A review of usage of these sometimes overlapping terms can usefully inform the debate. Several scientists have observed a relative increase in the use of the term 'ethnicity' rather than 'race' in the medical science literature from the mid-1960s to 2000. 'Ethnicity' emphasises cultural differences while 'race' is a term frequently used to describe pan-ethnic population groups differentiated primarily by markers of visibility.
According to one recent analysis, 'ethnicity' overtook 'race' in 1991-95 in the USA. The same switch occurred in the rest of the world between 1976 and 1980. In a corpus of more than five million digitised texts prepared by Harvard University and Google Books, the frequency of the word 'ethnic' exceeded that of 'racial' around 1980 in UK-published books and in the mid-1970s in US-published books. Its displacement has been more marked in UK books.
Professors Afshari and Bhopal suggest that the rapid rise of the compound term 'race/ethnicity' in the US medical literature is an interim step whereby race will be superseded by ethnicity and that race '…may soon be a relic of history'. Among social scientists the view is emerging that race is a social construct whose use should be limited to studies of racism and discrimination, and the history of race science.
The use by population censuses of terminology like 'ethnicity' provides a 'gold standard' and acts as an important catalyst for adoption across government, statutory bodies, and elsewhere. Indeed, the census classification is now used in virtually all ethnic monitoring and administrative data collection.
In the UK, the conceptual base in the 1991, 2001, and 2011 censuses has been 'ethnic group'. The UK's Office for National Statistics (ONS), for example, acknowledges that ethnic identification is a multifaceted and changing phenomenon. It may include aspects of ancestry, country of birth, nationality, language spoken at home, religion, culture, skin colour, and national/geographic origin.
By emphasising the subjective nature of ethnicity, the ONS argues respondents will interpret these dimensions in ways relevant to them. But they also acknowledge that the ethnic group options presented are not completely about self-identity.
In the census development programmes, emphasis is placed in cognitive research and question testing on identifying categories 'that reflect people's own preferred ethnic descriptions of themselves'. This recognises the subjectively-developed nature of the concept. Community involvement has become increasingly important in this process through 'special populations' advisory groups.
The term 'race', by comparison, has been conspicuously avoided (even in the 'mixed race' options now termed 'mixed/multiple ethnic groups'), except in the 1991 Great Britain Census question instruction that referred to 'ethnic or racial group'.
Different terminology is used in the USA. The census question and tick options continue to be termed 'race'. However, the US Census Bureau indicates that the racial categories used on the census form 'generally reflect a social definition of race recognized in this country and are not an attempt to define race biologically, anthropologically or genetically'. The Census Bureau also indicates that they recognise the racial categories 'include racial and national origin or socio-cultural groups'.
As in the UK, the US agency emphasises that the categories should be familiar and acceptable to the people described 'if the principle of self-identification is to be honoured'. Again, the census data are extensively used in federal programmes and by official bodies for equal employment opportunities monitoring and to assess racial disparities in health and environmental risks.
It is against this background of a continuing increase in the use of the term 'ethnicity' rather than 'race' in the US, UK, and elsewhere that the aforementioned views have been gaining ground. To reiterate, these are: evidence of a switch to ethnicity in new compound terms like 'race/ethnicity'; the consensus among social scientists, many in the biomedical sciences, and census agencies that 'race' is a social construct; and the view that 'race' should only be used in the contexts of racism and history of race science.
It is, therefore, pertinent to ask whether the reclaiming of the term 'race' to describe clusters representing important genetic differences between human populations is judicious. Clearly, given the salient view that 'race' is a social construct, the introduction of the term 'race' in this biological context could potentially lead to confusion and ambiguity in the scientific literature.
Moreover, the concept of 'race' has a burdensome history. In the past, it was used as an indicator of biological difference and associated with eugenic practices in Nazi Germany and racist apartheid policies in South Africa. Although the current context is entirely different, it is questionable whether the term 'race' can be incorporated into scientific work without reopening its fraught past.
The debate, however, is two-sided. An increasing number of geneticists in highly-cited papers are using 'race' to describe geographically structured genetic clusters. In some cases, these are reported to correspond with major self-identified ethnic/racial groups. There have also been calls from within sociology for social scientists to re-evaluate their position that ethnicity/race is a social construct that has no biological basis.
But the question, surely, is: why would you want to requisition 'race'? As has been noted, it has an unfortunate past, there's a settled understanding of the concept as a social construct, and its usage is currently declining vis-à-vis ethnicity. Antenatal screening programmes for genetic conditions including Down's syndrome, for example, have successfully introduced the concept of 'family origins' going back at least two generations. This eschews ethnic group and race. Francis Collins, a leader of the Human Genome Project, has suggested that geographical or continental ancestry may be a useful designation for these genetic clusters.
There is an urgent need for fresh thinking on the full range of race/ethnicity/ancestry concepts to address the potential disadvantages of using race/ethnicity as a biological construct, such as essentialising of groups, biological reductionism, and the prospect of racialised risk categories. Essentialising is where the characteristics of people in a category are seen as fixed and enduring over time. Biological reductionism is where 'natural' differences are ascribed to people from different cultural backgrounds. An example of racialised risk categories is the mapping of race groups to particular diseases or therapies.