Advice from the Methodological Assurance Review Panel on the Office for National Statistics’ proposals for the Future of Population and Migration Statistics

Published:
17 June 2025
Last updated:
17 June 2025

Panel’s feedback prior to the consultation

Further to the feedback given on methods papers over recent years, during the drafting of the consultation document, ONS approached the panel to ensure the consultation covered all the areas the panel thought relevant.

At the 29th meeting of the MARP, on 21 April 2023, FPMS were discussed. This was ahead of the consultation going live on 29 June 2023.

This document represents the view of the panel as a body providing independent advice to the National Statistician, based on the evidence it has been presented. Note that panel members are interested in population statistics as individuals, or as part of other roles they hold, and may have provided feedback in those capacities separately.

In addition, the panel discussed the consultation document (PDF, 1334KB) at a meeting on 10 October 2023. Key topics discussed included:

Guidance on statistical methodology underpinning ONS statistical production and research

  • Through previous papers and methods brought to the MARP by the ONS, the panel has provided guidance on a range of statistical methodology, including the coverage adjustment approach for census, methods for variance estimation, fractional counting, calibration methodology, and quality assessment methods. For more information, see the minutes to individual meetings as listed in annex 1.
  • Further to proposed methods making increased use of administrative data, the panel highlighted the potential of other non-traditional data sources such as mobile phone data and other geospatial datasets that could contribute to the future of population and migration statistics.
  • The panel emphasised the need for high-quality surveys to complement administrative-based statistics, for example as coverage assessment tools to assess the quality of the administrative data or to collect information on other variables that are outside of the scope of administrative data.
  • The panel also stated the need to compare proposed approaches with a baseline of continuing with the current approach, that is a census cycle of approximately every 10 years. The dimensions of any comparison should include a cost-benefit analysis as well as aspects of quality, such as relevance, reliability and timeliness.
  • The panel recognise the challenge of survey burden being linked to decreasing survey response rates, however some surveys will still be necessary even with increased usage of administrative data.

Reliance on Administrative Data, and mitigation plans

  • The panel noted that the proposals rely on administrative datasets which are outside of ONS control, raising potential risks of continued access, or discontinuation of data sources. The panel recognised that the ONS have planned mitigations to allow the system to react to changes that may affect the ONS’s goals given external changes. The need for sustainable access to datasets, considering the requirements for such, was highlighted.
  • The panel recommended the ONS consider its ability to return to a census if required, and whether this would be listed as a feasible mitigation option. An alternative to a full census would be a partial census for example, and the panel highlighted historical usage of such an approach.
  • In addition to the examples given, the panel stated there was potential to include additional datasets in the estimation in future, potentially including private sector and alternative data.
  • The panel commented that other national statistics institutes have used hybrid census-admin data systems, and that analysis and commentary on such systems could be included where relevant.

Public trust and related issues

  • The panel commented on public perceptions on how data is used and raised potential trust issues of the ONS linking datasets without good transparency and explanation, noting that this can be challenging.
  • The panel commented that cost-benefit examination of proposed transitions to alternative data-based estimates, and the value of making variables more timely and fit for purpose versus the costs of doing so, would add to transparency and trust, and would help to inform the National Statistician’s recommendation.
  • The panel appreciated the consideration given to more distant future users, genealogists, family historians and historical demographers, and encouraged the ONS to keep these in mind.

Bernard Silverman
Chair
15 November 2023

Back to top