Ed Humpherson to Paul Lowe (NRS): Assessment of 2021 Censuses in the UK

Dear Paul

Assessment of 2021 Censuses in the UK– National Records of Scotland (NRS) Response to Preliminary Findings

I am writing to you to share our feedback on the progress NRS has made to meet the standard of the Code of Practice for Statistics in its developments for Scotland’s Census. There has of course been a significant announcement by the Scottish Ministers to move Scotland’s Census to March 2022. This important decision has been informed by option analysis undertaken by NRS and we recognise the work undertaken to support this decision-making.

NRS will be going through a period of re-adjusting its plans given this decision and we recognise, through for example the news release and frequently asked questions on your website, that you have started to share information with users. As this re-planning progresses, we welcome your proposals to provide further updates and engagement opportunities to keep Census data users in Scotland fully informed.

Throughout this decision-making process NRS has worked collaboratively with the Office for National Statistics and Northern Ireland Statistics and Research Agency. As you continue to work closely to understand more about any impacts given the change of Census date in Scotland, Census offices should also ensure that UK Census data users have information on how their data needs will be affected.
The work of NRS to deliver a successful Census in 2022 will continue to be considered by the Office for Statistics Regulation (OSR) as part of the ongoing assessment process. The designation decision for Scotland will only be considered when Scotland’s Census is completed and it is our aim that the UK Statistics Authority will decide whether to confirm the National Statistics designation, based on OSR’s advice, prior to publication of Census outputs in March 2023.

OSR’s advice on the National Statistics designation will be based on the trustworthiness, quality and value secured by Census offices. That advice will take account of the actions taken to address findings or requirements from OSR as part of our assessment process. Our assessment work to date remains relevant and will support you in delivering Census outputs that meet user needs. As such, we have reviewed the actions that your team has taken to address the findings from our Assessment Report 345: 2021 Censuses in the UK – Preliminary Findings. These findings aim to build on the existing work being undertaken by Census offices, providing further direction and focus on pre-existing plans.

While many of the Phase 1 findings can only be fully actioned or addressed over longer timescales, it is my view that NRS has responded positively to the assessment findings so far. I am also pleased to see the ongoing work of NRS to secure a successful Census in Scotland and to deliver high quality data and statistics. I recognise the challenges and changes in ways of working that have had to be managed during the COVID-19 pandemic and the resilience of NRS staff during this period.

Since we published our assessment report in October 2019, my team has engaged with various members of your staff working on the Census to understand the improvements you have made since then. In evaluating how NRS has responded to our findings we have taken account of published materials from NRS including the evidence report Response to Actionable Findings from Phase 1 of the National Statistics Accreditation, alongside information gathered through meetings and workshops held with your teams. I consider that some actions taken by NRS are of note.

  • NRS has added to the material available to users through its website and improvements have been made to make the user journey more straightforward. NRS also uses several tools to highlight when new content is added to its website. In its development plans for its outputs website, NRS has incorporated strong elements of user testing and has actively sought involvement from Census data users in its development process.
  • NRS has published methods papers which have been discussed by its External Methodology Review Panel. This more-technical information provides detailed methods information for expert users, alongside stakeholder events which were run for the non-expert users and stakeholders.
  • NRS also published its Statistical Quality Assurance Strategy in December 2019 which was a detailed and well-written document providing assurance of the processes and strategies in place to assess and measure levels of quality. This document should give Census data users with greater confidence in NRS’s commitment to quality and quality management.

We consider that further action is still needed in some areas identified for improvement in our assessment report – this is partly due to changing circumstances such as the impacts on programme delivery given the pandemic. For example, NRS provided the assessment team with greater assurance on its approaches to the use of administrative data. However, little information is currently available in the public domain. It is important that NRS provide users and stakeholders with information on the data sources used in the Census and on its judgement on data quality and appropriateness for use. NRS should consider what provisional information could be made public and build on this as its research and understanding continues to develop.

We have included more detail about our judgement in an annex to this letter. I, or my team, would be happy to talk you or your colleagues through any aspects of this letter or Code compliance more generally.

I am copying this letter to Sir Ian Diamond, National Statistician; Professor Roger Halliday, Chief Statistician and Data Officer, Scottish Government; Peter Whitehouse, Director of Statistical Services, NRS.

Yours sincerely

Ed Humpherson
Director General for Regulation

Temporary exemption from Code of Practice for National Records of Scotland

Dear Pete

Following discussions with your vital events team, I am happy to confirm the exemption from the Code of Practice for Statistics’ standard publication time of 9.30am to permit a later release time of noon each Wednesday for NRS’s weekly deaths statistics. Given the ongoing Covid-19 pandemic I welcome your endeavours to publish information on deaths weekly rather than to your usual monthly timescale.

I also welcome your decision to try to publish statistics in as timely a way as possible, supported by clear narrative and explanation of the data to bring clarity to public debate. This results in a tight turnaround of the death registrations. The exemption from 9.30am publication, and publication at noon, allows more time for appropriate quality assurance of the data and preparation of the statistics each week.

Yours sincerely

Ed Humpherson

Director General for Regulation

 

Related links:

Covid-19: Changes to Statistics

Compliance check of Healthy Life Expectancy Statistics (Scotland)

Dear Amy

HEALTHY LIFE EXPECTANCY IN SCOTTISH AREAS

We recently carried out a short review of compliance with the Code of Practice for Statistics of the Healthy Life Expectancy (HLE) in Scottish Areas statistics, prompted by NRS taking over responsibility for producing these from Information Services Division. Our review also covered the changes in methods used to estimate HLE introduced by the Office for National Statistics (ONS) in December 2018 that NRS has also adopted. Thank you for notifying us of these changes.

We considered the impact on users of the decision to move production of Scottish HLE statistics to NRS and the processes that ONS undertook to develop the new estimation method, including ONS’s consultation with users. Moving the Scottish HLE statistics to NRS adds value for users by bringing all life expectancy statistics in Scotland together in one place. Your decision to adopt ONS’s new estimation method means that the HLE estimates for Scotland produced by ONS and NRS are now consistent and will enable cross-national comparisons of HLE between Scotland and the rest of the UK. Local area estimates of HLE in Scotland will also now be available on a more regular basis. These developments will be welcomed by users. The methodological changes made by ONS have clearly improved the value of these statistics for users by removing some of the volatility that affected some local area estimates. As we indicated to your team in February, I am pleased to confirm that these should continue to be designated as National Statistics and welcome the statement you have included in your bulletin describing the recent developments to these statistics. Today I have also written to Iain Bell, Deputy National Statistician at ONS, outlining the rationale for our judgement on UK HLE statistics in more detail.

We have noted in our letter to ONS that the new estimation method relies on Census data which might not be available post-2021, depending on decisions taken about questionnaire content. We will be encouraging ONS, NRS and the Northern Ireland Statistics and Research Agency to work together to ensure that harmonised measures of health and disability continue to be available after the next Census.

Yours sincerely

Mark Pont

Assessment Programme Lead

 

Related links: 

Mark Pont to Iain Bell (March 2019)