FOI reference: FOI-2026-3547
You asked
This request concerns the Office for Statistics Regulation's exercise of the monitoring commitment set out in the UK Statistics Authority's letter of 25 March 2026 to 18 leaders of major political parties in England, Scotland and Wales, in which the Interim Chair stated that the Office for Statistics Regulation would make statements as needed during the election campaigns to clarify the appropriate use of statistics.
Please provide the following information, which I understand is held by the Office for Statistics Regulation:
(1) Any internal record, log, briefing note, written assessment, or monitoring document produced by the Office for Statistics Regulation between 25 March 2026 and 7 May 2026 that records the review, monitoring, or assessment of statistical claims made publicly by any political party leader or their party during the election campaign period, including any assessment that did not result in a published letter or statement. I am not requesting any published correspondence, which is already in the public domain.
(2) The criteria, guidance, threshold, or decision-making framework, whether set out in a formal document or recorded in any internal communication, applied by the Office for Statistics Regulation in determining which statistical claims made during the campaign period warranted a published intervention and which did not.
(3) Any internal record of a decision not to intervene in response to a specific statistical claim made by any political party or leader during the campaign period, including the reasons recorded for that decision.
(4) Any internal communication discussing the application of the pre-election monitoring commitment to the statement made by Kemi Badenoch MP at Prime Minister's Questions on 29 April 2026, including any record of who identified that statement as warranting intervention and the basis on which that decision was reached.
We said
Thank you for your request, which concerns how the Office for Statistics Regulation (OSR) exercised the commitment set out in the UK Statistics Authority's letter of 25 March 2026 to 18 leaders of major political parties in England, Scotland and Wales, in which the Interim Chair stated that OSR would make statements as needed during the election campaigns to clarify the appropriate use of statistics.
You have asked for information in relation to four requests. Unfortunately, we cannot provide the internal documentation that you have asked for in relation to work across this period. As we will explain, this would exceed the cost limit for this FOI request. However, we hope the answers we have given provide a thorough insight into how OSR conducted itself and carried out its role during the pre-election period.
The Head of the OSR, Ed Humpherson, would welcome the opportunity to discuss OSR's role and how this is delivered with you in more detail. Please let us know via following email if you would like to meet: regulation@statistics.gov.uk
1) As the Authority's regulatory arm, OSR is responsible for managing its casework function, undertaking monitoring to identify issues regarding the production and or use of statistics and investigating issues raised with the Authority.
OSR undertook both proactive and reactive work in the run up to and during the election campaigns that took place across England, Scotland and Wales during March to May 2026. Our aim was to help ensure the appropriate and transparent use of statistics during the election campaigns, so that members of the public could understand and verify the statistics they heard and have confidence in them.
In advance of the elections, we published explainers about statistics on key issues that are likely to arise in the election campaigns, alongside our existing guidance on the use of statistics in a pre-election period. This was to help reduce misuse and to help others identify potential misuse more easily.
During the pre-election period, which ran from the end of March until 6 May 2026, we convened a time limited OSR 'election stand up' team to keep abreast with how statistics were being used in public debate. There were 12 people in this team, drawn from different business areas within OSR. Collectively the team reviewed published manifestos from six major parties in Scotland (Conservatives, Labour, Liberal Democrats, Greens, Reform and Scottish National Party) and six major parties in Wales (Conservatives, Labour, Liberal Democrats, Greens, Reform and Plaid Cymru). The team also performed daily, high-level reviews of major press outlets in Scotland, Wales and England, and of major social media platforms (X, LinkedIn, Bluesky and Tik Tok), and periodically reviewed fact checks published by the fact checking organisation Full Fact and by BBC Verify. Finally, team members listened to or watched major radio and TV election programmes, held on BBC Wales, BBC Scotland, Channel 4, ITV and STV. Across all this monitoring work, the role of the stand-up team was to identify and log statistical claims they thought might warrant investigation. These claims were discussed among the stand-up team in bi-weekly meetings and ones that the team collectively felt needed further investigation were passed to the relevant regulatory team within OSR to investigate, in line with our normal casework process. The election stand up team flagged statements with regulatory teams if they failed one or more of the questions set out in this published article on navigating statistical claims.
Statistical regulators throughout OSR also continued to conduct their own horizon-scanning, throughout the pre-election period, reviewing media and social media outlets for statements that drew on statistics relating to their individual topic areas.
Finally, throughout the pre-election period, OSR continued to respond to concerns raised with us by people from outside OSR, about the production and use of statistics.
The work of the election stand-up team produced a log of 85 individual claims/concerns, though not all of these were passed on to regulatory teams to investigate.
Within the period of the request, we added 41 issues to our formal casework issues log (8 in 2025/26 and 33 in 2026/27), which were investigated by our regulatory team. These 41 issues originated from a combination of: statements highlighted by the election stand up team; issues identified through normal regulatory domain horizon-scanning and; concerns raised with us by people outside of OSR.
Locating, extracting and collating the internal information (which would include emails, Teams chats, briefing documents and logs across at least 30 people) that relates to each of the 85 statements noted by the election stand up team and the 41 cases investigated, as well as wider team horizon scanning work, would take more than 24hrs and would therefore exceed the cost limit for this FOI request. As such, Section 12 of the Freedom of Information Act 2000 (FOIA) applies.
To illustrate this more clearly, we estimate that (across the 12 members of the stand-up team) it would take 10min to search for items in relation to each of the 85 statements the team logged, which is equal to approximately 14hrs. It would take a further 10min to search for documents and communications relating to the 41 cases that were investigated, totalling approximately 7hrs. Finally, each of our 24 regulators would need to search for items relating to horizon scanning they did during the pre-election period, which we estimate would take 30min per person, totalling another 12hrs. These times do not include extraction of the information found. We do, however, publish a summary of our formal casework issues log: it can be downloaded from our website casework page. The published issues log provides a date, short title and high-level outcome for each case. It includes whether the response was private or published, or (in rare cases) if there was no action.
2) OSR uses its voice to stand up for statistics and to represent the public, reporting publicly where we have concerns and highlighting good practice. It is extremely important that we work independently and impartially whenever we are investigating concerns: we are very conscious that not doing so could damage public trust in us and in statistics.
We have a clear framework to help us decide if a statement is misleading, and a well-established process to agree when we intervene and the most appropriate form of intervention. Application of this framework and process across all our casework enables us to work consistently and fairly when we investigate concerns.
To support transparency and accountability, we publish several key documents, which explain our framework and processes, and set out how we deliver our work:
Our published Interventions Policy sets out how OSR decides when it will or will not intervene in cases where there are specific concerns about statistics, especially their use in public debate.
The published FAQs on our Raise a concern webpage include further information about how we approach cases, including specifically 'How do you judge whether something is misleading or potentially misleading?' and 'How do you decide when to publish your responses/views?'
OSR can continue to intervene publicly during formal pre-election periods, under its role of promoting and safeguarding the production and publication of official statistics. However, as set out in our interventions policy, our threshold for intervening publicly is higher. We are likely to limit public interventions to occasions where a statement is materially wrong or misleading to the point it could significantly impact on the political debates that are presented to voters.
In all our investigations we focus on driving positive change, to strengthen public confidence in how statistics are produced and used: we believe this is best enabled when people can verify and understand what they hear.
3) As we have explained in section one, we are not able to share internal documentation relating to our work during the pre-election period, as it would take too long to identify and collate. Throughout this period, decisions on whether and how to intervene on the use of statistics in public debate were made in line with our published interventions policy, as set out in section 2.
4) Our earlier response to FOI-2026-3542 provides details of the investigation OSR conducted into the statement made by Kemi Badenoch MP at Prime Minister's Questions on 29 April 2026, including who identified that statement as warranting intervention and how the decision to write a public letter was reached.
Finally, we would like to reiterate that, the Head of the OSR, Ed Humpherson, would welcome the opportunity to discuss OSR's role and how this is delivered with you in more detail. Please let us know via following email if you would like to meet: regulation@statistics.gov.uk.