4. Findings and recommendations

This section of the report presents our findings and recommendations. These are structured into four areas:

  • Changing landscape of data sources
  • Understanding user needs and comparability
  • Strengthening strategy and governance
  • Overcoming finance and resource barriers

Changing landscape of data sources

The data underpinning UK statistics are changing; we increasingly see challenges to traditional survey data collection in the context of reducing response rates, and more use of administrative and other big datasets. Typically, these administrative datasets occur as a by-product of the provision of government services, which may be specific to a UK country or local area, for example the collection of council tax, rather than being directly collected for statistical purposes.

For example, the Office for National Statistics (ONS), through its Future of Population and Migration Statistics programme, is developing admin-based population estimates as an alternative to its current approach of using mid-year estimates, which uses a rolled-forward stock from the decennial census of England and Wales. These administrative-data-based estimates use sources such as health, education, housing, and income, tax and benefits data provided by other government departments and public agencies in England and Wales.

Traditionally, surveys have been the primary mode for achieving input harmonisation, as they consistently pose the same questions on a national scale, even when they are developed and administered in different parts of the UK. An example is the collection of data on wellbeing as part of the Labour Force Survey and the Annual Population Survey. Harmonised concepts and questions can also be used in administrative sources. However, these sources tend not to be in the control of statisticians, and therefore it is harder to make changes for statistical purposes, resulting in less standardisation. Consequently, comparability between statistics on common topics but based on different administrative sources may be reduced. For example, the statutory homelessness statistics produced by the individual UK nations are not comparable, as they are based on administrative data recorded in line with devolved homelessness policies. However, the GSS coherence team has successfully established that statistics on the number of homeless households in temporary accommodation in each nation can be meaningfully compared.

Increasing levels of devolution and policy variation across different parts of the UK mean that administrative datasets are often collected at country, regional and local levels with varying degrees of harmonisation in concepts. This can particularly impact the comparability of data in nationally devolved topic areas, such as education and health, but goes far beyond these areas.

The increasing reliance on local administrative data resulting from the devolution of powers to local government, such as the Mayoral Combined Authorities in England, along with the new role of the Mayoral Data Council, which seeks to integrate Mayoral Strategic Authorities with the UK Government digital and data function, provides further challenges to the dominance of national survey sources and for maintaining meaningful comparability at multiple geographic levels within nations.

We also heard about the need for statisticians to engage more heavily with managers of operational data systems in the future, as statisticians try to shape the systems on which they will depend for future data feeds in line with common data standards that will support comparable outputs.

Managing the impact of statistical improvements on comparability

There are a lot of parties with stakes in the debate about the adequacy of UK comparable data (ONS; the Ministry of Housing, Communities and Local Government/Cabinet Office; Scottish Government; Welsh Government; Northern Ireland Assembly/Northern Ireland Statistics and Research Agency; and local data leaders). These actors have different governance and legal gateways for data sharing, technological capability, and people and financial resources available to support comparable UK statistics and data.

For example, it can be challenging for statistics producers operating in different parts of the UK to have similar access to administrative or linked datasets due to the different legal and data sharing frameworks and practices in place. The Digital Economy Act provides ONS with a legal gateway to receive data from all public authorities and Crown bodies and new powers to mandate data from some UK businesses. However, these powers are not shared with other statistical producers, and only in limited circumstances can data held by ONS be shared with the devolved governments, and solely for statistical purposes under strict conditions. NISRA told us that there is currently no legal framework that would allow it to access linked data from Northern Ireland government departments. This is an additional challenge for producing comparable UK-wide statistics and limits the improvements in enhanced understanding and decision-making that data sharing offers.

As the statistical system makes significant changes to priority areas of statistics, these imbalances have the potential to exacerbate comparability issues, because some parts of the system may develop more quickly than others. To mitigate these risks, it is important that the system focuses on the principles set out in the Concordat on Statistics. The Concordat sets out the agreed framework for co-operation between the UK Government (including the UK Statistics Authority and Office for National Statistics) and devolved administrations, in relation to the production of statistics, specifically, to:

“work together to deliver an agreed suite of coherent, reliable, consistent and timely statistics about and across the UK” and to “collaborate to ensure the coherence of statistical infrastructure across UK Government and the devolved administrations to better enable the production of cross-UK data and analysis, where appropriate”.

Maintaining a level playing field for statistics production capability will therefore be an important enabler of comparability.

Planning for the various 2031 censuses in England and Wales, Scotland and Northern Ireland, which are likely to draw on a mix of survey and administrative data sources, provides a significant opportunity for the UK statistical system to work collaboratively and strategically. Ahead of 2031, the main UK producers should seek to identify any required legislative changes to support the balanced development of data sharing, access and capability across the individual UK nations.

Recommendation 1: In the context of supporting comparable data capability across the UK, the Government Statistical Service, under the direction of the Inter Administration Committee, should conduct a review of legal frameworks relating to data sharing and access across the UK nations. This review should focus on considering the imbalances between the data sharing and access opportunities and, if necessary, identify legislative changes that can level the playing field in data sharing, access and capability across the UK nations.

Back to top

Understanding user needs and comparability

Identifying types of need

Comparing performance and outcomes across the four nations is often cited as a key driver for having comparable UK data. However, achieving this goal is becoming increasingly challenging due to differences in data collection and reporting practice.

Political and policy needs often drive the demand for comparable UK-wide data, while simultaneously making data less comparable due to the different policy contexts across the UK. In areas like health and education, there is an increasing interest in being able to assess the effectiveness of different policy regimes across the UK and the outcomes for citizens. A comparable baseline of information is essential for informed decision-making and policy development and understanding ‘what works’ in terms of delivering positive policy and societal outcomes for UK citizens.

The UK Government’s response to the Lievesley review noted that “It is fundamental that citizens in all parts of the UK should be able to compare their data with other parts of the country. Comparable data are also crucial to enable all administrations to identify, design, and deliver interventions which benefit people and communities across the United Kingdom.” We have heard similar views through our engagement as part of this review.

Balancing the individual needs of UK countries and local areas with the need for UK-wide comparability is a complex challenge. Where data are collected at a sub-UK level, budgetary and resource constraints mean local needs, and the needs of the funder of the statistics, are often prioritised. This can conflict with a UK-level view of the public good, making it difficult to address these issues effectively. Often administrative data from the implementation and delivery of a policy can determine the scope of the available information. These data may lack characteristics that people might be interested in for research or comparability purposes.

The UK statistical system must be empowered to provide clear direction and to overcome these challenges. Currently, there are insufficient incentives for individual parts of the system to come together to prioritise and resource the production of comparable UK-wide statistics on priority measures where there is an identified and agreed need to do so.

This is not just about producing comparable statistics across the four nations. It can also provide valuable lessons to support comparable local data and ensure coherence in the context of local data and English devolution.

The Office for National Statistics (ONS) told us that there is a lot of interest in having comparable subnational statistics across the UK nations. This is something that users request regularly via the ONS Local service, wanting to be able to compare subnational areas in one nation with similar areas across the UK. ONS told us that its ONS Local and Subnational teams are leading a substantial programme of work in this area, which includes building analytical and data capability at local levels. This work presents a real opportunity for ONS to support the comparability of local-level data, particularly in the context of English devolution and the new Mayoral Data Council, with statistics produced both regionally and nationally. However, enhanced governance and political buy-in at different levels may be needed to realise these benefits.

In addition to needs for comparison at national, regional and local levels, we have also identified needs for global comparability, including by international organisations, for example the Organisation for Economic Co-operation and Development and the United Nations, who collect and curate data on global Sustainable Development Goals.

Making international comparisons between the UK and the rest of the world can be complex, in part because of the challenges of bringing together data from the four UK nations to get a consistent estimate. We have heard of instances where data for England are used to represent the whole of the UK, and the potential for this to lead to biased estimates in cases where there is divergence across the four nations. For example, for UK reporting against the UN Sustainable Development Goals, English statistics are commonly used as a proxy for all UK nations in UK submissions due to extensive data gaps. This theme was also highlighted in the Lievesley review.

Collating and understanding needs

A clearer understanding of users’ needs for UK-wide comparable data across various topics is needed. For statistics to serve the public good, producers need to understand user and potential user needs for information. This need is best understood through wide user engagement, both within and beyond government.

Responsibility for developing the evidence base of need for UK-wide comparable statistics is shared among various stakeholders, including ONS, UK Government departments and the devolved governments. While producers’ engagement with users within their own nations may be strong, it’s inevitably focused on needs in the relevant part of the UK, with less focus on comparability between geographies. There is a need for better mechanisms for cross-UK engagement to identify and address UK-wide comparability issues. We heard about recent, positive engagement led by the GSS UK-wide and coherence teams to understand needs more broadly across UK government departments.

However, we consider that the delivery of comparable UK-wide statistics that meet the need for change requires a more joined-up approach across the GSS. This approach should consider demand for UK-wide statistics beyond the different governments of the UK and to consider a ‘UK public good’ lens – taking this holistic view based on broad user engagement would better serve the needs of users.

While the GSS previously conducted qualitative research to understand UK-wide data needs beyond government, its scope was limited. Additionally, there are currently insufficient established processes and engagement approaches to understand and bring together a broad, joined-up view of the needs of users for UK-wide comparable data on different topics, and to fully understand the value that users place on this.

The UK Statistics Assembly in January 2025 provided clear evidence of a good step in this direction, and it demonstrated that there is strong desire from a broad range of users to engage with this topic. There are also examples of good practice; these are typically where the need for UK-wide comparability is considered from the outset, as in, for example, the new data collection for the Homes for Ukraine settlement scheme. In this case, comparability across the UK was planned from the beginning, with all countries pausing at the start of data collection to collaborate on developing comparable data.

The absence of a joined-up picture of users’ needs inhibits the effective prioritisation and development of comparable UK-wide statistics and data for the public good. This conclusion is in line with the first overall priority that the UK Statistics Assembly identified for the statistical system to address:

“As a priority the GSS should re-invigorate sustained and effective user engagement, in which official statistics producers take a lead in understanding the needs for statistics and curating relevant sources, to help answer the questions that the public, businesses, local government, the media and academics, as well as policymakers have about the economic, social and environmental situation. This would also help understand and increase the value of statistics.”

The UKSA and the GSS are currently developing plans to address findings from the UK Statistics Assembly but are doing so in the context of constrained financial resources.

Assessing the comparability landscape

The UK Statistics Assembly highlighted the opportunity to consider if it is possible to unpick existing measures from the different governments of the UK to produce something ‘comparable enough’ to meet users’ needs. It also highlighted the increasing importance of local data in the devolution context.

A key aspect to meeting user needs is understanding the current comparability landscape. A better understanding of this landscape, informed by a more comprehensive understanding of users’ needs for UK-wide comparable statistics, would help statistics producers understand how best to address the gaps. Publishing details of the landscape would also provide valuable information to users to help them effectively use the available data and understand the cases where data might be ‘good enough’, versus the cases where comparisons should not be made.

In 2013 the GSS published guidance on comparing official statistics across the UK. The guidance was aimed at producers of official statistics to help them assess the comparability of their statistics with other parts of the UK, thereby enabling users to better understand what was comparable. The guidance included a comparability scale for the four nations comprising ten detailed categories broken down across three main classifications: fully, partially and not comparable. The guidance was archived in 2019 and is no longer in active use.

Drawing on more recent work by The Fraser of Allander Institute (as part of the research programme of the Economic Statistics Centre of Excellence) and internal guidance issued by the thenDepartment for Levelling Up Housing and Communities in 2023, OSR proposes the use and further development of a new comparability framework tool (Annex A). This tool would support the UK statistical system and users to better understand current statistics and data comparability issues, and to prioritise statistical measures and topics where comparability should be improved.

Recommendation 2: The Government Statistical Service, under the direction of the Inter Administration Committee, should:

  • conduct cross-UK user engagement to better understand specific needs and potential uses for comparable UK-wide statistics and data for the public good. This should build on known UK and devolved government needs to identify statistics on priority topics that should be comparable at national, regional and local levels to provide a more informed strategic direction to UK-wide data work
  • assess and score the comparability of the identified priority statistics at national, regional and local levels against OSR’s proposed comparability framework (Annex A) and publish these findings
  • publish a strategy to produce and maintain comparable UK-wide statistics and data on selected priority measures at national, regional and local levels. This should be based on a cross-UK commitment where the National Statistician and Chief Statisticians all agree that meaningfully comparable statistics (as a minimum) should be provided
  • monitor progress in the development and maintenance of comparable UK-wide statistics for agreed priority measures on an ongoing basis, refreshing understanding of users’ needs for comparable UK-wide statistics at least every three years, potentially as part of the triennial UK Statistics Assembly process

Enacting this proposal can build upon the planned work of the UK-wide coherence team, which has recently undertaken extensive cross-government engagement to inform its priorities for 2025/26. We suggest that this engagement could provide a useful starting point for establishing ‘known UK and devolved government needs’, as discussed in Recommendation 2, ahead of establishing broader public needs for comparable UK-wide statistics and data.

OSR also plans to trial this framework in some of its upcoming reviews and will consider the case to embed it in our assessment practice and make it a standing requirement of all producers to steer users on the extent of comparability of their statistics using this standard framework.

As we develop the final version of our revised Code of Practice for Statistics during summer 2025, we will also consider the responses we received to our Code of Practice consultation to determine if the third edition of the Code should have a greater emphasis on the comparability of statistics. We will also provide guidance on the Code’s use, outlining an expectation that producers report on the extent of comparability of their statistics with other statistics on the same topic.

Back to top

Strengthening strategy and governance

Concordat on Statistics and Government Statistical Service progress

The governance of the UK statistical system for producing comparable UK data involves multiple parties, including the UK Statistics Authority (UKSA), the Office for National Statistics (ONS) and the devolved governments of Wales, Scotland and Northern Ireland. It is supported by the Concordat on Statistics.

The Concordat is the key framework for cooperation between the UK Government, including the UKSA and ONS, and the devolved administrations of Scotland, Wales and Northern Ireland. The Concordat is overseen by the Inter Administration Committee (IAC), which is chaired by the National Statistician. It aims to ensure the production of coherent and comparable statistics across the UK, while recognising the unique policy contexts within each administration.

Key aspects of the Concordat state that while recognising that the policy context within administrations might not always be identical and that official statistics should reflect local as well as UK user needs, producers of official statistics across the administrations should look to develop:

  • new or existing statistics in such a way that, while meeting the needs of primary users, aids coherence and comparability in the UK as well as internationally
  • coherent infrastructure, including interoperable platforms, definitions and classifications for the linking and analysis of data across the UK Government and the devolved administrations to better enable cross-UK data and analysis

The Concordat also states that “where new statistics are needed, under new and evolving devolution settlements, the statistical producers will consider, as far as appropriate, whether they should produce equivalent statistics where feasible, to a comparable level of quality, subject to there being a demonstrated user need and the necessary resources being available.”

The current Concordat is primarily focused on collaboration and engagement and does not commit the main UK statistics producers to providing meaningfully comparable UK-wide measures on any specific topic. In our view, more is needed to provide strategic-level prioritisation and to overcome barriers to delivery. We have heard from some stakeholders that a stronger governance approach through a revised Concordat could be helpful in ensuring the effective comparability of data.

The Lievesley review also recommended that “the UKSA should build on existing work and lead discussions between the four nations and strengthen the Concordat on Statistics to encourage more UK wide data by creating common standards and improving harmonisation where appropriate and mutually agreed. HM Treasury should ensure that funding is available to support the harmonisation of key data.”

There are ongoing efforts within the UK statistical system to better understand how comparable UK statistics are, which are being led by the GSS UK-wide and coherence teams coordinated by ONS. In September 2024, the GSS published its 2024/25 coherence work plan. This sets out achievements and planned work across a range of topics: adult social care, crime and justice, environment, climate and nature, equalities, health, housing planning homeless and rough sleeping, income and earnings, population and migration, trade and investment, and UK-wide comparable data.

Overall, this approach appears to be working well, with significant programmes of work delivered. For example, on the coherence of health statistics across the UK, the team has published, over the past year, coherence articles exploring the comparability of NHS experience and satisfaction data across the UK, NHS planned care waiting times statistics across the UK and cancer waiting time statistics across the UK. It has also updated the GSS Health and Care Statistics Interactive tool to provide UK coverage of information. The tool signposts to published official health and care statistics across the UK in one location, which was previously available for England only.

The team advised us that for their 2025/26 update due to be published in September 2025, they plan to take a more integrated approach by combining previously separate coherence and UK-wide workstreams. To inform this, they sought government user input and developed a priority matrix for assessing coherence risks and which coherence and UK-wide topics should be prioritised.

The work of the GSS coherence team is effective. But much of the team’s planned work is carried out on a project-by-project basis, with delivery based within each project’s governance structure. This work is agreed across UK-wide topics using a ‘hub and spoke model’ approach, with ONS staff coordinating work with relevant topic-based teams across government. With some notable exceptions, the GSS’s efforts are primarily focused on understanding and explaining the differences that exist between the related statistics produced across the UK.

However, the GSS coherence work will not on its own be sufficient to achieve meaningful UK-wide comparability across a range of priority measures. A strengthened, system-wide strategic and enhanced governance approach will also be required to overcome both the longstanding barriers and new challenges to producing comparable UK-wide statistics and data.

Strengthening strategy, governance and collaboration

In the previous section, we identified the need for the GSS to carry out a comprehensive review to understand users’ needs and potential uses for meaningfully comparable UK-wide data on priority topics. Further work will then be needed to agree a new UK-wide strategy for meeting identified needs on priority measures that serve wider public information needs.

Decisions on the priority measures that should be comparable should reflect a wide range of user interests. But these decisions should not be made by politicians. Instead, they should be made impartially by the National Statistician and the Chief Statisticians of the devolved nations through the IAC. The chosen measures should also be aligned with established international or harmonised UK-wide data standards wherever possible.

As the IAC takes this work forwards, it should develop a longer-term work programme. We heard that the current short-term focus tends to result in less-strategic decisions, whereas a longer-term perspective could help address this issue.

There is also a need for local data leaders and users to input into these processes; government statisticians should act as advocates for comparability and harmonised standards with local data leaders. It is currently unclear whether existing governance structures are able to facilitate this approach, or whether new governance might be needed. The Concordat on Statistics should also be strengthened to include a fuller commitment to, and accountability for, the production of a baseline of agreed comparable UK-measures on priority topics, at national and also regional and local levels.

The Fraser of Allander Institute (as part of the research programme of the Economic Statistics Centre of Excellence) has recently launched work exploring improvements to cross-organisational collaboration in UK regional data. We hope that this work will uncover insights that will lead to even more effective collaborative work across the system, supporting evidence-based decision-making and moreefficient resource distribution. 

Recommendation 3: The Government Statistical Service (GSS), under the direction of the Inter Administration Committee, should conduct a review of the strategic effectiveness of the current governance arrangements in place for ensuring the production of comparable UK-wide statistics at national, regional and local levels. Acting on the findings of this review, the National Statistician and UKSA Board should strengthen these arrangements to ensure that they are sufficiently strategic and adequately resourced and that they represent the interests of all four UK nations and local data leaders. This should be done with a view to achieving meaningful comparability across a range of agreed, UK-wide measures on priority topics, as part of a strengthened Concordat.

Back to top

Overcoming finance and resource barriers

There are inequities in the resources, capability and infrastructure required across the UK system to safeguard existing comparable statistics and achieve UK-wide comparable statistics and data in priority areas. These differences can lead to gaps in the availability of measures across the UK.

The analytical and data resources and capability available to the devolved governments are much smaller than those available to the UK Government Whitehall departments and ONS and consequently might not benefit from some economies of scale. These differences in resourcing impact the UK’s ability to achieve increased data comparability. The UK Government spending review process is a key mechanism for allocating resources. Often, resourcing for comparable data is centred in ONS, but any increases in funding for ONS do not generate a Barnett funding consequential increase for Wales and Scotland.

The Lievesley review recommended that HM Treasury ensure funding is available to support the harmonisation of key data. The UK Statistics Assembly also identified that to be effective, coherence and harmonisation work requires dedicated funding. A more effective and equitable distribution of resources is therefore needed to effectively support the production of comparable UK-wide statistics and data on key measures.

However, given that each nation prioritises the data requirements of its users, collecting new comparable measures represents an additional cost, as effectively this may mean that more than one measure is required on the same topic. The Authority response to the Lievesley review noted this as a significant barrier to harmonised data.

The UK statistical system should seek cross-UK political commitments to adequately finance meaningfully comparable statistics on agreed priority measures at national, regional and local levels. The 2031 Census recommendations offer an imminent opportunity to initiate such conversations within the framework of a long-term strategic plan that utilises survey, administrative and third-party sources.

Recommendation 4: The National and Chief Statisticians and the UKSA Board, supported by the governments across the four nations, should seek a cross-UK political commitment to adequately finance statistical developments for producing meaningfully comparable statistics at national, regional and local levels. They should establish this agreement through intergovernmental relations and a strengthened Concordat. The 2031 Census recommendations offer an imminent opportunity to initiate such conversations within the framework of a long-term strategic plan that utilises survey, administrative and third-party sources. In addition, they should explore alternative financing models for achieving comparable UK data that ensure a long-term and equitable distribution of resources at regional and subnational levels along with more efficient sharing of human and non-tangible resources, such as the sharing of data. This work should also consider building local level analytical capability, taking into account the context of English devolution, the Mayoral Data Council and the services provided by ONS Local.

Back to top
Download PDF version (336.66 KB)