Panel Members

• Sir Bernard Silverman (Chair)
• Dr Oliver Duke-Williams
• Dr Nik Lomax
• Prof David Martin
• Prof Natalie Shlomo

Office for National Statistics

• Julie Stanborough (ONS Lead)
• Owen Abbott (ONS Deputy Lead & Presenter)
• Cal Ghee (ONS Attendee)
• Jon Wroth-Smith (ONS Attendee)
• Helena Rosiecka (Presenter)
• Ed Dunn (Presenter)
• Leone Wardman (Presenter)
• Steve Smallwood (Presenter)
• Gareth Powell (Secretariat)
• Rob Turner (Secretariat)

1. A73 –Owen Abbott – ONS to work through the implications of altering the field operation timings
2. A74 – Helena Rosiecka – ONS to produce paper on options for the Sex Question
3. A75 – Viktor Racinskij -ONS to bring paper on Adjusting for dependence bias in the coverage estimation of the 2021 Census of England & Wales

1. Chair’s Introduction

The chair welcomed the panel, ONS attendees and Marie McGhee, an observer from OSR who attended the first four items to observe MARP processes.

2 – Review of Actions

• The panel agreed to close action 18, due to the ongoing transparency provided by publishing on UKSA.
• The panel agreed to close actions 24 and 56, following presentation at MARP11 of EAP145 – Performance of 2021 coverage adjustment method with lower response and contingencies.
• The panel agreed to close action A62 following the scheduling of a rolling Covid-19 presentation.
• The panel agreed to close action A64 following update on response targets from authors of EAP137 – Household response profiles for the 2021 Census
• The panel agreed to close action A66 after ONS conveyed that the suggestion had been integrated into the ongoing administrative data work and would be followed up as part of that work
• The panel agreed to close action 69 following presentation at MARP11
• The panel agreed to close action 61a following work by secretariat
• The panel agreed to close action 61b as work underway and paper to be presented at future MARP
• The panel agreed to close action 61c following feedback from ONS on outcome of simulations

3 – COVID-19 Update

ONS updated the panel on current COVID-19 workstreams. Panel suggested looking into the effect on modelling and estimation of extending field operations in some areas.

Action 73 – ONS to work through the implications of altering the field operation timings

4 – EAP147 – Overcrowding statistics derived from VOA number of (bed)rooms

ONS presented on the strategy and analysis undertaken for assuring overcrowding statistics derived from Valuation Office Agency (VOA) number of rooms variable, replacing the number of rooms question on Census 2021. Also discussed was the similar strategy and analysis undertaken for these statistics when derived from VOA number of bedrooms.

ONS presented analysis used to compare the number of bedrooms in a property, as measured by VOA and the 2011 Census. Following panel feedback, ONS provided a greater level of detail on the mean number of rooms across each property type according to each data source. The panel noted that using a mean average to measure rates of overcrowding might cancel out the households at the high and low extremes, in particular for owner-occupied properties.

ONS discussed regional differences in discrepancies between VOA and Census 2011 number of bedrooms, highlighting in particular for owner-occupied homes in London and part of the North-West of England. The panel suggested a likely explanation might be due to differences in how local authorities submitted updated information to VOA. The point was additionally raised that this proactive process of submitting VOA updates could result in fewer updates from owner-occupied properties that are less likely to have a change of residents than rented properties. Overall, the panel recommended discussions with VOA to discuss these results.

The panel enquired about floor area data present in VOA and whether this might be useful for overcrowding statistics. ONS have done some feasibility research into this finding differential treatment by VOA on floor area for different types of property.  Energy certificates were raised as an area of interest, due to consistency in how they measure floor area across property types.

The panel suggested clarifying what criteria for overcrowding were followed and highlighted the importance of being transparent about how number of rooms is calculated. They discussed the potential for providing data that would allow users to analyse overcrowding themselves.

The panel questioned the recommendation for a simple subtraction from the number of rooms.  ONS explained that they do not have data to model more precisely and hence were suggesting a simpler way of adjusting that will remain accurate in future.

5 – Methodology for decision making on the 2021 Census guidance for the sex question

The panel engaged with ONS in a detailed discussion of the Census 2021 Sex Question. No conclusions were reached, but the discussion focussed on the kinds of evidence and documents the panel would like to receive. ONS agreed to produce a detailed paper for consultation in time for the next meeting.

Action 74 – ONS to produce paper on options for the Sex Question

6 – Adjusting for the dependence bias in the coverage estimation of the 2021 Census of England & Wales

ONS presented on the work currently in progress surrounding the adjustment for dependence bias in coverage estimation. Panel expressed interest in receiving a full paper on progress as soon as possible.

Action 75 – ONS to bring paper on Adjusting for dependence bias in the coverage estimation of the 2021 Census of England & Wales

7 – Any other business

No other business was discussed