Sikh ethnic tick box in the 2021 Census and a question about research and methodology

The views and opinions expressed in this article are those of the author and do not necessarily reflect the official policy or position of the University of Birmingham

“It is essential that policy makers, such as the ONS, make decisions based on a robust evidence base.”  


A recent article in The Times, ‘Sikhs may get ethnicity status in census’,  which seemed  to suggest that the inclusion of a separate Sikh ethnic tick box in the 2021 census is a real possibility, has generated a lot of discussion within the Sikh community. Many Sikhs have cried foul about the potential bias and flaws in the Office for National Statistics’ (ONS) consultation process regarding the issue. 

To give some context, lobbying for a separate ethnic status within the UK census has been going on since the early 2000’s by certain Sikh groups. The argument to include an amendment to the 2011 Census was that the existing questions on religion and ethnicity undermined the Sikh community and the services they were entitled to. The lobbying ultimately failed even after the Sikh Federation (UK) threatened ONS with court action in 2010, in an attempt to force the issue. The ONS is now facing calls from the Sikh Federation (UK) and the All Party Parliamentary Group (APPG) for British Sikhs,  for the same amendment to the 2021 census.  The Chair of the APPG, Preet Kaur Gill MP, has already stated that Sikh Gurdwaras have contacted the APPG to discuss the establishment of a legal fund to challenge any negative decision by the ONS. 

Advocating for the change, there are a number of reasons often cited by Sikh Federation (UK) and others for the need for a Sikh ethnic tick box. These include under-estimation of Sikh numbers, inadequate allocation of resources to Sikhs based on current census statistics and the 1983 Mandla vs Dowell-Lee ruling which recognised Sikhs as an ethnic group. There is also a claim that many Sikhs simply do not want to be labelled as ‘Indian’.  

In the 2011 Census only 1.6% of respondents had recorded themselves as ethnically Sikh with a religious affiliation other than Sikh. This compares to a 2017 survey commissioned by the ONS in Hounslow and Wolverhampton (areas with high numbers of Sikhs) to test differences in responses to Sikh religion and Sikh ethnic tick boxes.  Three alternative versions of the ethnic group question of an online survey were tested. These were census in style and identical, except for the ethnic group question. All respondents to the 2017 Ethnic Group Question Test who had selected Sikh ethnic group also selected Sikh as their religious affiliation. In the words of the ONS, ‘There is no indication from the findings [2017 survey] that the religious affiliation and ethnic group questions are capturing different Sikh populations. All respondents who stated they were ethnically Sikh also stated their religious affiliation was Sikh. This is in line with findings from the 2011 Census data’. These results suggest that there is minimal, if any, underestimation of Sikh numbers in the UK using existing Census data.

In terms of inadequate resource allocation to Sikhs, to date no data has been presented in the public domain to substantiate this claim with reference to healthcare and educational funding.  Data on religion is routinely collected by healthcare providers and educational establishments and this may be a better source of information than census data when making resource allocation decisions to particular groups. 

The definition of Sikhs as an ethnic group in UK law stems from the 1983 House of Lords ruling in the Mandla v Dowell-Lee case which used ethnicity to protect Sikhs under the 1976 Race Relations Act.  As Lord Fraser stated in 1983 "The Sikhs are a "racial group" defined by reference to ethnic origins for the purpose of the 1976 Act". However, this is no longer relevant today because the 1976 Race Relations Act has been replaced by the Equality Act 2010

To the claim that Sikhs simply do not want to be labelled as ‘Indian’, this is difficult to prove without further data. A simple check of the number of respondents in the 2011 census who ticked both the Sikh and ‘Indian’ tick boxes would allow one to gauge where the majority view lies.  

Finally, community representation should happen through a process of consultation and feedback. Representatives are elected by the community and work across the community to represent their full range of views and experiences.  It is not clear if any of the groups purporting to represent the Sikh community on the matter of the ethnic tick box fulfil this definition of a ‘representative’.   It is also not clear what process these groups have undertaken to reach their position on the Census, or what evidence base informs it. This leaves a concern regarding how widely and robustly the ONS has consulted the Sikh community.  

In evaluating the consultation process, and the evidence presented by the Sikh APPG and the ONS, there seems to be little justification to amend the 2021 census to include a Sikh ethnic tick box and the data from the 2011 Census supports this.  It is essential that policy makers, such as the ONS, make decisions based on a robust evidence base, and further work is needed to establish, what, if any, benefits could be seen by this change.