National Statistician’s Independent Review of the Measurement of Public Services Productivity

Published:
13 March 2025
Last updated:
14 March 2025

Report summary

In the UK, around 20% of Gross Domestic Product (GDP) is accounted for by the outputs of public services, comparable to most other western economies. The recent coronavirus (COVID-19) pandemic, and the desire to find innovative ways to improve public services without increasing spend or taxes has brought the focus back onto public services, both in terms of measurement and making improvements to delivery itself.

The coronavirus pandemic highlighted that public services can be subject to significant changes which measurement systems designed for more standard times can struggle to accommodate. In 2023 the then Chancellor of the Exchequer asked the National Statistician, Professor Sir Ian Diamond to review the measurement of public service productivity, noting the changing data and policy landscape that may arise in coming years. Automation, the use of Artificial Intelligence (AI) and other forms of innovation also offer opportunities to transform public service delivery.

The measurement of the productivity of public services is historically and internationally acknowledged as being challenging. This is because of the absence of prices to demonstrate the value to citizens of what is produced. This report builds on previous work undertaken by the Office for National Statistics (ONS) to make improvements to published statistics, building on the Atkinson Review: Final report Measurement of Government Output and Productivity for the National Accounts (2005) and the Bean Review (2016).

The following principles, which are references in recommendation 1 of this Review, are reproduced from recommendation 4.1 of the Atkinson Review (2005) and continue to apply:

  • Principle A: the measurement of government non-market output should, as far as possible, follow a procedure parallel to that adopted in the national accounts for market output.
  • Principle B: the output of the government sector should in principle be measured in a way that is adjusted for quality, taking account of the attributable incremental contribution of the service to the outcome.
  • Principle C: account should be taken of the complementarity between public and private output, allowing for the increased real value of public services in an economy with rising real value of public services in an economy with rising real GDP.
  • Principle D: formal criteria should be set in place for the extension of direct output measurement to new functions of government. Specifically, the conditions for introducing a new directly measured output indicator should be that (i) it covers adequately the full range of services for that functional area, (ii) it makes appropriate allowance for quality change, (iii) the effects of its introduction have been tested service by service, (iv) the context in which it will be published has been fully assessed, in particular the implied productivity estimate, and (v) there should be provision for regular statistical review.
  • Principle E: measures should cover the whole of the United Kingdom; where systems for public service delivery and/or data collection differ across the different countries of the United Kingdom, it is necessary to reflect this variation in the choice of indicators.
  • Principle F: the measurement of inputs should be as comprehensive as possible, and in particular should include capital services; labour inputs should be compiled using both direct and indirect methods, compared and reconciled.
  • Principle G: criteria should be established for the quality of pay and price deflators to be applied to the input spending series; they should be sufficiently disaggregated to take account of changes in the mix of inputs; and should reflect full and actual costs.
  • Principle H: independent corroborative evidence should be sought on government productivity, as part of a process of ‘triangulation’, recognising the limitations in reducing productivity to a single number.
  • Principle I: explicit reference should be made to the margins of error surrounding national accounts estimates.

This report summarises the work of the Review over the last 18 months, presenting the key challenges identified; the cross-cutting methodological improvements the Review has developed; the service specific improvements implemented; and recommendations for further work. The Review offers a substantive step-change in UK public services measurement, which over the next few years can be applied into the UK National Accounts as well as productivity measures, to support the coherence and accuracy of the UK economic measurement system.

Status of published ONS estimates pre and post Review

Tables 1 and 2 illustrate the status of the ONS published estimates, by public service, prior to the Review, and the position achieved by Spring 2025 when improvements will be incorporated into the Annual Public Service Productivity (PSP) statistical publication on an ongoing basis.

In these Tables a yes indicates robust estimates; a no indicates the use of simpler methods. The use of ‘Research’ indicates the Review was able to undertake research and develop recommendations for further work, but not implement improvements.

Table 1: Published estimates, status prior to this Review, June 2023

ServiceRobust inputsActivity based outputsQuality adjusted
Social Security Administration*NoNoNo
Taxation Administration**NoNoNo
Policing and Immigration*NoNoNo
Defence*NoNoNo
EducationYesYesYes
HealthcareYesYesYes
Public Order and SafetyYesYesYes
Environmental Services***NoNoNo
Adult Social CareYesYesYes
Children’s Social CareYesYesYes
Other Local Government Services***NoNoNo
*Input = Outputs (and hence by definition the growth of productivity is zero)
** Pre-review, Taxation was part of “Other” grouping
***Environment and Other Local Government Services are within “Other” grouping

Table 2: Published estimates, status post this Review, Spring 2025

ServiceRobust inputsActivity based outputsQuality adjustedIncorporated into PSP published estimates
Social Security Administration***YesYesYesYes
Taxation Administration***YesYesResearchYes
Policing and Immigration*YesResearchResearchNo
Defence*YesResearchNoNo
EducationYesYesYesYes
HealthcareYesYesYesYes
Public Order and SafetyYesYesYesYes
Environmental Services**ResearchResearchResearchNo
Adult Social CareYesYesYesYes
Children’s Social CareYesYesYesYes
Other Local Government Services**NoNoNoNo
*Input = Outputs (and hence by definition the growth of productivity is zero)
**Environment and Other Local Government Services remain within ”Other” grouping
***Statistics under Development (experimental statistics)

 

Further work has been identified to build upon this progress. See the List of Recommendations.

Main findings

Key challenges

  • Even before the global coronavirus pandemic, the delivery models of some public services were changing substantially (for example, the transition of the benefits system to Universal Credit) necessitating new methods to better account for these. The Review makes key proposals to address this change and make the system more robust to any future changes.
  • As well as changes to services, over time there has been, and will be, changes in data availability. This may be because data are no longer available or sufficiently robust, or new data becomes available which allows further disaggregation of measures or improved robustness in measurement.
  • Some services have multiple outputs or outcomes which makes it difficult to create robust metrics. A number of considerations need to be taken into account such as mapping inputs to each activity, establishing relative weights and attributing the output and outcomes to participating bodies.
  • The coronavirus pandemic, as a significant shock to the UK economy, fundamentally impacted models of service delivery and has necessitated intervention in some public service productivity models. In some cases the need for this will continue as the pandemic’s impact continues to affect public service delivery outcomes.
  • The coronavirus pandemic is impactful not just on service delivery, but also directly on the statistical models themselves. This is particularly the case with regards to ‘seasonal adjustment’ (the statistical methodology employed to smooth out the effect of seasonal impacts from the quarterly time series) where new structural breaks have required the ONS to refresh it’s understanding of when different activities happen and how to best reflect this. This Review is very much in step with wider domestic and international methods development across the public sector.
  • Using cost to value services, as Atkinson (2005) noted, leads to a failure to recognise the value added by public services. This is most extreme for preventative services, which are often low cost but deliver significant later savings, for example a lower-cost tobacco cessation programme which would reduce a future higher-cost expense, in this example – cancer operations. The Review has carefully considered how to address this challenge.
  • The UK is not alone in recognising the importance of this topic. In 2024 the United Nations (UN) launched a Review of the Classification of Functions of Government (COFOG). This presents the opportunity for the UK to feed the findings of this Review into international negotiations on recommendations to update the current classification system used for public service measurement to better reflect current (and expected future) patterns of government expenditure. Areas the Review has proposed which merit particular attention by the international process include gaining a better line of sight of Environmental spending, Tax Administration, and Defence.

 

Adding to the evidence base, improving methods and data

  • The Review has undertaken a line-by-line audit of existing methods and proposed multiple improvements based on new data, new methods and the experience accrued in delivering estimates since the Atkinson Review (2005).
  • The Review has resulted in implementation of improved methods and data into the accredited annual and experimental quarterly Public Service Productivity statistics published by the ONS, improving transparency for users.
  • Investment in new pilot surveys of how public sector workers spend their time, and the management practices employed in the public sector, has yielded valuable insight into the drivers of productivity in the public sector.
  • Qualitative research on the lived experience of public sector workers has provided insights into understanding the impact of administrative tasks on productivity and highlighted opportunities and barriers to innovation, including the use of automation and Artificial Intelligence (AI) to improve productivity.
  • The Review has worked with ongoing work on the measurement of Government Research and Development funding to help ensure the understanding of this key driver of productivity is well understood and evidenced.
  • The Review has worked to improve deflators used to derive public service volume estimates of inputs and outputs, with particular focus on improving the relevance and accuracy of intermediate consumption (IC) inputs deflators across all service areas (apart from Healthcare and Adult Social Care). There have also been small improvements to labour inputs deflators, and exploratory work on the capital inputs deflator for Defence.
  • The Review has focused methodological development work on addressing outstanding gaps relating to the measurement of preventative services and latent capability, to better reflect the value of such services where current cost weighting models may under-represent their importance in modern public service delivery.
  • The Review has challenged the accepted concept of ‘collective services’, which has been used to brand some areas of public services as ‘too hard to measure’. Whilst the ONS has not been able, in the time available, to fully develop methods in every area, it is increasingly clear that, given modern data availability, it should move away from dividing services into ‘individual’ and ‘collective’, and address opportunities to tailor solutions given the nature of the service.
  • It remains the case that some services deliver multiple outcomes and how to reflect these in quality adjustment, and particularly how to weight these different outcomes together remains to be addressed. Similar issues apply over time – is a GCSE point in 1998 worth the same as a GCSE point in 2023 – and the Review has developed recommendations for further work in this regard.

The Review has led to a number of countries to directly approach the ONS to discuss methods and concepts, and in some cases to launch their own parallel domestic processes.

Improving timeliness of public services productivity estimates

  • Timeliness of published National Statistics accredited estimates of annual public services productivity has been improved through development of experimental ‘nowcasting’ modelling, to address the two-year time lag in the necessary administrative data sources becoming available. These will continue to be under review given the challenges of modelling associated with the coronavirus pandemic.
  • Publication of experimental quarterly estimates of productivity by service sector has begun, with the release of estimates for the Healthcare service from 10 February 2025 (the largest service by spending) with plans for release of Quarterly estimates of Education productivity, followed by a roll-out to other services.

 

Coherence of public services productivity estimates

  • The principles of the Atkinson Review (2005) remain relevant, but this Review has developed innovative methods to implement these, building on the knowledge the ONS has gained in producing these statistics over the past twenty years.
  • Coherence between annual accredited National Statistics and quarterly experimental statistics on public services productivity can be improved, and a programme of methodological work to achieve this has been developed.
  • Coherence of estimates across the devolved governments of the UK could be improved. The Review has found numerous examples where different measures have evolved in different systems and where consistency across data sources would improve coherence in the UK statistics produced.
  • The development or proliferation of productivity measurement in local areas of the UK, and the establishment of organisations focused on increasing productivity (and efficiency) in specific public services, means that coherence between these measures and the national measures produced by the ONS becomes increasingly important.

 

Summary of sector specific findings

  • Recognising work which was recently undertaken on Children’s and Adult Social Care following the Bean Review (2016), this Review has identified and implemented further improvements in the measurement of public services productivity in the Healthcare, Education, Social Security Administration, and, Public Order and Safety services, exploiting new datasets.
  • In Healthcare, Education, and Public Order and Safety the Review has also identified key improvements which react to the impact of the pandemic on these services and better reflect the true pattern of productivity growth during and following the pandemic.
  • Material improvements in the measurement of the productivity of the Police service are possible: first by separating policing from Immigration & Citizenship services, and secondly by categorising police outputs into three themes: crime and criminal investigation, public safety and welfare outputs, and crime prevention, to provide a structure to understand the diversity of services delivered. The Review prioritised ‘crime and crime investigation’ and improvements should be implemented in 2026 following final refinement. Quality adjustments have also been considered but require data: the Review has explored the Home Office ad hoc Police Activity Survey and recommends that this survey is made regular, with the possibility of inclusion in the ONS survey portfolio.
  • Defence productivity, a long-standing conceptually challenging area to measure, is an international measurement issue rather than a UK-specific one. It has been possible to make some improvements on measuring UK Defence inputs, which will be implemented in the Spring 2025 forthcoming statistical release. Conceptual improvements to output measures have also been researched but appear dependent on work at the Ministry of Defence, who has developed a measure of output “readiness”, which is a defined and measured state of military preparedness.
  • Given the increased awareness of, and public spending on climate change, net zero and environmental protection it is clear that the classification of environmental services needs to be better brought together to aid visibility. Currently these are dispersed across multiple public services. The Review has submitted its considerations on this issue to the UN COFOG Review. The Review recommends improvements need to be in the light of the changes expected from the UN process, and anticipates this will not report until 2027, but notes this area, when fully mapped should be seen as a large service area, at least comparable in size to policing.
  • High quality data sources, which exist for several smaller local government services, including waste management, environmental services, highways and planning, would allow improvements to be made at pace, although whether at least some of these will be classed within environmental services by the COFOG UN Review needs to be resolved.
  • Historically the ONS has grouped smaller areas of government into an “Other” category, which has grown through time and is now the second largest reported unit in ONS statistics. This is analytically unhelpful, so the Review has explored how to disaggregate this “Other” category and has prioritised improving estimates of the productivity of Tax Administration working with HM Revenue and Customs, with implementation from Spring 2025. As a result, the “Other” grouping will have a lower expenditure share and in some years will now be the largest in terms of expenditure after Healthcare and Education. Further work is recommended, particularly on environmental services, which should reduce this further.
  • Social Security Administration improvements can be made by applying similar innovative methods as those for Tax Administration. That is, the ONS should measure the productivity of services which award benefits to (or receive taxes from) citizens in a comparable way. In addition, the Review identified and implemented a solution to the transition of the legacy benefits system to Universal Credit, but continued work is required to identify the transition of recipients of Tax Credits to Universal Credit.

For an ‘at a glance’ summary of improvements and further work for each service examined by the Review see Annex A.

Impact of this Review on published ONS estimates of public services productivity

  • In developing improved estimates of public services productivity the Review has worked extensively with relevant UK government departments, arm’s length bodies and other organisations in the quality assurance of these data. The impact on published estimates of public services productivity has been extensively analysed and considered. The ONS productivity statistics publications incorporated improvements in Healthcare and Education from March 2024, with further improvements being implemented in Spring 2025 for other services as shown in Table 2. Subsequent continuous improvement will be possible if recommended research is successful.
  • New and improved estimates with associated revised back series are being implemented in the annual statistical publication of public services productivity estimates by the ONS in Spring 2025. A separate, accompanying article detailing the impact on previously published estimates and any revisions is planned for release simultaneously.

 

Impact on the UK National Accounts

  • The National Statistician’s acceptance of the recommendation of the National Statistician’s Committee for Advice on Standards for Economic Statistics’ (NSCASE) for the UK to move from the European System of Accounts 2010 (‘ESA10’) to the revised System of National Accounts in March 2025 (‘SNA25’) necessitates the incorporation of public services ‘quality adjustments’, such as those used in public service productivity estimates into the UK National Accounts and compilation of Gross Domestic Product (GDP) for the first time since 2010. The work of the Review to move towards a comprehensive set of quality adjustments is a substantive contribution towards this goal. A high level ‘roadmap’ to continue this work has been compiled as part of the current ONS planning for implementation of SNA25 more widely. This roadmap will need to be maintained and updated in alignment with the final SNA25 implementation plans. The roadmap can be seen at Annex B.
  • The incorporation of quality adjustments into the UK National Accounts is one of a number of changes that will be incorporated as a result of implementing SNA25. Incorporation of quality adjustments is one of the more significant changes in measurement along with the recognition of data as an asset. The impact of these changes on UK GDP and National Accounts estimates will be evaluated by the ONS in future as part of the wider plans for implementing SNA25.
  • The ONS will also update outputs measures used in the national accounts to mirror those implemented in productivity estimates as a result of this Review.

 

Conclusions and further work

  • The Review has made significant progress delivering a step-change improvement in the published ONS estimates of public services productivity, which will be made available from the Spring 2025 release of data, improving the accuracy, granularity and timeliness of these statistics.
  • With continued focus, a number of further improvements should be achievable in 2026 and beyond. In addition a work programme to improve the coherence of the annual and quarterly public services productivity estimates in 2026, has been produced.
  • Valuable new insight into public sector workers’ use of time and managers’ views on the opportunities and barriers to use of automation and Artificial Intelligence can be obtained through pilot surveys. Future investment would be needed to add these to the ONS’ survey portfolio.
  • The coherence of devolved governments’ data underpinning the UK-wide public service productivity estimates could be improved if sufficient user need exists and investment into development is made available.
  • There is a growing network of service-specific bodies exploring public service productivity measurement at more granular levels than the ONS can deliver under current budgets, such as the Centre for Police Productivity. These propositions should make local data more transparent. Consistency of methods will require continued engagement.
  • A significant number of farther reaching recommendations for further work have been identified utilising new data and methods which could deliver further improvements. Some impact or depend upon other government departmental programmes or initiatives, and may need additional funding to bring to fruition.
  • Further step change improvements in productivity measurement are possible, particularly in Policing, Immigration, Defence and Environmental services, to continue to improve the evidence base for policy decisions relating to the productivity of public services.
Back to top