58th Report - Government services: Identifying costs

Most departments lack a sufficient understanding of their service costs and departments need help from HM Treasury and the Government Finance Function on the practical steps they can take to improve and upskill. While standards for service costing and financial management exist, they lack the necessary sophistication and are applied inconsistently across departments. There is limited practical guidance and systematic support, with some progress in granular costing seen in local frontline services like schools and hospitals, which are easier to benchmark than unique central government services. Forums such as the Finance Foundations Group aim to share best practice. The Government Finance Function (GFF) is conducting a skills assessment and working to improve its understanding of cost drivers. It recognises the need for a significant shift, aligning with its new strategy that prioritises better use of data to enhance value for money. Departments are receptive to this push, especially given the efficiency targets set in the 2025 Spending Review. GFF acknowledges it does not have all the answers and is seeking to co-produce guidance with departments demonstrating good practice. To support this, GFF will work through the Finance Foundations Group to explore what can be done in more detail. recommendation HM Treasury and the Government Finance Function should set out concrete ways in which departments must start to identify and record service costs within six months. This should include setting out what needs to be improved and practical guidance on how to make improvements.

Legacy IT systems are a significant contributory factor in the cost of government services and an impediment to being able to gather better data to bring about improvements. Departments face major challenges in collecting and analysing cost data due to legacy IT systems and siloed data structures. These outdated systems are a significant cost driver, requiring additional people and processes to work around system limitations, and hinder consistent data gathering. The Cabinet Office recognises the need for Permanent Secretaries to understand their legacy estate, while DSIT has baselined government legacy systems and reviewed departmental bids for funding in the 2025 Spending Review. However, quick wins often mask deeper issues like poor data quality and entrenched cultural barriers. The lack of common data standards across government complicates benchmarking and granular cost analysis. The Shared Services strategy offers potential for improved consistency in the future, but technical solutions alone are insufficient–cultural and process reforms will be constantly needed. There is no standard policy on time recording in the civil service, and despite ERP systems offering 4 this capability, its use remains optional. The desired levels of productivity improvements will not be achieved without effective use of time for every civil servant as would be the norm in the private sector. A shift in mindset is needed to better understand how time is spent and to make use of data for meaningful insights. recommendation a. DSIT should provide the Committee with a baselined list of legacy systems identified and the services they support alongside the Treasury Minute response to this report. b. In addition, DSIT should,within six months from the publication of this report, indicate which legacy systems should be targeted as a priority for further investigation into how far their limitations result in additional people and process costs in the operational business areas they support

The lack of Single Service Owners with accountability for all aspects of an end-to-end service inhibits departments’ ability to identify the visibility of a service’s end-to-end cost and the incentive to reduce it. The absence of Single Service Owners (SSOs) with the right mandate and visibility prevents a full view of end-to-end service costs and weakens incentives to reduce them, as focus remains on individual components rather than the whole process. Achieving larger benefits requires deeper service understanding and, in some cases, re-engineering of those services. The SSO role is not well understood across departments compared to other senior roles, and more can be done to raise its profile. There is an opportunity to build on the existing community of Service Owners and the methodology from the Top 75 programme, which included industry benchmarks for segmented digital services. Permanent Secretaries should be responsible for appointing SSOs with appropriate skills and talent, though the Cabinet Office has not set a timeframe for this. recommendation a. The Cabinet Office should, within the next six months, require Permanent Secretaries to appoint Senior Single Service Owners for all remaining services identified by the Cabinet Office and the Government Digital Service which do not yet have one in place. b. The Cabinet Office should set out a deadline by which Permanent Secretaries must complete the identification and appointment of SSOs for their remaining services. 5

There is a lack of urgency and clear leadership for resolving the legacy systems and poor data issues that inhibit the realisation of benefits through greater productivity and efficiency that departments will reap from new technology such as AI. Government transformation and efficiency efforts increasingly rely on artificial intelligence (AI), but legacy systems remain a major barrier, limiting access to the data needed for effective implementation. Despite this, departments are already experimenting with AI. Resource constraints at the centre of government mean less support is available compared to previous initiatives like the Top 75 programme. Although the digital pay framework has been revalorised, it still falls short of industry rates in London and the South East, affecting recruitment and retention. Addressing data and legacy issues requires sustained effort and resilience. AI offers potential to connect disparate data sources, but progress must be focused and purposeful–trying to solve everything at once risks losing momentum. Prioritisation and clear milestones are key to avoiding drift and ensuring that artificial intelligence delivers meaningful value. recommendation As part of its response to this report, DSIT should set out how it will take the lead on systemic data and systems improvement across government in order to exploit new technologies such as AI. At the same time, the department should set out what further steps it will take to ensure it can both secure and retain board level CDIOs and senior specialists of the right calibre given it cannot match industry pay rates in all areas of the country. 6 1 Accountability, support and guidance Introduction

On the basis of a report by the Comptroller and Auditor General, we took evidence from the Cabinet Office, HM Treasury and the Department for Science, Innovation and Technology on improving government’s productivity through better information on the costs of services.1

Responsibility for holding departments to account for improving their cost information is joint between the Cabinet Office and HM Treasury. The Cabinet Office sets the overall performance management framework expectations, including financial minimum standards.6 However, it does not systematically enforce compliance, relying instead on departments and internal auditors. HM Treasury’s role is to make sure departments and accounting officers have the guidance they require and, through the Government Finance Function, arrange and conduct end of year performance reviews.7

However, current obligations are focused on wider overall value for money and there are no specific objectives or accountability for Permanent Secretaries or senior leaders to collect detailed data on costs at the service level.8 Although the Cabinet Office and HM Treasury expect permanent secretaries to understand their cost drivers and how they evolve, the level of sophistication is often insufficient.9 Departments should be better incentivised to collect the information that is currently lacking. HM Treasury told us this could be done through clarifying accounting officers’ responsibilities for understanding their detailed costs and following this up through end of year financial assessments.10 The Treasury plans to clarify expectations for permanent secretaries in an upcoming “Dear Accounting Officer” letter.11 Providing central visibility of departments’ maturity levels would also help by shining a light on what others are doing across government.12 5 C&AG’s Report, para 1.4 6 Qq 84, 106 7 Q 84 8 Qq 105, 106 9 Qq 86, 89, 90, 94, 106 10 Q 107 11 Q 93 12 Q 110 8

The Cabinet Office told us that political sponsorship and support is considered essential as well as that of senior leaders within departments.13 This aligns with government’s goal of creating a more “productive and agile state”.14 Introducing targets for getting processes and systems in place to identify the costs of services would sharpen focus; while this would require a significant cultural shift, we were told that this would be “pushing at an open door”.15 Helping departments with the practical steps they can take

Some standards for service costing and financial management exist.16 However, we repeatedly heard that they lack the necessary sophistication and are applied inconsistently across departments.17 We heard examples of where costing activities were taking place at a more granular level, but these mostly related to front line health and education services delivered locally.18 Such services are replicated across many organisations and largely similar in nature. They are therefore intrinsically easier to benchmark for cost of delivery compared to services delivered once by a single central government department.19

We found that there is limited practical guidance and systematic support for people in departments who own the services and are responsible for identifying their costs.20 Forums such as the Finance Foundations Group aim to share best practice.21 While we heard a good description of the theory of sharing best practice, we were less convinced that the arrangements as described to us have led to tangible improvements, raising concerns that these forums may be little more than “talking shops” and not as effective as they ought to be.22

We were pleased to hear that the Government Finance Function (GFF) recognises the need for a significant shift.23 It told us that it has conducted an internal skills assessment and will recruit additional experts to help 13 Qq 147, 148 14 Qq 123, 146 15 Q 140 16 Q 85 17 Qq 86, 89, 90, 94, 106 18 Q 88 19 Qq 95, 96 20 Qq 91, 107 21 Q 91 22 Qq 92, 93 23 Q 139 9 upskill the finance function.24 This aligns with the GFF’s new strategy that prioritises better use of data to enhance value for money.25 GFF told us this upskilling will include working to improve the understanding of cost drivers and spending more time on this as an area of focus.26

GFF is seeking to co-produce guidance with departments demonstrating good practice.27 Both Cabinet Office and GFF said that departments are receptive to this push, especially given the efficiency targets they are expected to meet as set out in the 2025 Spending Review.28 To support this, GFF told us that it will work through the Finance Foundations Group to explore what can be done in more detail.29 24 Qq 126, 127, 128 25 Qq 93, 141 26 Qq 93, 139 27 Q 157 28 Qq 105, 110, 113, 141 29 Qq 98, 109, 158 10 2 Legacy systems, single service owners and overall leadership Addressing the barriers posed by legacy systems

Legacy systems are problematic for two main reasons. Firstly, they are a significant driver of cost, comprising around one-third of government’s technology estate but representing almost half of its costs.30 This may not include the excess costs that arise from people having to work manually around cumbersome, old systems and poor-quality data from fragmented data sources.31 Secondly, legacy systems are an impediment to being able to gather better cost data as data extraction often requires manual effort as repeated point-in-time exercises.32 The Cabinet Office recognises the need for Permanent Secretaries to have a good understanding of their legacy systems. The Department for Science, Innovation and Technology (DSIT) has baselined government legacy systems and reviewed departmental bids for funding in the 2025 Spending Review to mitigate the risks such systems pose. It told us that although there has been a step- change in how senior leaders are thinking about their legacy systems, there were not as many bids for funding as expected.33

“Quick wins” can be valuable. However, we are concerned that they often mask deeper and more persistent issues like poor data quality and entrenched cultural barriers. As the Cabinet Office has itself identified, these will need ongoing tenacity and effort if substantial progress is to be made in addressing them.34 30 Q 115 31 C&AG’s Report, para 4 32 Qq 139, 149 33 Qq 114, 115 34 Qq 111, 125, 150 11

The lack of common data standards across government further complicates benchmarking and granular cost analysis because it makes data more difficult to analyse and interpret.35 GFF told us that the introduction of new Enterprise Resource Planning systems and new data standards for back-office processes such as finance and human resources offer the potential for improved consistency.36 However, this will not necessarily address the costs of services provided by operational departments to citizens and businesses. Furthermore, technical solutions alone are insufficient–we heard that cultural and process reforms are also needed.37

Departments do not generally have a good view of the additional time and people costs associated with specific business processes or different stages of the customer journey.38 We were therefore surprised to learn that there is no standard policy in the civil service on how staff time is tracked, and despite many systems offering time recording capability, its use remains optional.39 A shift in mindset is needed to better understand how time is spent and to use the data collected for meaningful insights on where the costs of services are high and potential efficiencies and productivity improvements can be made.40 Appointing Single Service Owners

It is difficult for departments to gain a full view of the end-to-end costs of a service where there is no Single Service Owner (SSO) with overall mandate and responsibility for the service concerned.41 This also weakens the incentives to identify and reduce costs overall, because the focus is on the individual component parts of a service rather than how they fit together as an overall whole.42 While we heard “quick wins” can give a sense of progress, more substantial benefits require a deeper understanding of a service and in some case, re-engineering of the service.43 There is an opportunity to learn from the methodology used by the Top 75 programme, which included industry benchmarks for segmented digital services.44 35 Q 149 36 Q 140 37 Qq 140, 150 38 C&AG’s Report, paras 1.6, 1.7 39 Qq 135, 136 40 Qq 137, 138 41 Q 143 42 Q 144 43 Q 111 44 Qq 87, 108 12

The SSO role is not well understood across departments compared to other senior roles in departments. Both Cabinet Office and DSIT told us that more can and should be done to raise its profile and ensure that an SSO is appointed for “every single service”.45 It is important to ensure that SSOs have the appropriate authority as well as the right skills for the role. The Cabinet Office told us that it has already identified where an SSO is necessary, but ultimately Permanent Secretaries should be responsible for their appointment. The Cabinet Office does not yet have a timeframe for this but has committed to set one out in response to this report.46 Pace and leadership to enable benefits from new technologies

Government transformation and efficiency efforts are increasingly reliant on newer technologies such as artificial intelligence (AI).47 Legacy systems and poor-quality data remain a major barrier to reaping the full benefits.48 We heard that addressing data and legacy issues requires sustained effort and resilience and remain concerned that the pace of addressing the challenge is not keeping up with what is needed in today’s world and that despite this, departments are already experimenting with AI.49

Resource constraints at the centre of government mean that less support is available to departments compared to what was available under previous initiatives like the Top 75 programme.50 We have often observed on this committee that government pay rates cannot match what highly skilled people could earn in the private sector. Government has attempted to address this through the digital pay framework, which DSIT told us has recently been uplifted to market rates.51 However, we are concerned that it still falls short of industry rates in London and the South East, and this will affect the recruitment and retention of people with sufficient seniority and calibre to lead the transformation required.52

Progress must be focused and purposeful. We accept that trying to solve everything at once risks losing momentum, and that having a roadmap is helpful. Nevertheless, prioritisation and having clear milestones to aim for are key to avoiding drift in addressing government’s 45 Qq 132, 144 46 Q 133 47 Q 137 48 Q 115 49 Qq 115, 145, 153 50 Qq 122, 145 51 Q 128 52 Qq 131, 134 13 long-standing legacy and data issues, so that are in a better position to reap the full benefits of productivity-enhancing technologies such as artificial intelligence.53 53 Q 154 14