Annex C: Review of press releases

Key information for interpreting these findings

This is the first exercise of this kind that OSR has conducted. The review is intended to provide an illustrative overview of the extent to which press releases cite sources for statistics – a key element of the second principle of intelligent transparency. The findings presented here are explorative and do not constitute a formal OSR regulatory judgement. It is important to note that there is a level of subjectivity present in this exercise, and it should not be considered an exact science. It is also important to note that if a citation and/or link to the source was not identified, this does not necessarily mean that the underlying analysis is not published.

Method

We identified all press releases published by departments in scope of this review between 14 and 20 July 2025 (see Annex A). The press releases were taken from the following sources:

• UK government departments: Press releases were sourced through UK. At the time of the exercise, it was not possible to filter specifically by press releases, and so we determined those in scope through manual checking for those marked as ‘Press Release’ just above the article title. Outputs with other labels, such as ‘news’ or ‘correspondence’, were not included in the sample.
• Welsh Government: Press releases were identified through the Announcements page with a filter applied for ‘press releases’.
• Scottish Government: Press releases were identified as those listed as ‘new’ on the News page.
• Northern Ireland: Press releases were taken from the dedicated Press Release page.

Definitions

As this is intended to be an explorative exercise, we have taken a broad definition of what constitutes a reference to ‘statistics’ as well as what was considered to be a source. There is a level of subjectivity in this exercise, and we acknowledge that some departments have identified a different number of statistics than we did in our review. This does not change our overall findings at the aggregate level, which was the intended purpose of this work. The next section sets out how we identified statistics to include in our review.

References to data and statistics

We judged something to be a reference to statistics if the statement either directly quoted a figure or if it was a statement that was based on an underlying calculation or piece of analysis. Statements were not considered in scope if they referenced funding announcements or numbers that were not based on underlying analysis.

The number of references to statistics and data identified within each press release is based on the number of unique claims rather than the total number of references within a press release. By this, we mean that if a specific claim is used multiple times in a press release, it will only be counted once in this analysis.

Some examples of statements that we considered to be a reference to data or statistics:

• “More than 50 million meals have been served since the introduction of Universal Primary Free School Meals.”
• “Research published alongside the Green Paper today also highlights the important role the Post Office still plays in the daily lives of people and businesses, adding social value of around £5.2 billion per year to households.”
• “With two-thirds of adults in England currently overweight or living with obesity”

Some examples of statements that we did not consider to be a reference to data or statistics:

• “More than 170 hospices across England will receive a share of the funding”
• “Today’s action will lower the Crude Oil Price Cap from $60 barrel to $47.60 directly hitting Russia’s oil revenues”
• “Chancellor launches new £500m Fund to break down barriers to opportunity”

Source

We judged the press release to be providing a source if the origin of the underlying analysis was mentioned. We also accepted references to the organisation that had produced the analysis where it indicated an underlying report. We recognise that this should not be considered the gold standard in providing a source.

Examples of sources that had been provided:

• “This year’s report showed that the majority of airports assessed were performing either in the ‘good’ or ‘very good’ category.”
• “The 153rd Open will be the largest ever sporting event held in Northern Ireland and is expected to generate more than £213million in total economic benefit, according to an independent forecast by the Sport Industry Research Centre (SIRC) at Sheffield Hallam University.”
• “Last August, Scotland was the first nation in the UK to introduce another new RSV vaccine, Abrysvo, for pregnant women and older adults – more than 70% of eligible older adults took up the offer, leading to a Public Health Scotland study, published in the Lancet, showing a 62% reduction in RSV related hospitalisations among this group.”

Links

We considered that a link was provided if there was a hyperlink to the statistical or analytical output that provided the figures in question.

Findings

We identified 80 press releases across 17 different departments. Of these, 47 press releases were identified as containing references to statistics, with 209 unique references. Of these 209 references, 36 (17%) cited the source and 18 (9%) provided a direct link to the source.

To support transparency, the full analysis of the 80 press releases identified in our review can be accessed by downloading the excel sheet listed below. This should not be used to form the basis for further analysis. Where press releases were produced by multiple departments, we have just listed the lead department. Departments may appear more than once if they published more than one press release during the relevant period.

Download the press release analysis