• 10:30 – Introduction
• 10:35 – Action and sub-group update
• 10:45 – Evaluation criteria for consultation on Ethnicity harmonised standard response options 
• 11:15 – Census Innovation Workshop 
• 15:00 – Any other business / Close

Panel members

• Sir Bernard Silverman (Chair) 
  
• Ana Basiri 
  
• Carl Emmerson 
  
• Natalie Shlomo 
  
• Nik Lomax

Office for National Statistics

• Owen Abbott (Presenter)
• Charlie Hayes (Presenter)
• Joy Preece (Presenter)
• Dan Hutchinson (Presenter)
• Sarah Henry
• Mary Gregory
• Tom Tarling (Secretariat)
• Emily Winstone (Secretariat) 
• Susan Williams (Secretariat) 

1. Introduction

1. Sir Bernard, the chair, welcomed the panel and attendees. Oliver Duke-Williams had given apologies. 

2. Actions and sub-groups update

Actions:

1. Action 114: ONS to produce and provide to the panel an explainer on papers covering the topic of Multiple System Estimation
   • This will be provided by correspondence to the panel, when available. ONS has previously discussed with the chair, and there was a brief discussion at this meeting ahead of the explainer being provided. 
2. Action 116: Secretariat to provide chair, and sub-group panel liaisons, with draft of the annual report 
   • This meeting would be the final one scheduled in the reporting period, the draft will be provided following the meeting. 
3. Action 117: ONS to hold a meeting of the Methodological Assurance Review Panel (MARP) discussing Census Innovation topics
   • This meeting fulfilled this action, and can be closed. 
4. Action 118: Secretariat to supply details on the National Statistician’s Inclusive Data Advisory Committee positions when advert is published and open 
   • This action was completed in July, and can be closed. Vacancy: National Statistician’s Inclusive Data Advisory Committee members – UK Statistics Authority
5. Action 119: ONS to structure and organise running of subgroups, integrating with the main panel to allow for review and sign-off
   • This action is on-going, and was discussed at the meeting. The panel agreed the subgroups should be convened by ONS, and chaired by an external panel member. Clarifications on this would be detailed in the Terms of Reference. ONS agreed to work with the chair and bring these drafts to the panel for review at a subsequent meeting. 
   • It was clarified the “Migration” sub-group is focussing on “International Migration”. Other subgroups include “Labour Market Statistics” and “Dynamic Population Model”.

3. Evaluation criteria for consultation on Ethnicity harmonised standard response options 

1. ONS presented the paper and addressed questions from the panel.
2. The panel discussed that there were multiple policy reasons for enumeration of small ethnicity groups. It suggested focussing on the ‘impact’ of measuring a group, rather than specific advantages and disadvantages.
3. ONS stated these criteria would be published and consulted on further, including continuing their engagement with other government departments to understand their needs, and needs of policy makers. 
4. The panel commented the strength of evidence should be weighted more heavily, as opposed to quantity of evidence, when scoring. ONS agreed to consider and revise wording of the proposed criteria. 
5. It was also commented the criteria wording assessing whether write-ins could be used as a proxy for the proposed ethnic group, ‘could’ was too theoretical. The panel suggested use of ‘effective proxy’, giving focus to practical reality. Framing of the question of ethnicity was also discussed, with ONS noting it was considering the ethnicity question and how that works alongside the cultural identity questions (National Identity and Religion). 
6. The panel cautioned about over-reliance on consulting community leaders, noting that in some circumstances this might yield a minority view of a group. It also agreed definitional issues were complex, with there being inclusive and exclusive approaches to grey areas. Therefore, terminology use was important. 
7. In the context of a census every ten years, the panel commented that interpretation of ethnicity may change more rapidly than it is measured, creating challenges for comparability and coherence. ONS agreed, clarifying it was seeking to keep comparability rather than using a zero base approach. 
8. The panel recommended enabling comparability should be an evaluation criterion in itself, and ONS could also consider ‘data quality’ as its own separate evaluation criteria. 
9. Following the ONS presentation, the panel discussed a broader approach to measuring ethnicity, highlighting circumstances where an individual may wish to identify with two ethnicities. ONS said it would consider this approach. 
10. ONS discussed the proposed assurance for the final scores. This could include Academics, policy colleagues and the Government Statistical Service (GSS) Heads of Profession. ONS commented they would also seek assurance from the National Statistician’s Inclusive Data Advisory Committee.  The panel recommended non-statistical experts be included, with representation from other social sciences beyond statistics.
11. The panel thanked ONS for the presentation.

4. Census Innovation Workshop 

1. ONS initially presented potential areas for innovation identified internally and via other workshops. From there, the session consisted of discussions and idea sharing. A topic summary of the workshop items includes: 
   • Geolocation technology, allowing targeted reminders via SMS and modern communication technology 
   • Use of Agentic AI to provide support, guidance, and potential as a route for census completion in future. It was noted technology in this area was rapidly evolving, and by 2031 the landscape could be very different 
   • Trade-off printing paper census on demand, versus pre-printing 
   • Varying requirements for enumerating communal establishments 
   • Potential for a ‘continuous census’ using online options, and ability for respondents to update answers 
   • Potential for such an online system to pre-fill some fields, reducing response burden 
   • Activities of other National Statistics Institutes (NSIs) was discussed, though it was noted some make use of national ID systems and registers that couldn’t be replicated in the UK currently 
   • Benefits of development of AI in-house and across government departments, versus benefits of building on developments from private sector investment 
   • Usage of administrative data in Census 2021, and potential to expand this further 
   • Risks of having modules and area-specific questions was discussed. ONS noted it had ruled out this option previously, the panel agreed there were significant risks to comparability and data quality, as well as technical costs and difficulties, plus requirements for methodology to consider such comparability issues. It was agreed having different output quality by area was undesirable 
   • Links between ONS and Ordnance Survey with the address base was discussed, with the panel questioning if there were further opportunities for collaboration 
   • The cost of a census versus mandatory surveys was discussed. The panel noted that large-scale surveys and census could be cheaper than smaller scale surveys that required full response to be representative 
   • Income was discussed, with it agreed gathering such information by survey tended to irritate respondents, and administrative data could be used to construct household income. ONS commented it was keen to add income to both 2031 and 2021 Census data 
   • Response impact of ‘prefer not to say’ options on questions 
   • Value of linkage from census to administrative data, including the use of geospatial data alongside the address base 
   • Public perception of GPS geolocation, and enumeration of non-address locations. Alternative technologies to GPS were discussed 
2. The panel also discussed the following thoughts: 
   • Contacting people by telephone will become increasingly difficult, both because landlines are becoming scarce, and because smartphone users often do not pickup calls. 
   • Growing numbers of people who don’t read or receive post by 2031 
   • Potential for online advertising to drive action, increasing responses 
   • Changing concept of a ‘household’ from the earlier census years to now 
   • Crowdsourcing potential to drive harder to measure groups, and groups with lower response rates 
   • Enumeration of homeless respondents 
   • User needs for new and expanded topics, such as mental health, working patterns, travel to work, carer status and disability 
   • Options to regard both religion and ethnicity as possibly intersecting cultural identities, and hence to consider merging these concepts in some way. 
   • Future topics of significance for future government and policy making was highly linked to invite development of Labour Market data, net-zero statistics such as transport mode usage and working patterns, digital exclusion, and education topics 
3. ONS thanked the panel for the discussion. It noted that these would be fed into the Census taskforce for their consideration and evaluation, alongside the other potential innovations identified. Papers on aspects of the Census 2031 design would be coming to the panel in future, as they did for Census 2021. 

5. Any other business / Close

1. ONS invited the panel to communicate any further ideas following the meeting via correspondence. 
2. The next meeting date would be polled, potentially looking at October 2025. 

Actions:

No new actions were raised in this meeting.