Office_for_Statistics_Regulation_Annual_Report_202223

Office for Statistics Regulation Annual Report 2022/23

Published:
13 July 2023
Last updated:
13 July 2023

Our performance

How statistics are produced

Delivery of the key commitments set out in our 2022/23 Business Plan.

How statistics are produced: We uphold trustworthiness, quality and value of statistics and data used as evidence

2022/23 Highlights

Outputs – Year ending 31 March 2023

Assessment reports:

9 reports published

Assessments closed:

10 closed (an assessment is closed once we have reached a final decision on designation; the publication of the assessment report is an interim stage).

Designations as National Statistics:

12 designations, of which: 3 confirmations of existing statistics and 9 new.

De-designations:

National Statistics status was temporarily suspended for 6 sets of statistics at the request of producers.

Compliance checks and rapid reviews:

38 closed

Strategic reviews including systemic reviews: 

8 reviews published

Outcomes

Focus on regulatory projects which are best placed to improve understanding and support society’s information needs.

Fully delivered:

This year we’ve demonstrated an increased maturity in adapting in-year the program that we specified at the start of the year to be demonstrably more focused on the really key issues. For example:

  • on the Scottish Census we adapted our assessment plans during summer 2022 to provide a stronger and more-timely intervention in order to facilitate better progress from NRS.
  • in response to concerns about assurances on economic statistics following the UK’s departure from the EU, we developed and are rolling out a new form of assessment focusing on quality.
  • we responded rapidly and decisively to concerns raised about ONS’ excess deaths statistics through a short review.

Other key topics covered have included: fuel poverty, covid infection survey, modern slavery, LFS transformation, regional GDP.

Highlight potential for, and positive examples of, data sharing and linkage in regulatory work.

Fully delivered:

Data sharing and linkage has become more embedded as a standard consideration in each of our regulatory projects and we have both championed and challenged producers. Key examples include: considering BEIS’ alternate ways to use wide range of data in fuel poverty statistics; engaging with ONS on new data sources for inflation measures; promoting benefits of merging data in production of Key Stage 4 statistics, and DfE’s onward sharing of data for analysis; assessment of Scottish Census drawing out aspects of using alternative administrative sources; recognition of effective data sharing as part of compliance check of Welsh school workforce statistics.

Use regulatory work to affect wider improvements, drawing insights from regulatory outputs to drive improvement across the statistical system and beyond.

Fully delivered:

Our work on Intelligent Transparency, Uncertainty, the role of Evaluation in Policy, and State of the Statistics System (SoSS) report have resonated and informed improvements by producers. We need to do more to capitalise on SoSS in 2023/24 and the project team is already well established to deliver this, informed by our Insight Committee (advisory group of producers). OSR’s international reach has also grown this year with requests to share our team’s expertise, for example, around the Code and Analytical Leadership.

How statistics are used

How statistics are used: We protect the role of statistics in public debate

2022/23 Highlights

Outputs – Year ending 31 March 2023

Cases opened:

372 – compares with 241 the previous year (included in this number are 187 duplicate cases opened with us and managed together from June-September 2022, concerning a twitter campaign on a BEIS consultation about weights and measures).

Time to close a case (days – median):

27 – compares with 11 the previous year (the increase in the median time to close a case from 11 days in 2021/22 to 27 days in 2022/23 largely reflects the more varied case topic areas that we dealt with. With COVID-19 we could often reuse or update responses to previous cases, with the more varied case load we have to build knowledge afresh. The high number of duplicate cases for the BEIS consultation also impacted the median time to close a case).

Time to close a case (days – mean):

26 – compares with 20 the previous year.

Cases still open at year end:

10 (equal to the previous year).

Outcomes

Improve our understanding of how public consume information.

Partially delivered:

Through the year, we have gleaned some direct insight into how the public find and consume information from the survey of Public Confidence in Official Statistics. We published our first specific reports on the topic of Statistical Literacy and made recommendations on how to improve communication. We have also steadily built our knowledge of work going on outside of OSR, which focusses on public understanding and trust in statistics, such as by ESCoE and the Winton Centre. We are scoping research work on how the public use statistics for 2023/24.

Intervene where there is greatest potential for misunderstanding and misuse.

Fully delivered:

We have shown that we are more than willing to step in publicly when we have concerns. We have also written and published statements on a number of high-profile topics such as the ONS excess death reporting, the differences between an asylum seeker and migrant, and financial sanctions after the Russian invasion on Ukraine.

Make statements about data sharing and its contribution to public debate.

Partially delivered:

We are preparing a report into data sharing and linkage across the public sector for publication in early Summer 2023.

Consolidate our position to support any public use of data and analysis by UK governments. Focus on system-wide change.

Fully delivered:

We have had great engagement from the analytical community and other stakeholders around our work on uncertainty and intelligent transparency. We have hosted several successful analytical seminars, bringing together analysts from different areas to discuss topics such as RAP, data linkage, uncertainty, and intelligent transparency. These sessions were well attended with over 150 attendees at each session and 450 attendees at our RAP session.

How statistics are valued

How statistics are valued: We develop a better understanding of the public good of statistics

2022/23 highlights

Outputs

Our National Statistics designation refresh project has developed and tested ways to reframe the designation to enable clearer messaging and the greater accountability in the production of official statistics – the recommendations have been submitted to the Authority Board in May 2023. We are piloting material to support producers to effectively communicate quality and to develop a ‘Code culture’, and we are refreshing our assessment process.

Outcomes

Understand what “public good” looks like.

Fully delivered:

We have taken big steps in developing our understanding of public good of statistics. We are now finalising a first-draft definition of the public good provided by statistics.

Develop OSR work to identify and evaluate our impact.

Partially delivered:

We are at the stage where we have gathered a lot of evidence of impact but more robust recording systems will improve confidence that we are capturing and capitalising effectively the knowledge we have. We have made gains this year in respect of the OSR culture around identification and evaluation of our outcomes and impact. We are in the final build phase of our new workflow system which will pose questions from 2023/24 at scoping and initiation of how outcomes and impacts will be measured and evaluated, with reporting at closure. We made good headway in developing an organisational Theory of Change in the first part of the year but that work has developed more slowly, reflecting limited resources. Our new Insight and Evaluation Manager is now in place and structuring our evidence base for evaluation.

Understand value of data linkage and sharing and share this within and beyond OSR.

Partially delivered:

As part of the data sharing and linkage review, we have engaged with a wide number of stakeholders. We will be following the publication of our report in Summer 2023 to engage and share what we have learned with our regulators and the statistical system.

Improve understanding of our impact and influence.

Partially delivered:

We have used social media metrics (Twitter and LinkedIn) to identify the audience that our work reaches. We have also seen our public good work referenced in the Scottish Government Health and Social Care Data Strategy, and voluntary adoption recommended for charities in the recent Pro Bono Economics report on civil society.

Encourage voluntary adoption of the Code of Practice for Statistics.

Fully delivered:

Awareness and support for voluntary adoption seems to be steadily gathering momentum. We have launched the 2023 TQV award and we continue to support analysts in adopting TQV, giving advice to new organisations considering adoption. We most recently added Fable Data – a private company – to the VA list.

Use our understanding of public good to influence beyond official statistics.

Fully delivered:

We have been accepted to two conferences (PERITIA conference on policy, expertise and trust as well as International Statistical Institute world conference), allowing us to widely spread our message about the public good. We also organised a government analysts event on the public good with speakers including the National Data Guardian and Data Science Campus.

Back to top
Download PDF version (1.12 MB)