Manchester expert assists major report into how to estimate UK's population

The Office for Statistics Regulation (OSR) has published a new report about the new methods that the Office for National Statistics (ONS) are developing to produce population estimates for England and Wales, with the help of an expert from The University of Manchester.

The Dynamic Population Model is a new statistical model that aims to produce more-timely and accurate estimates of the population by age, sex and local authority. The DPM is part of ONS’s wider ambition to transform population and migration statistics and make more use of administrative data. 

The report, aided by the expertise and guidance of Professor Arkadiusz Wiśniowski, an expert on Social Statistics and Demography, recognises the ambition of ONS in developing the DPM – an innovative method for estimating the population using administrative data.  OSR has identified eleven requirements for ONS to improve the trustworthiness, quality and value of the ABPE: 

• Developing and publishing criteria to support the decision about when the ABPEs will replace the MYEs, in consultation with key stakeholders such as the Welsh Government; 
• Strengthening the governance structure and decision-making responsibilities for the production and development of the ABPE, and managing any risks associated with funding, capability and prioritisation; 
• Publishing a plan for addressing the recommendations and suggestions in Professor Wiśniowski’s independent review of the data and methods used in the DPM.  
• Understanding and quantifying the quality and uncertainty of the data inputs to the DPM, especially the Statistical Population Dataset (SPD) and the migration data, and implementing regular assumption checking and validation; 
• Publishing more information on the quality and methodology of the ABPEs, including the strengths and limitations of the data sources, the coverage of special population groups, and the impact of data and methods changes over time; 
• Improving its quality assurance process; 
• Implementing and publishing a revisions policy for the ABPEs, based on user feedback and analysis of the scale and frequency of revisions; 
• Creating and implementing technical and expert user groups to provide independent and external challenge and assurance on the methods used in the DPM; 
• Developing and implementing a user engagement strategy specific to the ABPEs, to understand and meet user needs, communicate plans and updates, and manage user expectations; 
• Being open to scrutiny from key stakeholders and users and responding appropriately to their feedback and queries; and 
• Improving and tailoring the communication of the methods and quality of the ABPEs, considering the differing technical expertise of population statistics users. 

These recommendations build on the positive progress ONS has made in publishing a suite of information related to these statistics today. Taking the actions outlined in the recommendations will ensure that users have more confidence in the new method, and therefore the ABPEs themselves. 

“We welcome the work of the Office for Statistics Regulation which, along with input from our users, helps inform our development of these important statistics,” said Mary Gregory, Interim Director of Population Statistics for ONS. “We welcome the work of the Office for Statistics Regulation which, along with input from our users, helps inform our development of these important statistics. Today, we’ve provided an update on our progress towards admin-based population estimates (ABPEs). We’ve sharing these new data to help users understand the new approach, share their feedback with us, and take time to consider what it means for them before we move to the ABPEs as our official estimates of the population.

“Although these ABPEs are at a research and development stage, our intention is for them to become our official measure of the population in 2025, dependent on meeting the acceptance criteria we will publish later this year. We will take into account the feedback we receive following engagement, and will only transition to the new approach once we are confident they are of the high standards that our users need. We will be working closely with the Office for Statistics Regulation as we seek accreditation of the ABPEs, and our long-term international migration estimates.”

Professor Arkadiusz Wiśniowski said: “The OSR’s report evaluates the progress the Office for National Statistics made with developing new population estimates that are based on administrative data sources. Population estimates are crucial for making decisions about our lives, such as funding of the A&Es, number of GPs per areas, new schools or infrastructures. They also underlie most of the economic, health and other indicators, including those used to measure progress towards Sustainable Development Goals. The new admin-based population estimates are meant to replace the current estimates that suffer from various issues, as well as potentially replace future censuses.

“My role was to assess a sophisticated statistical model (Dynamic Population Model) that is developed by the ONS and how it is used to produce those population estimates. I was happy to be involved because, firstly, the development of such models is my main area of research and, secondly, I believe the ONS’ work is ground-breaking and highly innovative globally. I think it is essential that all assumptions made in the DPM are appropriately described and tested, which will ensure that the population estimates are of highest possible quality. This, in turn, will ensure user trust and confidence in population statistics.”