottish population as a whole, especially outside Glasgow and Edinburgh. Groups could therefore be too small for statistical analysis. Respondents aggregated groups to help with this, for example, into groups such as 'Asian'.

Dissemination of statistics on ethnicity

1.24 Respondents were extremely mindful over issues of anonymity; they feared identifying individuals and the potential for harm that could result from this.

1.25 Ethnicity data was shared for a variety of reasons:

1.26 Overall there was a sense that data dissemination outside an organisation was in an 'official' capacity and in response to external requests. There was nothing to suggest that data users would be averse to disseminating data so long as confidentiality needs were addressed as this was a major concern for respondents.

Current classifications being used by data users

1.27 Data users interviewed for this study described using three different types of classifications for a variety of reasons:

• The 2001 census question on ethnicity (or something very similar) so that comparisons could be made with national data and also to enable joint initiatives
• Ethnicity classifications tailored to the specific needs of their organisation
• Ethnicity classification developed many years ago that were maintained to ensure compatibility over time.

Data users' views on the ethnicity classifications they were using

1.28 Generally respondents thought that the classification currently used by their organisation was useful but could benefit from being improved or fine-tuned. This view stemmed from uncertainty over whether the current ethnicity categories reflected best practice.

1.29 A number of specific issues were raised in relation to current classifications:

The impact of changing to a new ethnicity classification

1.30 A number of concerns were raised by data users in relation to any changes to their current systems that might:

• Inhibit trend analysis
• Disrupt the organisation, in particular their computer systems
• Be expensive

1.31 However, data users generally thought they would cope with such changes and stressed that the most important thing was for data providers to have a framework that worked for them.

The ideal approach

1.32 A range of suggestions were made for improving and adding to the Census question on ethnicity:

1.33 Data users were keen for any new approach to better identify changes in the Scottish population. For some this meant having a much longer list of pre-codes (ethnic groups/categories) to choose from, which it was hoped would allow people to be more accurate in their responses. Others suggested that having a better understanding of a changing population would mean looking more closely at people's responses to the 'Other' categories.

Phase Three: views of data providers: perceptions of ethnicity

Definitions of ethnicity

1.34 Respondents defined ethnicity more generally in terms of a person's background or identity. It was clear that many respondents had not thought about this before in any great depth. Consequently, their thoughts on how they defined themselves developed over the course of their interview, with subtle differences emerging on how they perceived the term ethnicity in relation to themselves and others.

1.35 Whilst both race and colour were mentioned by some respondents, ethnicity was generally related closely to nationality. Religion, accent, culture or language could also have a bearing but this was variable with ethnicity being simple for some and more complicated for others.

Colour in relation to ethnicity

1.36 Colour was a contentious issue for some respondents but not others. There were people who saw colour as very much tied to their ethnicity whilst others felt strongly that it should not be, either for themselves or for others. Such people expressed fears over stigmatising people because of their skin colour and a few respondents commented that such associations could also be misleading.

Single and multiple ethnicities

1.37 Generally, respondents in this study described themselves as having multiple ethnic identities and drew on a number of reference points, which varied from person to person, these included: parents being from different ethnic groups, countries or cultures; where they were born; where they were brought up; where they currently live (or have lived); their citizenship/passport; religion; and finally a range of life experiences.

Descriptions of identity

1.38 People referred to a wide variety of descriptors, underlining the complexity of trying to establish any kind of ethnicity framework. For some people there was a clear tie-in with the 2001 Census categories and how they wished to describe their ethnicity. However, for others the categories did not capture the level of specificity they might have wanted or needed. Having the opportunity to write in their answers helped some respondents when faced with the 2001 Census form.

Fluid and evolving identity

1.39 A number of the respondents taking part in stage 3 were not born in Scotland and they talked about developing a sense of Scottish-ness for a variety of reasons, for example because they had been living in the country for a number of years or because of other factors like their children being born in Scotland. Changing contexts were also explore but generally people either did not change how they described themselves or occasionally varied what they said because they were in a specific situation. Many respondents commented that the way in which they referred to their ethnicity was unchanging.

Other comments

1.40 It is worth noting that a few comments were made by respondents in stage 3 that suggested that people defined themselves according to the labels or categories they were familiar with. This might be because they had not thought about their ethnicity or identity in any great depth, perhaps because they did not see a need to.

Attitudes towards data collection and different classifications

1.41 In general respondents found it difficult to remember occasions when they had been asked about their ethnicity, although they remembered being asked about it for employment and in some other contexts. They found it more difficult to recall specific detail about the frameworks used, although some were aware that certain elements had changed over time, for example the addition of a 'Mixed' category or descriptions like Asian or Black Scottish.

1.42 Respondents varied in their reactions to being asked about their ethnicity, but generally speaking were happy to answer the question. Others were more reticent, feeling that it depended on who wanted to know, their reason for collecting such data and the way the question was asked. There were fears that the information might be used to discriminate. Some respondents worried about more detailed categories, wondering why an organisation would need so much information; others found broad categories pointless.

1.43 Respondents thought that organisations might want to know about ethnicity for the following broad reasons: for information purposes, including monitoring of diversity; to enable the most efficient allocation of services or resources; or to serve other agendas.

Strengths and weaknesses of different classifications

1.44 As the concepts being discussed were very abstract, three different methods of classifying ethnicity were shown to respondents during the interviews and group discussions.

1.45 Some respondents felt that the categories used in the example classifications were the wrong ones and proposed their own. However, others suggested ways in which the approaches adopted by the example classifications might usefully be combined in order to make a workable framework.

1.46 Generally, the preference was for a national identity question, with the current Census categories forming a second part asking about ethnic background. Respondents liked the opportunity to emphasise their national identity as well as their ethnic background, instead of their ethnic background alone.

1.47 It was thought to be important to have a wider 'white' background section and to have the opportunity to describe 'other' as necessary.

1.48 Respondents fell into three broad types in terms of the amount of specificity they wanted: those who wanted a short classification framework, which they could fill in quickly, and were willing to accept lack of detail and clarity; those who acknowledged the theoretical benefits to greater specificity of categories, but felt a line had to be drawn somewhere in terms of numbers; and those who wanted specificity to a potentially infinite degree.

1.49 Respondents differed in their attitudes towards the 'Other' category. For some, being forced to use it could be an annoying experience, which made them feel excluded and people would sometimes select categories that did not describe them adequately just to avoid ticking 'other'. Others assumed that the categories on the frameworks were based on population sizes and did not therefore mind being 'unrepresented' if they were part of a numerically very small community.

1.50 It was thought to be important to list the different UK nationalities if a nationality question was given. Many thought that the Republic of Ireland should be listed with these, for cultural, numerical and geographical reasons.

Conclusions

1.51 The stakeholder respondents, data users and data providers recognised that data on ethnicity was collected for a variety of reasons. Data providers and users in particular saw the validity of collecting ethnicity data. Stakeholders believed that collecting ethnicity statistics had to move beyond compliance to contributing to equal opportunities. They also thought it was vital that ethnicity data had local relevance.

1.52 Data providers and stakeholders commented on the lack of evidence showing how ethnicity data was being used. On the other hand, data users commented on how data was helping them to improve practice. There was a distinct gap between these two views.

1.53 In general, it was recognised that the current classification systems would benefit from adaptation to better record increased diversity. There was recognition that some data providers might resent and avoid using the 'other' category. However, respondents were unclear whether a longer list of categories was going to be helpful or desirable.

1.54 Stakeholder respondents, data providers and users also recognised the fluidity of concepts like ethnicity and identity. Stakeholder respondents found the conflation of characteristics such as nationality, ethnicity and colour to be highly problematic. They also viewed the current ethnicity question within the Census as being conceptually flawed and confusing. Stakeholder respondents in general saw the need to decouple colour from ethnicity or nationality, but the issue of colour was rarely mentioned by data users. There were mixed views among data providers. Nationality, however, was not contentious. A number of the respondents, from all of stages of the research, were clear that colour was still a clear trigger for racial discrimination and had to be addressed. Consideration may therefore need to be given as to how 'colour' is included within an ethnicity classification framework.

1.55 Religion was mentioned in passing by the stakeholder respondents and data users. The stakeholder respondents suggested there was a need to expand the religion question to include the different sects/denominations of the other major faiths. Data users on the other hand queried the relationship between religion and ethnicity. For some data providers, religion was perceived to be an important facet of their identity, but for others questions relating to it represented an intrusion into their personal lives.

1.56 Stakeholder respondents raised the importance of language as part of identity but also the lack of being able to communicate in English as a potential source of exclusion or discrimination. Data providers also talked about language being part of ethnicity.

1.57 During the research, the following were identified as areas to focus on for future consultation: