Dear Pete 

Scottish Population Census

The Census in Scotland continues to be a major undertaking by National Records of Scotland (NRS). We welcome the ongoing engagement between your team and ours as we carry out our assessment against the Code of Practice for Statistics. I am writing now, ahead of our formal assessment report which we hope to publish in the autumn, to outline some key expectations of NRS that will form a crucial part of our assessment of compliance with the Code of Practice.

The disruption caused by the COVID-19 pandemic and the change in both timing and mode of data collection to digital first mean that the context of this census is noticeably different from previous ones. There has been much coverage recently of the level of response and concerns raised about the possible impact of this on the quality of population estimates from the census. We share your concerns that the focus on the census response rate as the sole indicator of the quality of estimates could become a distraction. In particular, we consider that questions about what response rate is sufficient do not adequately capture that population outputs do not rely on census data alone, but a combination of census and other data.

Having said that, we welcome the steps taken by NRS to achieve the best response rate possible and to maximise the usefulness of the census coverage survey. The additional steps that NRS is taking, such as investigating what additional non-census data can be used to further improve the quality of census estimates and establishing an International Steering Group to provide additional expert input, all point towards an endeavour that continues the aim of producing the best census outputs possible.

In the lead up to census day, we saw NRS be open and transparent about its planned approach to quality assurance and use of administrative data through publishing materials on its website. We also note your commitment to provide detailed supporting materials alongside census outputs to help users make best use of census estimates.

However, given the range of challenge that has been evident over recent months, we consider that it would be in NRS’s interests to be more transparent now about the steps that it is taking to generate good quality census estimates. We consider that being transparent about the various current activities, plans, processes etc would assure users of NRS’s trustworthiness and reassure users that they can confidently expect high quality estimates from the 2022 census. Various practices within the Code of Practice1 reinforce this need for openness even ahead of the release of the statistics. Demonstrating this trustworthiness is especially vital given the importance of census estimates.

It will therefore be important for NRS to update the existing materials soon to reflect how you are adapting your processes, and the possible impact on the expected quality of census estimates. We are mindful of the need to be proportionate in doing this, especially when the focus of your team is on processing census data. At the very least we expect a clear description of how the various steps you are taking help to assure the quality of census estimates fit together, and to provide updates to users about the level of quality that they may expect from census outputs.

Thank you again to you and your team for engaging with us through this process.

Yours sincerely

Mark Pont
Assessment Programme Lead