Summary of the meeting with the Sikh community: 23 October 2017

On 23 October 2017, the Office for National Statistics (ONS) hosted the “2021 Census and Statistics About the Sikh Community” event. This was an open meeting for organisations and individual experts who have an interest in statistics about the Sikh community. The aim of the meeting was for us to share progress in developing the questions for the 2021 Census and to discuss how the need for statistics about the Sikh community can be met.

This is a note of that meeting and also covers feedback provided at, or after, the meeting by those who attended. This note is structured around the themes raised in the meeting and from the feedback provided afterwards from those who attended. Slides presented at the meeting have been published.

Background

Since 2015, we have engaged with the Sikh community to discuss requests for the addition of a Sikh tick box within the 2021 Census ethnic group question. We have met members of the All-Party Parliamentary Group (APPG) for British Sikhs, the Sikh Federation (UK), Sikh Network, Sikh Council and Lord Singh of Wimbledon. The 2021 topic consultation and a follow-up survey on ethnicity received representation from Sikh community groups for the addition of a Sikh tick box within the ethnic group question.

We hosted the “2021 Census and Statistics About the Sikh Community” event to provide an update on the research since the topic consultation and listen to views of the Sikh community. An email invitation to the meeting was sent to individuals and organisations who had previously expressed an interest in data about Sikhs through our consultation process. It was also sent to Census Advisory Group members, government departments and was posted on our website and StatsUserNet.

There were 47 attendees, in addition to the ONS staff at the meeting. The majority of attendees at the event declared affiliation or membership to the following organisations: City Sikhs, National Sikh Youth Federation, Sikh Council UK, The Sikh Federation, The Sikh Network, Sikh Women's Alliance, The Sikh Archive, Sikh Missionary Society UK, Sikhs for Labour, Statistics User Forum, Gurseva and The Khalsa Jatha British Isles Est. 1908. Individuals also attended.

Other meetings with the Sikh community include a meeting in the West Midlands on 1 December and in Hounslow on 16 December.

Our presentation

We began the meeting by sharing an overview of the Census Transformation Programme. It showed the journey from the topic consultation to where we are now in the process towards finalising recommendations on the topics and questions for the 2021 Census. This included details of the evaluation criteria and timelines for finalising questions.

We presented slides covering the engagement and assurance of the approach taken to develop the ethnic group question. This included the evidence gathered, question testing undertaken, and opportunities and constraints associated with asking the question online. We also highlighted that consistent and comparable questions will need to also form part of the paper questionnaire.

In the slides we presented we said:

• the meeting was convened to listen to the views of the community

• decisions will be made based on the evidence gathered so far, the views of the community and other considerations such as space and impact on harmonisation and legal advice

• no decision has been made on whether to make changes to the ethnic group or religion tick boxes

The meeting then turned to the views of the participants. We set out that logically there were the following options:

1. Include a Sikh tick box in the ethnicity question and the religion question

2. Do not include a Sikh tick box in the ethnicity question but include a tick box in the religion question (the 2011 scenario)

3. Include a Sikh tick box in the ethnicity question but not in the religion question

These options were used to stimulate debate during the meeting. At the meeting there was strong support for option 1 over the other options suggested. In a vote taken at the end of the meeting only two people voted against option 1.

The following views were expressed by attendees of the meeting either during the discussion, in feedback forms or through email feedback after the meeting.

Definition of ethnic group

There were differing views over what the ethnic group question was trying to capture. Comments received highlight these issues:

“…the previous census seems to conflate/confuse race, ethnicity, and nationality and perhaps that needs to be reorganised.”

“The ethnicity question itself requires better definition and categorisation. It is something that has developed over the years but requires a major rethink in order to clarify its purpose and a stricter definition to help make it easier to understand if a category can be added or not.”

In the meeting, we explained that the census first asked a question in the UK on ethnicity in 1991 and the question has evolved over time. The question is a mixture of geographic and colour terms. It is designed to try and best capture ethnicity and is a question which allows individuals to identify as they wish. We are trying to devise a question that can best reflect society in England and Wales.

Within the meeting there were suggestions that the whole question needed a fundamental rethink. In response, we said that overhauling the ethnic group question would be a large task, which would not be possible now within the timelines of the 2021 Census. We are working to balance what will work well now as well as enabling users such as local authorities to be able to make comparisons with previous censuses. The Race Disparity Audit highlighted a real need for better information on ethnicity. We are working with the Race Disparity Unit to improve equality monitoring.

Reasons to add a Sikh ethnic group tick box

The majority of the attendees at the meeting expressed the view that including a Sikh tick box in the ethnic group question, in addition to the Sikh religion tick box, was their preferred option. The following opinions were expressed, through discussion and feedback, as reasons to include a Sikh ethnic group tick box:

• the case of Mandla versus Dowell-Lee ruling (1983) suggests that Sikh ethnicity has been recognised as an ethnic group in law

• a Sikh tick box would allow Sikhs to self identify their ethnic group with a tick box (instead of through a write in box as was the case with previous ethnicity questions)

• the tick box is required for ethnic monitoring of Sikhs (because government departments and other data collection exercises mostly use the census ethnicity question for monitoring and allocate resources)

• ethnic Sikhs are currently missing from official data

• the religion question is voluntary, and therefore the size of the Sikh population may be understated as they choose not to answer the question

• the opinion that there is a consensus within the Sikh community to add an ethnic group tick box

• it is the preference of many Sikhs not to be labelled as “Indian”

• the proportion identifying their ethnic group as Sikh doubled in testing when a Sikh ethnic group tick box was added

• it would be discriminatory not to add a Sikh tick box

We provided the following feedback to some of these points.

The religion question was added in 2001. It provides additional information on identity alongside the ethnic group question.

We confirmed that the House of Lords insisted that the religion question be voluntary when introduced in the 2001 Census.

The percentage of religiously affiliated Sikhs who identified with a Sikh ethnic group was 24% when a Sikh tick box was available in testing. This was double the 12% who wrote in Sikh when no tick box was available. A report of findings is available in 2017 ethnic group question test – Sikh ethnic group and religious affiliation findings. This also details the findings that from the test there was no evidence 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.

Concerns were raised about the sample size of the ethnic group 2017 question test. We confirmed that the sample, of 40,000 households, was a random sample across the areas selected. These areas were chosen due to a relatively high Sikh population.

Testing was conducted to get a good, broad, reasonably representative view across a number of topics. We explained that the test findings are a reasonable indication of evidence to feed into our evaluation. While we would have liked more responses it is likely that this would not have changed the findings. The results are in line with findings from the 2011 Census data.

We noted that we would be happy to review any evidence of any undercount of Sikhs backed up by sources of such evidence.

Better data about the Sikh community

There was a consensus in the meeting that there was a need for better data on British Sikhs.

Many respondents noted that it was important to get information on Sikhs using the most appropriate method to do so. This included discussion around the importance of Sikh data beyond the census, including what data could be achieved before the census data would be available.

Suggestions were provided for question design and operational issues including on the position of a Sikh tick box, should one be included, and the importance of materials being provided in Punjabi (translation material was provided in Punjabi for the 2011 Census).

Reasons not to add a Sikh ethnic group tick box

The following opinions were expressed, through discussion and feedback, as reasons not to include a Sikh ethnic group tick box:

• larger ethnic groups than Sikh (Polish was given as an example) are not represented as tick boxes within the ethnic group question

• potential confusion to respondents of having both a Sikh ethnic group and religion question

• Sikh is not an ethnic group; the Mandla case was used to protect Sikhs under the Race Relations Act at that time

• being Sikh is a choice, not something you belong to at birth (there are white and African Sikhs)

• if monitoring is done on ethnic quota alone (monitoring those who identify as ethnically Sikh and may not wear visible signs of the Sikh religion), it would not reveal discrimination against the visible signs of the Sikh religion

• the majority of people writing “other ethnicity” as “Sikh” are doing this for political reasons

• there is no evidence that Sikhs are not being allocated funds that they are entitled to

Format of the meeting

Some commented on the meeting we hosted to gather feedback from users of Sikh data.

The following opinions on the format of the meetings were shared:

• the meeting was causing disharmony and exploiting divisions within the community

• some believed the timing of the meeting was too late in the decision-making process while others questioned whether the meeting was needed at all (as the community had already taken part in earlier consultation processes)

• the meeting objectives and how the feedback of the meeting would be used to influence a final decision was unclear

• the meeting didn’t address the areas it was set up to address and lacked a rational data-focused discussion

Those requesting a tick box often expressed the opinion that the views of the community were clearly in favour of this addition. Others felt that dominant voices left no room for debate and expressed concerns about the representation at the meeting. Some individuals provided feedback that views voiced at the meeting were not representative of the full Sikh community.

Comments on the approach and processes so far

Criticisms were raised regarding the processes we had employed including:

• lack of transparency of the evaluation

• ambiguity of the timeline for finalising recommendations about the questions to be included in the 2021 Census

• that harmonisation guidelines need updating to consider the need for Sikh data

We provided information on the processes we employed within the meeting. The evaluation criteria can be found in the slides (and were presented by us at the event). The timeline for decisions can be found on slide 10.

We accept that the harmonised guidelines could be improved and committed to review them.

Next steps

No decision has been taken on whether the census will include an ethnic group Sikh tick box and we will continue to work with all parties on the best way forward.

We will use published evaluation criteria together with legal advice to make an assessment for any additional tick boxes to include on the 2021 Census. The National Statistician will not be making his recommendations until 2018. In doing so he will take into account the range of research available and input from stakeholders.