Radical
Statistics
Count me in?
Joanna R. Southworth
Introduction
The 2001 Census will give us greater detail about the cultural make-up of Britain than any previous census. In the years prior to the 2001 census discussion occurred as to whether and how to alter the 1991 ethnic group question to reflect the changing nature of the British population. In addition a proposal was made to include a new question on religion. In the event both of these changes occurred. As such when the results of the 2001 census emerge we should know more about, for example, the mixed ethnic population in this country, the Irish population in Britain, and affiliation to broad religious groups. Yet, whilst the 2001 census will provide the government, both local and national, as well as businesses, organisations and social scientists, with some potentially valuable information, the picture presented is far from complete. Taking the examples of the ethnic and religious questions on the 2001 census in England and Wales this paper considers briefly some of the groups of people who are not identified by the census. Prior to this, however, a moment is taken to consider why people might wish to be identified by a separate box in the census.
Why is it important to be counted?
To be counted in the census entails social and political visibility for the group(s) concerned. For example some Muslim organisations who campaigned for there to be a question on religion in the 2001 census referred to the fact that should a move would give official recognition as to the importance of religion as a factor in the social life of Britain (see Southworth 2001). Also, speaking on America, Prewitt (1987: 270) stated:
being 'measured' is to be politically noticed, and to be noticed is to have a claim on the nation's resources... political visibility follow[s] on the heels of statistical visibility.
Therefore being counted on a statistical survey, of which the census is probably the most comprehensive and well known, can lead to social, political, and, often in turn, economic recognition. The census itself also has a somewhat unique status of being regarded as an official provider of definitive statistics, which form the denominator for many other data sources. Therefore the census can become the de facto standard (Jones 2000). The extent to which such a reputation is deserved is the topic of another paper. Suffice to say the standard set by the census is frequently emulated by other organisations, as has been demonstrated by the form and content of many equal opportunity forms that have arisen since the inclusion of a question on ethnic minorities in 1991. Thus, to be counted in the census often entails representation in many other areas as well.
How invisibility creeps in
However, if being counted in the census promotes social and political visibility, it follows that the opposite may also be true: that is, being excluded from the census may lead to social and political invisibility. There are two ways in which people can be made to feel invisible by the census: being placed in too broad a category, and having no category. These two scenarios are considered in turn using examples.
1) Categories too broad?
The first people rendered invisible by the census are those who can choose one of the designated boxes on the census form, but feel it masks, or does not accurately represent, who they are. As regards the ethnic question, the Welsh have exemplified this situation most prominently in the run-up to the 2001 census. The 2001 census ethnic question in England and Wales sub-divided the White group into three separate categories - 'British', 'Irish', and 'Any other White background' - marking a change from 1991. This decision was due to a long-fought campaign by Irish groups to be identified separately. Strong evidence was presented, for example by the 1997 report from the Commission for Racial Equality (Hickman and Walter 1997), as to the incidence of discrimination against Irish people, which made it pertinent to record them as a separate group. Such a move was also strongly supported by users of census data, including local government departments (Aspinall 1996). For example Manchester and Bolton had already incorporated a separate Irish category into their own census forms, but usage of this data had suffered from a lack of denominator data, such as would be provided by the census. Trial forms were produced during the testing phase that sub-divided the 'British' category into 'English', 'Welsh' and 'Scottish'. However the reasoning behind such a move was to make the question more acceptable to Irish people, and it was later changed because of the complexities that would arise in having to potentially offer such options to other ethnic groups as well, and due to lack of user requirement for such distinctions.
The decision to only single out the Irish as a separate ethnic group may have gone through smoothly had it not been for the last minute announcement by Scotland, which under devolution devises its census separately from England and Wales, to include a 'Scottish' category. The lack of an equivalent 'Welsh' category in the census in England and Wales sparked outrage from many Welsh quarters, and a Census Reform Group was formed, supported by the Western Mail, to ensure a separate Welsh box would appear on the census form. This resulted in a 10,000-name petition being given to Tony Blair in early December 2000. The request was rejected, however, as the census forms had already been printed and costs could run to £ 2 million to reprint. In response Len Cook from the National Statistics Office said £ 1 million would be spent on an advertising campaign highlighting the fact that people could write 'Welsh' in the other box, along with the production of a special report on this group from the 2001 census. Such moves were not regarded as adequate however, and in the event Plaid Cymru produced 200,000 stickers for people to place on their forms to declare their Welsh identity, though questions over their legality and ability to be processed have been raised. In addition, on census day a further group of protesters posted their uncompleted forms in a coffin ( Anon 2001a). It remains to be seen whether and how the Welsh will be counted.
A similar situation of the census rendering groups invisible by encompassing them within a broad category arose concerning the question on religion The move to have a question on religion on the 2001 census was supported by the majority of faith groups. Whilst they succeeded in securing this, they did not obtain their desired format. It was strongly recommended by the Religious Affiliation Sub-Group, formed to discuss the question by the Office for National Statistics, to sub-divide the Christian category into component groups. The reasons given included that denominational differentiation was well developed in Britain and so it would have a bearing on many social policies such as education and abortion, and affect lifestyle choices such as smoking and alcohol use, which are linked to social services and health care provision (Francis 1998). The request was rejected, however, on the grounds that there was no user requirement for such a division, and the main need for the religious question was to provide a finer description of ethnic minorities in order to target services more effectively (Office for National Statistics 1998). Opinion seems to vary as to whether such a division in England and Wales would have in fact been useful. However, without it the religious question will primarily give data only about minority communities and not the rest of the population (1).
2) Categories omitted?
The second group of people that are made to feel invisible by the census are those who do not fit into any box and so are relegated to one of the 'other' categories. The 2001 census has alleviated for some the dilemma of where to place oneself as regards the ethnic question. For example, in 1991 178,000 people marked the Black-Other box, 58,000 of whom wrote-in 'Black-British'. As a result 'Black-Other' constituted 5.9% of all minority ethnic groups (Owen 1996) (2). At the time of the 1991 census there did not exist a developed sense of an Asian-British identity, but as the number of second-generation migrants has increased and grown older a parallel identity to that of Black-British appears to be emerging. In addition the 1991 census ethnic question did not offer a section for those of mixed ethnic background. The issue of how to categorise those of mixed racial or ethnic background was considered in both Britain and America over the past decade. Much controversy surrounded the topic in America, whereas in Britain the change was more smooth, attention instead being focused upon the religious question. However, the 2001 ethnic question, whilst acknowledging that a multiplicity of backgrounds exists, has designated boxes that only accommodate 'White and Black Caribbean', 'White and Black African' and the broad 'White and Asian' mixed ethnicities. The remainder are contained within 'Any other Mixed background', yet again being ascribed as 'other'. Such a move does not accommodate those of mixed European background, who may not consider themselves solely 'White', or many of the other myriad of backgrounds people living in England and Wales encompass.
As regards the religious question, as well as including a box labelled 'none' for those who want to opt out altogether, there is also an 'other' box. The Multi-Faith Directory (Weller, 1997) recognises nine major religions in Britain: Bahá'í, Buddhist, Christian, Hindu, Jain, Jew, Muslim, Sikh and Zoroastrian. The 2001 census question, however, only acknowledged six of these, excluding Bahá'í, Jain and Zoroastrian and forcing them to use the 'Any other religion' box. Many members of the Religious Affiliation Sub-Group viewed this as a serious omission, and even if members of these faith groups do provide write-in answers, it is unlikely that figures for the individual groups will be available, and so they will remain invisible as regards the religious picture of the country. They were excluded due in part probably to space considerations, but also because it was not felt it was important to know figures for these groups. The other category that people are petitioning to be recognised is Jedi. This is an idea that stemmed from the most recent New Zealand census, emerging from an email circular which said: 'Send this on to all your friends and tell them to put down "Jedi" on their census form... do it because you love Star Wars, If not.... then just do it to annoy people'. Rumours suggest that many people may have opted for this decision, and it remains to be seen whether the government will acknowledge it. However it does suggest that whilst many want to get their religion noted, others are perhaps registering protest at such intrusive questions. The extent to which such feelings are true may be demonstrated by the number of people who do indeed answer Jedi, or who opt to not answer the question altogether, a move made possible by the voluntary nature of the religious question.
A change in the nature of the census?
There appears to be a conflict arising between user perceptions of the census and those who answer it. If you were to ask a hundred people what their ethnic or religion was, straight off, free choice, no boxes, you would likely get almost as many varying responses. Some decisions have to be made regarding what categories to employ. If it is borne in mind who the census is ultimately designed for - the government - then the decisions made about what categories to use within these questions reflects a fairly narrow margin of interest of census users, government and the Office for National Statistics. The main official purpose of both the ethnic question and the religious question is to identify and monitor incidences of discrimination and disadvantage. As such previous similar questions on country of birth, along with the 1991 ethnic question had overtones of asking 'What are you?'. In addition, due in part to considerations of cost and space information is collected on a need-to-know as opposed to interesting-to-know basis. However, the perception of the census has been changing slowly, and now for many who complete the form the ethnic and religious questions are more closely linked to identity, asking 'Who am I?'. This is reflected in the support for a religious question by some Muslim organisations who felt that religion, not ethnicity, was the true expression of their identity (Southworth, 2001).
Conclusion
Overall, despite the expansion, slightly, of the ethnic question, and the introduction of the religious question, there are still likely to be many people ticking 'other'. Even though the choice available to people has increased, so too has the complexity of our backgrounds, and is likely to continue to do so. The advertising campaign for the 2001 census used the bold declaration 'Count me in' to inspire people to complete their forms. Even if people completed the form, many are likely to have felt that the census did not count them, at least not in their own terms. Although for the purposes for which the census was initially intended these finer distinctions might not have been particularly relevant, increasingly it is going to be difficult to ignore the changes that are happening in England and Wales. If the census is to retain its status of information provider then it might have to find ways to embrace greater multiplicity and provide options for expressing this. In doing so it will have to perhaps venture further into the more 'fuzzy' area of identity. Whilst this may be harder to deal with quantitatively, less oriented towards practical policy concerns, and of more interest to social scientists than government, it may be a necessary move in order to ensure the census continues to yield relevant and quality information. In the advent of devolution, awareness of Scottish, Welsh and English identities have become heightened and this has not been taken into account by the census, focusing as it does primarily on minorities. In turn, ethnicity is becoming more complex. Recent estimates suggest that 50% of all black children born in Britain have one white parent, meaning Britain has one of the highest rates of inter-racial relationships in the western world (Anon 2001b). This can render the notion of having an ethnic group difficult, even redundant. For example, in a recent Radio 4 programme addressing the issue of people who tick 'other' in the census, the narrator, Chris Simpson, had a Rwandan Greek mother, and an Irish father. Initially growing up in Dublin he moved to London whilst still a child and has remained there. In his words: 'I consider myself to be a Londoner. I'm not English or British' (Simpson 2001). In turn, the emergence of the religious question was largely due to pressure from Muslim organisations, which argued that defining themselves by ethnic group was meaningless and out-of-date (Southworth 2001).
Mixed ethnic, Welsh, Londoner, Muslim, Jedi - whatever terms we use to describe ourselves, it is apparent that we are becoming an increasingly ill-defined population as regards our ethnicity and religion. As a participant in the Radio 4 programme stated:
I think most people in the world are mixed... There are many more people like us than we think. I think I am on the periphery of the census. That doesn't mean I'm on the periphery of human life.
If the census wishes to continue informing us about the state of the nation's population, and not become a survey on the periphery, then in future decades it may well have to find better ways of representing the diversity of this country, and give statistical and social visibility to a greater number and variety of people.
Notes
- It should be noted that in Scotland there were two questions: the first asked about present day religious affiliation, and the second enquired about what religion people were brought up in. In addition the Christian category was sub-divided into 'Church of Scotland', 'Roman Catholic' and 'Other Christian'. This is to reflect the importance of these denominational distinctions in Scotland.
- The remainder of the 'Black-Other' category who did not write in 'Black-British' mainly had one black parent, the other being either white or from another minority group (Owen 1996). It is also interesting to note the variance in median age between the 'Black-Other' category, at around sixteen, and the 'Black-Caribbean' and 'Black-African' categories at mid to high twenties. This also suggests that those in the 'Black-Other' category were more likely to be second generation migrants (Diamond 1998).
References
Anon (2001a) 'Warning over Welsh census stickers' BBC news website : news.bbc.co.uk/hi/english/uk/wales/newsid_1298000/1298826.stm
Anon (2001b) 'Mixed race, mixed feelings' BBC website: www.bbc.co.uk/worldservice/people/highlights/010831_mixedrace.shtml
Aspinall, P. (1996) The development of an ethnic group question for the 2001 census: The findings of a consultation exercise with members of the 2001 Census Working Subgroup on the Ethnic Group Question Unpublished manuscript.
Diamond, I. (1998) 'The Ethnic and Religious Questions in the British Census: a Symposium - Introduction' Patterns of Prejudice 32 (2) pp.3-4.
Francis, L.J. (1998) A religious question for the 2001 Census? Should the category 'Christian' be subdivided into denominations? Unpublished manuscript.
Hickman, M.J. and Walter, B. (1997) Discrimination and the Irish community in Britain: A report of research undertaken for the Commission for Racial Equality London: Commission for Racial Equality.
Jones, G.C. (2000) Personal Correspondence 13th January 2000.
Office for National Statistics (1998) Summary points about the question on religion ONS Census Division 20th May 1998 Unpublished manuscript.
Owen, D. (1996) Towards 2001: Ethnic Minorities and the Census Coventry: Centre for Research in Ethnic Relations, University of Warwick.
Prewitt, K. (1987) 'Public statistics and democratic politics' in W. Alonso and P. Starr (eds.) The Politics of Numbers New York: Russell Sage Foundation.
Simpson, C. (2001) Other Radio 4 Friday 20th July 2001.
Southworth, J.R. (2001) The Changing Nature of the Census? Counting race, ethnicity and religion in Britain and America Unpublished PhD manuscript University of Bristol.
Weller, P. (ed.) (1997) Religions in the UK: a multi-faith directory Derby: University of Derby in association with the Inter Faith Network of the United Kingdom.
Jo Southworth
Lecturer
School of Geography and Environmental Sciences
Birmingham University
Birmingham B15 2TT
Tel: 0121-415-8022
E-Mail: j.southworth@bham.ac.uk
