HoC 85mm(Green).tif

 

Public Accounts Committee

Oral evidence: Digital transformation in government: addressing barriers to efficiency, HC 1229

Monday 22 May 2023

Ordered by the House of Commons to be published on 22 May 2023.

Watch the meeting

Members present: Dame Meg Hillier (Chair); Olivia Blake; Dan Carden; Sir Geoffrey Clifton-Brown; Mr Jonathan Djanogly; Mrs Flick Drummond; Nick Smith.

Gareth Davies, Comptroller and Auditor General, National Audit Office, Yvonne Gallagher, Digital Director, National Audit Office, Adrian Jenner, Director of Parliamentary Relations, National Audit Office, and David Fairbrother, Treasury Officer of Accounts, were in attendance.

 

Questions 1 - 80

Witnesses

I: Alex Chisholm, Chief Operating Officer, Civil Service, and Permanent Secretary, Cabinet Office; Megan Lee Devlin, Chief Executive, Central Digital and Data Office, Cabinet Office; Paul Willmott, Chair, Central Digital and Data Office, Cabinet Office.

 

 

 

 

 

 

 

 

 

Report by the Comptroller and Auditor General

Digital transformation in government: addressing the barriers to efficiency (HC 1171)

 

Examination of witnesses

Witnesses: Alex Chisholm, Megan Lee Devlin and Paul Willmott.

Chair: Welcome to the Public Accounts Committee on Monday 22 May 2023. One of Governments main challenges is achieving efficiency savingssomething that is this Committees bread and butter. In 2020-21, central Government Departments expected to spend £456 billion on the day-to-day running costs of public services, grants and administration. In order to try to make that work better, the Government have established a new Central Digital and Data Office, the CDDO, which the NAO, in its Report, has found is laying good foundations, but many of the critical and more challenging milestones still need to be tackled.

We have looked at this as a Committee before, particularly with reference to certain Departments, finding that there are many legacy systems from the 1970s and lots of issues about how data is collected. It is a huge task across Government, let alone in areas like the Home Office and the Ministry of Defence, where there are additional challenges that we have looked at separately.

Today, we are doing an overview of the wider subject, how it is progressing across Government and what the barriers and challenges are. I would like to welcome our witnesses: from the Cabinet Office, we have Alex Chisholm, the chief operating officer for the civil service and the permanent secretary at the Cabinet Office; Megan Lee Devlin, the chief executive of the Central Digital and Data Office; and Paul Willmott, the chair of the Central Digital and Data Office.

Q1                Nick Smith: Welcome, everybody. All power to your elbows—we really want you to do well with this important task. The prize is huge. Introducing digital services and improving public services is something that everybody wants to see, so thumbs up for your efforts.

However, according to page 4 of the NAO Report, under Key facts, just 4% of civil servants are digital professionals, compared to between 8% and 12% as an industry average. There have been 11 cross-governmental digital strategies launched over the past 25 years, all seeking to address usability, efficiency and legacy systems. There has been a reduction of 20% in digital, data and technology apprenticeships from October 2021 to 2022a very difficult backdrop to your work. What is your view on the progress that Departments have made in delivering the roadmap?

Alex Chisholm: Could I echo your words that we are very keen to see that transformation of digital services? It is so important to the delivery of top-quality services to citizens today. We also really appreciate both this hearing and the Report that we had from the NAO, which we agree with and support. Indeed, it is one of a series of NAO Reports on digital that have really been very helpful to us in guiding our thinking in this area.

In the NAO Report, you were kind enough to say that this is a good quality roadmap, and so you should, I say, because it incorporates a lot of the recommendations and findings from previous PAC hearings and NAO Reports. That is a good plan. The question is, can we now deliver against it?

Nick Smith: And how are Departments doing?

Alex Chisholm: Yes, and how are Departments doing? The fair answer is “Good, so far. We launched the plan last summer. We did a six-month stocktake and reported on that in the NAO Report. The things that have been most impressive and important, I suppose, are first of all that we have not just set aside a lot of money for legacy remediation, which is £2.6 billion, but that Departments have committed. We have been through 153 legacy projects and identified all the top high-risk ones. We have funded and programmed plans for remediation, which we have been through and are satisfied with. That is a big step forward. A number of your predecessors and the NAO have said, “Just tell us where the legacy is and what the plan is to remediate it, so that is really important progress.

A second point that I wanted to highlight is that there are about 7,500 services out there, and we have picked out the top 75, or 1%, which account for the vast majority of services that affect the public and are of the highest value and the highest risk. Those are the top ones. When we were talking to the NAO during its work five or six months ago, we had baselined 15 of those with Departments. We have now done all 75, so we have made terrific progress with Departments, which have been really keen to work with us on that.

That gives us now a very good basis to understand where we are already great, which we are in a relatively small number—about 10—of those, I have to say.

Q2                Chair: Which ones are great?

Alex Chisholm: They include two in the Cabinet Office, I am proud to say.

Chair: Blow your own trumpet—why not? Use the chance.

Alex Chisholm: There are another eight or so elsewhere. There are about 20 that we think are good, and the remainder45 or sohave a long way to go to get to the high standard that we want. Rather than being defensive about that, I should say that we have set this standard about the cost to serve, the speed of the performance for users, the error rate and all these other standards to make sure that you are genuinely great if you hit all of those things. Once you are great, you want to be better again, so I would regard that as very important indications of the support from Departments, that they have gone with us and agreed to that end-to-end baselining.

The third area of big progress with Departments, even since this Report was doneI know it was delivered in March, and the research a few weeks before thatis that the One Login programme has continued to be rolled out and we now have over a million users of that. Again, that is terrific progress in a relatively small space of time.

Rightly, the NAO Report asks, “Can you continue to keep up that pace of progress? Can you continue to get the top-level leadership? Can you continue to make sure that this gets the priority for resources that it needs?” So far, there has been good progress and no complaints.

Q3                Nick Smith: Thank you, Mr Chisholm. I have a few more supplementary questions before we go to your colleagues. You talked about Departments having the money to crack on. You have talked about having identified legacy problems. You said that a bunch of work that was taking place was great. Why was it great?

You also said that of the 75 elements that you were working on, 45 still had a long way to go. Tell us what was great, so that we can learn about that, and then tell us more about these 45 that had greater progress to make.

Alex Chisholm: Let me introduce Megan, who is chief executive of the CDDO. [Laughter.]

Chair: It is a difficult question.

Alex Chisholm: If I get the chance later, I would be happy to talk about the four Cabinet Office ones.

Chair: That is the second time today. I think we should just acknowledge that the Cabinet Office has some good work going on, to save Mr Chisholm from doing it himself.

Megan Lee Devlin: Good afternoon, Committee, and thank you for the opportunity to appear today. The assessment that Mr Chisholm references is what we call the great framework for service performance. It uses industry-standard metrics for usability and efficiency to assess the performance of live services.

To give you an example, usability is assessed through metrics such as citizen satisfaction scores or user satisfaction scores, as well as digital adoption rates and digital completion ratesin effect, how many people start an experience online and how many end that experience online, which is a good indicator that that service is a positive experience for the person who is seeking an outcome from Government.

In terms of efficiency metrics, we look primarily at cost to serve, or cost per transaction, and we are working really closely with departmental teams in all corners of Government on those top 75 services to apply that framework, to assess performance against it, and to understand the opportunities for improvement and how we can support them to achieve it.

Q4                Nick Smith: My question was initially about how well the Departments are doing with these important qualities that you have identified. Could you tell us a bit more about that?

Megan Lee Devlin: In terms of service transformation, we have set a target that 50 of our top 75 services will be at a great standard by 2025. We are in the process of doing our one-year update on the roadmap, given that we published it last June. As part of that, we have been reviewing performance, as Mr Chisholm notes, of all 75 services. Approximately 10% of those are great at the moment, 20% are in the good category, and the remainder are in the requires improvement category. That was largely in line with expectations. We know that our users want and, indeed, expect a high degree of service performance. Therefore, we are working very hard to make sure that our services reach those expectations.

To give you an example, when we first did the assessment against that framework and the standard last summer, one of the services from HMRC, for instance, did not meet that standard. It was a PDF form that a user, who might be in a very difficult position because it was child benefit, had to print and either take in person or post, in order to access the service. I am very pleased to say that there has now been a significant improvement. It is a digital service; it can be applied for online, which will absolutely have an impact on the usability and efficiency scores.

Our role in the Cabinet Office is very much about setting those standards and ensuring that the frameworks and support are in place for departmental teams to be able to meet what we have defined as a great service.

Q5                Nick Smith: Are you confident that you will reach your 50% target of teams getting up to be great services by 2025?

Megan Lee Devlin: The roadmap was very intentionally ambitious. In designing it in that way and setting some measurable metrics, we took on advice from colleagues in the National Audit Office that we needed to strike the right balance between high ambition and achievabilityhow realistic those targets were. We made sure that it was deliverable within spending review settlements. We are in a very difficult economic environment and Departments are making difficult decisions there. In line with the most recent NAO Report, we are ensuring that we are continuously reviewing deliverability.

It is an ambitious target and, as the NAO Report highlighted, there are challenges around capability and making sure that we can attract and retain the right skills. That is going to be essential, as is retaining the right level of business focus on those transformation activities.

Q6                Nick Smith: That sounds like possibly.

Megan Lee Devlin: Possibly.

Alex Chisholm: Just to offer some perspectives that may be of interest to the Committee, I have seen a long period of involvement with Government digital services. This spending round period has been much the most positive and significant, because it is the first time that we have really been able to invest at this level—£8 billion, of which £4 billion was for digital transformation. Those were bids that came in and were assessed by Megans team and the Treasury working together, and the best ones got their money. That is very good.

I have also been really impressed by the way in which Departments over the last two years have responded to the challenge from us and others saying, “You are too reliant on external resources, consultancies and contractors, and we have begun to develop better in-house capabilities.

Chair: We are going to come to that a bit later.

Alex Chisholm: Fine, but I wanted just to highlight that as an area where we have made really palpable progress. They have been very focused on cyber risks and improving the customer experience.

Where we have the next chapter to go, if you like, which is recognised in the NAO Report, is that we do not have great visibility and control over end-to-end services. Also, a number of those are not really fully digital and require quite a lot of manual administrative effort and toggling between systems, which is frustrating for people working in those admin roles, as well as for the end user.

Chair: We have also seen a number of examples on this Committee of things that can go wrong where there is no end-to-end oversight, which we will come to later.

Alex Chisholm: Indeed.

Q7                Nick Smith: Mr Willmott, what do you have to say?

Paul Willmott: First of all, progress has been good. Departments are at the table, literally, with the digital and data board, and engaging on all of our six missions. We remain on track, but I do agree with the points picked up in the NAO Report that there are risks to delivery that we need to take seriously, and I would point at two in particular.

The first is leadership engagement. Senior civil servants have an awful lot on their plate, and we need to continue to push for priority. The second is capability in the senior leadership group, which we are addressing through our various training programmes, but which need continual attention.

Q8                Olivia Blake: I just want to ask about that £4 billion. Is that all going on the 1%—the 75 projects?

Alex Chisholm: No. In fact, three main categories for bids were looked at in SR21digital transformation, legacy remediation and data. They are not perfectly distinct categories, but, broadly, that is the way we went about it. That was SR21, so it would have been set in November 2020, whereas this particular process of the top 75 came after that. Of all the different areas of spend, some were already identified and properly set up.

I know that you have looked at personal independence payments and very complicated transformation programmes like that. Future Farming is another one that you have looked at; that would have been funded as part of SR21. In other cases, we looked at the top 75 and said that we do not see a funded programme, but we see a gap, and that is where we want to focus our attention.

Q9                Olivia Blake: How much would it cost to get all 7,500 up to the standard that they need to be at?

Alex Chisholm: It is probably an infinite amount.

Chair: Ms Blake is giving you an opportunity to bid to the Treasury here, but “infinite” might not get you across the line.

Alex Chisholm: Todays brand-new is tomorrows legacy, so you are always at it in that sense. We have tried to look at it more from the point of view of asking, “What is the standard that you need in terms of the performance level for the end user? What is the cost to serve? What unacceptable risks are we are taking with legacy?” If we focus on those areas, we will be spending our money wisely, I would suggest.

Q10            Olivia Blake: You are getting to the root of the problem, then. I just want to move on to what skills shortages there are and the impact they are having. What are the most pressing skills shortages that are holding up the process, Mr Chisholm?

Alex Chisholm: Let me start, and Megan will probably want to come in as well. First of all, it is the overall volume of people working in DDAT, as we call it, or digital, data and technology—you will be familiar with the acronym that we use. As the name already describes, you have quite a wide variety of people with those skills, some of whom would have—again, apologies for the jargonfront-end skills, who are people working in the presentation of materials on the web. Other people might be software engineers working at the back end.

Within that community, there are people who are looking after infrastructure in very practical ways, others who are developing and coding new services, and others who are data scientists and the like. It is quite a big community. That said, last time we measured, in 2021, we found 19,000 and, when we did so again at the back end of last year, we were up to 26,000, which is a very big increase in the number of people identifying as members of that digital, data and technology function.

Q11            Olivia Blake: How many would be the best complement for the work that needs to be done?

Alex Chisholm: This is the fact that was picked up by the member earlier: overall, it is about 4.5% of the workforce. When we do benchmark tests with other professional services-type organisations, we tend to get a range from 8% to 12%. Even though we have a lot more than we had, when you ask how many people you could have doing digital roles, it is probably the same again. From a taxpayer perspective, you will want those to be substituted rather than on top of.

Q12            Olivia Blake: Ms Lee Devlin, what is the most pressing shortage out of all those roles that have just been described?

Megan Lee Devlin: It is important to look at it through two lensesboth technical skills and for the generalist civil servant population. In terms of technical skills, the roles that we find hardest to recruit are software development, engineering and architecture roles. In the generalist population, we know, through a survey that we ran in collaboration with industry, that 75% of civil servants would like more digital and data training, so we certainly see appetite among the broader civil service for greater exposure to digital and data essentials.

However, in recognising the extraordinary efforts of many of my colleagues across Government, since launching the roadmap we have seen some real progress in terms of skills. For instance, the size of our profession in terms of roles has grown by 10%. We have reduced our vacancies from 17% to 13.6%. We also now have more than 80% of roles outside London, which exceeds the target.

Olivia Blake: Is that 18% or 80%?

Megan Lee Devlin: It is 82%. There has been real progress and a battle hard won by many colleagues across Government Departments working hard on insourcing skills, but there is, of course, a lot more work to do. We are specifically focusing our efforts on those harder-to-recruit roles in terms of early career talent—for instance, our software developer programme for graduates. We are also making sure that we do more bulk and brigaded recruitment, so that we can go to market once and attract a vast number of colleagues, who we can then seed to Departments. We are doing that specifically for data architects at the moment, for example.

Q13            Olivia Blake: We will talk later on about recruitment and attracting, but I do have a question about benchmarking of pay and just how we compare in the UK Government and civil service with other civil jobs internationally, as well as with the market and the private sector, which I know we are very far behind. I would just like to know if you have any information on both of those sectors and how that is affecting recruitment.

Megan Lee Devlin: In this context, I would like to share a little bit on our digital, data and technology pay framework. We currently have 32 organisations across Government on the pay framework. That enables us to pay more market-competitive rates for specific capabilities across the DDAT profession. The average total increase for those colleagues who are on the pay framework is around 30% compared to the average of civil servants within that pay grade, so it is certainly an increase and does help to improve the attractiveness of the civil service to technical talent.

We have seen real success as a result of using that pay framework. For instance, in DVLA, we were able to shift the balance of insourced capabilities, so civil servants on the ground, to suppliers from about 50:50 to 95:5. It is extraordinarily helpful in that context, and hence we are working closely with colleagues across Government to ensure that all Departments are adopting that framework.

Your question had two parts, firstly around how it compares to other civil servant roles. Hopefully that has demonstrated that we are able to pay more competitively through that framework. As compared to other Governments, there are a range of models used. In some Governments, such as Singapore, they pay at market rates for technical skills—indeed, for all of the civil service. It is very country-dependent, but certainly, here in the UK, we are ensuring that we can be competitive in the market.

Alex Chisholm: Just to reassure the Committee, the DDAT pay framework is entirely self-funded from the savings we make from not employing extensive consultancies and contractors.

Q14            Olivia Blake: That is good to hear. It would also be useful to know whether you feel there is a brain drain in the same way that we talk about in other areas where we struggle to recruit. We are losing nurses to Australia, for example. Do we have that situation here, or are you confident that the offer is strong?

Alex Chisholm: It is probably right to recognise that cyber skills are at a super-premium in the market, and there are a great deal of contests with other organisations for that. Some of the people we get in, who work both on the agency side and within the Government security group, are sometimes made very tempting offers by people externally. That is a particular area.

It is also very important for people to be able to look at the senior levels of the organisation and see opportunities for themselves. It was the case a few years ago that the top of the digital tree stopped a bit short of the executive committee. That is not true now in most of the big Departments, and not true now in the Cabinet Office. As well as being able to attract top talent in from the industry, looking to my left and to my right, we have also been able to develop long-term career prospects for people to say, “Gosh, I can see myself as the chief digital information officer of this Department or this public organisation in the future, and that is good for long-term motivation as well.

Q15            Olivia Blake: That is a good answer, but I just want to push back, because I dont think I have heard the answer to my initial question, which was about what activities the lack of skills is stopping from happening. What is the fundamental impact of the skills shortages that you are seeing?

Alex Chisholm: When you look at the opportunities for digital transformation, they are three or four times the performance that we have now. If we do not have the right resources and capabilities at all levels, we tend to make incremental changes, as is well documented in those NAO Reports. That is a correct analysis and therefore we are trying to invest, through the upgrade, in our skills and capabilities. We are also very focused on that transformational potential of those key services, so rather than making a small incremental improvement, we make a genuine transformation. It is probably that opportunity cost rather than something that is a bit more visible.

Q16            Olivia Blake: Just to challenge you again, are there any activities that are at the top of the list for what you would stop or defer if you were to continue to have the challenges in skills? How would you respond, if you cannot recruit?

Megan Lee Devlin: I would emphasise that that adds risk to the deliverability of the strategy and our competitiveness in the market. It is incredibly challenging, even for the leading private sector organisations, to attract the right technical talent. It is a real challenge. We are ensuring that we keep across that, and the NAO Report highlighted the importance of those continuous deliverability checks in light of the current mix of DDAT skills.

As I mentioned earlier, we were really pleased with progress at the six-month mark, and I am really quite proud of the work of colleagues across Government and, indeed, our own team in the Cabinet Office in achieving that. We will be doing a similar exercise—indeed, it is mid-flight at the moment and coming up to the one-year mark of the roadmapto scrutinise progress to date and look ahead at our priorities for years two and three in light of the current skills mix and challenges there.

Q17            Olivia Blake: Is there a mechanism to scale down some of the projects that you have going if necessary, if you do not have the skills to deliver them?

Megan Lee Devlin: If I take top 75, for example, they are spread across all central Government Departments and, indeed, some arms length bodies. We have 26,000 colleagues across Government Departments who are very committed, purpose-led technologists working quite tirelessly to achieve our transformation aims.

We know that that is not enough, as Mr Chisholm has pointed out, but we are tracking progress, particularly through that performance framework that I spoke to earlier, so that we can stay on top of that and call out risks as and when necessary. Indeed, the NAO Report highlighted that we did that at the six-month mark. We had a conversation with the permanent secretaries through the digital and data board to raise potential risks, one of which was the continued challenges around recruitment and retention.

Alex Chisholm: I will try to give as succinct an answer as possible, because it is a very good question. The real consequences of not having enough capable resources are to defer opportunity and to prolong legacy risk. That is it, in a nutshell.

Q18            Chair: Once you have started a project, you are determined to see it through. The top 75 will keep going.

Alex Chisholm: Yes. Let us say that you needed 100 software developers to deliver this important transformation, and you had 20. Rather than giving it up entirely, you might just go at much slower rate. That is the same, I fear, with legacy remediation.

Q19            Chair: There might be a risk, then, that if you go that slow you are out of date with what you were originally planning.

Alex Chisholm: Exactly, and that is one of the reasons why, when we did the Organising for Digital Delivery review in 2020, we introduced these quarterly business reviews. You might say that it is a relatively straightforward discipline, but we were concerned that there was too much annularity and that you would find out a year on that something was running behind in the transformation and, rather than having a September delivery date, it was suddenly March and six months were lost.

The same is true with legacy. One Department was funded to deal with five major legacy programmes and, in the course of a five-year spending review, none of those was remediated. It is those problems in the past that we are trying to confront here, and that is why every Department now has a quarterly business review with CDDO, and jointly with Treasury, again reflecting learning from the past, to make sure that money given to Departments is used for the purposes intended. Otherwise, the temptation, when you are dealing with things like legacy, is to extend and use the resources for something else.

Q20            Chair: I am very interested that that is now happening. Is that because you have Mr Willmott and Ms Lee Devlin there? I do not want to be rude about mandarins who did not have this experience before. Is it because you have specialists who have come in and said, “A quarterly review is what you need”, or is that something that the system worked out for itself?

Alex Chisholm: Organising for Digital Delivery was a report done for us by the Digital Economy Council, led by Doug Gurr, so it may have had some influence from the private sector-type approach to running businesses.

Q21            Chair: Mr Willmott, is there anything that you wanted to add from your perspective on cutting across private and public sectors in that approach?

Paul Willmott: It is very common practice in the private sector to assure progress on a more frequent basis. Typically, you see different levels of governance as you cascade down the organisation. You might see a semi-annual review of project progress by the board, and then a quarterly review, and every few weeks down at the working level. We are moving towards that kind of system, which provides better value for money, because you can redirect projects earlier, but also better outcomes for users, because you can course-correct to what users need.

Q22            Olivia Blake: Would you hope, Mr Chisholm, that you are moving away from sharp surprises at the year end to more of a planned approach to the issues and challenges around skills?

Alex Chisholm: Yes, and I would also emphasise that it is good on both sides for more of a partnership-type approach. I have had experience on both sides of that conversationat the centre, trying to review progress in Departments, but also in those Departments being reviewed. It can sometimes feel like a dialogue of the deaf, which is what we do not want. You can avoid that by having a regular quarterly engagement to say, “These are the plans we are familiar with. How have you progressed those over the last 12 weeks?” That is a much better approach than, as you say, once-a-year accounting for use of money or trying to get more.

Q23            Olivia Blake: I found it interesting to see in paragraph 3.26 on page 34 that you have had a 20% drop-off in apprenticeships. I wondered what the reason for that might be: is it quality, or is it just a change in the way that it has been done? What is the cause of that quite significant drop-off?

Megan Lee Devlin: You are right to point out that we had a decrease in apprentices, from 800 to 637, between the April 2021 and October 2021 workforce commissions. I am very pleased to say that that trend has reversed in the following six months, and we are now at 679 apprentices as of the end of March this year.

To address your question quite specifically, the decrease was due to headcount restrictions, which meant that Departments had to make some challenging trade-offs around hiring specialists who are more trained, or apprentices who require a bit more investment in skills. That was very challenging for us as a function.

Our apprentices are a core part of our technical community. Indeed, the Government chief technology officer, who we hired in CDDO quite recently, started his career as an apprentice in HMRC, learning to code. He has since had a very successful career in the private sector and is returning to Government here in CDDO.

We know that there are some challenges around apprentices. We would like that number to be moving at a faster rate in terms of our apprentice community, and so we are looking at our choice of providers for apprenticeship programmes in order to ensure that that is as strong as it can be. We are also ensuring that Departments have the managerial capacity to develop and support apprentices as they come in, noting those challenges around capacity that Mr Chisholm alluded to earlier.

Q24            Olivia Blake: Are there standards set around the apprenticeships that are offered, or is it a bit piecemeal, Department by Department?

Megan Lee Devlin: There absolutely are. We work very closely with civil service HR on those standards and ensure that Departments are aligning to them.

Q25            Nick Smith: Ms Lee, you had 800 apprenticeships in October 2021. Even though you have increased them following that drop of 20%, you are still well below that 800. Until now, I have been hearing you say, “We are thinking long-term. We are trying to keep our work well organised and increase our technical staffing, but it is still a shocking drop-off, is it not? How on earth did you allow that to happen?

Chair: Where did they go?

Megan Lee Devlin: I would again point to the headcount restrictions, which led to difficult decision making. As we have noted, we have approximately half the industry average when it comes to the capacity of digital teams to deliver, and yet the work of many digital, data and technology colleagues across Government is underpinning live services that are crucial to the public accessing the outcomes they need. My colleagues across Government had to make some difficult trade-offs there.

Q26            Chair: To be clear, you have dropped apprentices to get more skilled people.

Megan Lee Devlin: Each Department made its own resourcing decisions. We did, of course, work closely with them through the quarterly business review process to track that over time. Indeed, that process enables us to detect these sorts of movements through what we call our digital dashboard, which measures a key set of metrics for each Department across digital, data, tech and capability. I acknowledge that we want there to be a higher number of apprentices in our pipeline.

Q27            Nick Smith: I am sure that you do, but, just for accountability, whose headcount restrictions were they? Who is carrying the responsibility for this decision, please?

Alex Chisholm: That would be in Departments and public bodies. Just to keep it in perspective, while we share your concern that the number of apprentices went down by 170, that is in the same period where we were halfway through an increase of 7,000 overall capabilities. As my colleague said, it looks that there would have been a shift at the margin towards slightly fewer apprentices and slightly more highly skilled people. We are now increasing both categories, which is good.

Q28            Nick Smith: So decisions are taken by the Departments.

Alex Chisholm: Yes.

Q29            Nick Smith: Can you let us know which Departments held on to their apprentices and which did not, so we can chase the accountability down to that next level if possible?

Alex Chisholm: We have that data.

Q30            Sir Geoffrey Clifton-Brown: Good afternoon, everybody. We all want you to succeed, but the Report is not uncritical. Therefore, you would expect from me not uncritical questions.

First, on top of the reduction in apprenticeships that Ms Blake and Mr Smith have outlined, paragraph 27 on page 14 says, “Only 4% of civil servants are digital professionals, compared with an industry average of between 8% and 12%. You have to have people at the top of every Department who are going to buy into this and have sufficient skills to get it to work.

That paragraph then goes on to say, The number of Government digital, data and technology vacancies rose by 7% from 3,900 in April 2022 to 4,100 in October 2022. The number of digital, data and technology apprentices has droppedas we have heard“from 800. Not only has the number of apprenticeships dropped, but the number of qualified staff throughout the Departments has as well.

I heard what you said in answer to Ms Blakes question about benchmarking, but if you are not going to pay the market rate as opposed to the civil service rate, how are you going to recruit the sufficiently skilled people you need?

Alex Chisholm: Let me start, and I can tell that Megan is keen to come in as well. Just on the number of people, I know that the Report, which is perfectly accurate, does say that the number of vacancies increased slightly. The number of underlying posts has increased a lot over this timeframe. As I say, that is the move from the 19,000 in 2021 to the 26,000 in 2022. The demand is going up, so we have more digital people than ever, but we need even more than that.

On the figure of 4%, that is absolutely the case. We probably need to double the number of people with digital expertise working in Government. Some of that will be in training up people we have now who do not have great digital expertise to acquire that. We will be shifting the type of people we are recruiting over time, and general skills administrators and caseworkers will become less valuable relative to people with coding and data science skills. That is a shift that lots of other organisations are experiencing, so we just need to be part of that.

The leadership at the top is very important. In some cases, we have been able to get people with a high level of digital expertise acquired elsewhere, which is true of the three of us. In other cases, we have had to train that in, so to speak, for people who had not had that previous experience. That is why, elsewhere in the Report, it mentions that we have a digital excellence programme for our SCS. The number of people who have now done that is 300, which is a promising start, but there is a long way to go. We want to make sure that all of our senior leaders are sufficiently comfortable to work in the digital domain.

Finally, part of the thinking behind the digital and data board, which Paul Willmott mentioned and which he co-chairs, is that we cannot only have fora that bring together digital experts to talk about their own world. You have to make sure that the people with the biggest control over resources and the authority in their organisationspermanent secretaries and chief operating officersare fully aware of the opportunities of digital, as well as the responsibility to deal with those legacies. The board is co-chaired by Jim Harra—again, a long-standing permanent secretary.

Sir Geoffrey Clifton-Brown: We are coming on to the board in a minute.

Alex Chisholm: The philosophy behind it is to try to make sure that the top of the shop, so to speak, is digitally aware.

Q31            Sir Geoffrey Clifton-Brown: You did not answer my key question: how are you going to recruit highly skilled people if you do not pay the market rate?

Alex Chisholm: We pay a premium to the regular civil service rate through the DDAT framework that you heard about from Megan a moment ago. We also findand I have done a great deal of recruiting myselfthat people are attracted to work for the Government because the scale of what we do is literally incomparable. No private sector organisation has to deal with the range of things that we do. The impact of what we do can be genuinely transformational, so it is very interesting and rewarding work. We try to make sure that we pay something. It is not going to be the most that you can get out there. We are public servants; we do not expect the best pay, but it is more competitive than it would be if we did not pay this premium for DDAT skills.

Our other responsibility is to make sure that people are engaged really productively with exciting, agile projects that achieve their goals, and do not feel that they are getting frustrated by the overall weight of administration. Those things are very important to retention and to the point that I mentioned about being able to develop your forward career and feel that you could have 10 or 20 years in Government as a digital executive rather than just a short period of time.

Q32            Sir Geoffrey Clifton-Brown: With that answer, you are trading on peoples goodwill. They might do it for a year or two, until they get a really attractive offer from the private sector. They are in a senior job within a Department. They are part of a key team, and then they are poached by the private sector. If your pay is consistently below what is being paid outside in the private sector, you are going to be continually vulnerable, are you not?

Alex Chisholm: You are absolutely right that there is a challenge there. Because Megan has come from the private sector and has recruited a number of people in recently, could I ask her to make a few comments?

Q33            Chair: Ms Lee Devlin, what made you come to the public sector?

Megan Lee Devlin: As Alex rightly notes, the scale of the opportunity in Government is enormous, and it is deeply rewarding. Before I answer that specifically, to come back to you on the numbers around our vacancy rates, the dichotomy is that while our vacancies have increased by 3% over that period, which was called out in the NAO Report, our total number of roles has increased by 40%, so we are growing the profession in terms of total number of DDAT professionals in Government. Naturally, as we expand the number of roles, we have vacancies accompanying that, and so for every 10 roles that we have added, that translates into one vacancy.

To your question around pay, which is, of course, a challenge, we work very closely with the Treasury to set the annual rates for DDAT pay, which is aligned to specific capabilities within our DDAT capability framework, and which acknowledges degrees of technical specialism and seniority. We work very hard to ensure that those are as competitive as possible. We are currently in the process of working with Treasury on the annual revalorisation of that framework, so it is something that is iterated and adapted over time to make sure that we stay in line, to the extent that we can, with market developments.

Q34            Sir Geoffrey Clifton-Brown: Mr Willmott, let me take you to the end of paragraph 30, on page 15: “Stronger digital expertise and capacity-building, sustained support from the centre of Government and the continued goodwill of Department senior business leaders are needed to maintain momentum. Without these”— I would ask whether you agree with the Report on thisthe roadmap will peter out as its predecessors have done and Government is unlikely to address the systemic issues and achieve the efficiencies the roadmap has identified. That is a fairly dire statement. Is it a real possibility, or will you be able to go on recruiting sufficiently skilled people to keep the roadmap going?

Paul Willmott: I remain optimistic. It is true that we have to have the right level of support from Departments, and that Departments need toas we have discussed already and as comes out in the Reportshow leadership and to build the skill base in their senior ranks.

As CDDO, we remain committed to that and, via the board, are holding Departments to account on a regular basis around that development. We need to continue to sustain that momentum and make sure that those issues are addressed, but I will not argue that it is simple or that it gets addressed overnight.

Sir Geoffrey Clifton-Brown: If it were simple, it would probably have been done by now.

Q35            Chair: Before you came in front of us, Mr Willmott, you thought that the existence of the CDDO was making a difference. If you were not there, would it be embedded enough to continue? Let us say that it was left to Departments to deal with. Would the system revert to its old ways of working?

Paul Willmott: There are two parts to that. One is that there are some cross-governmental enablers that CDDO has put in place and only CDDO can deliverfor example, data standards, a common framework for legacy risk, recruiting, since most talent moves around between Departments and pay frameworks, and so on. Leadership must come from Departments as well, and many Departments are stepping up. They understand the size of the opportunity and of the risks, but as we have all agreed, it is a journey.

Chair: Everyone is on a journey.

Q36            Sir Geoffrey Clifton-Brown: If I could then take you to paragraph 2.18 on page 27, perhaps I would address this to you, Ms Lee, because you are really into staff and training and everything else. Training is critical to all of this. The paragraph says, “To date, around 200 non-specialist executives have received digital awareness training, but the target set out in the roadmap was more than 6,500CDDO has proposed that the digital and data board ask all their senior civil servants to promote digital training in 2023-24. This comes to my earlier point about buy-in from permanent secretaries downwards. Are you getting sufficient buy-in from permanent secretaries and senior directors in order to be able to make this, or is it variable throughout Departments?

Megan Lee Devlin: We have fantastic support from permanent secretaries, particularly our digital and data board, which Paul chairs, alongside Jim Harra from HMRC. We have also appointed a range of permanent secretary executive sponsors for each of the missions. For the capability mission, that is Matthew Rycroft from the Home Office. Those permanent secretaries meet through the digital and data board every two months to discuss progress against the roadmap, together with any risks, challenges and issues. There absolutely is senior leader sponsorship, and we are very fortunate to have the support of Alex in that as well.

In terms of SCS upskilling, that is, of course, a very ambitious target, as you rightly point out.

Sir Geoffrey Clifton-Brown: You set it.

Megan Lee Devlin: Yes, we did, because we think it is extremely important. We have already upskilled 8% of SCS in digital, data and technology essentials against the standard we published last year, and we have a target to reach 15% by this September. We have the glide path to reach the 90% target, but it will, of course, take sustained effort and focus from our senior leaders across Government.

Q37            Sir Geoffrey Clifton-Brown: Can I take you to paragraph 3.5 on page 29? It says that half the services have “no single senior individual who has ownership of the full service which encompasses both online citizen-facing portals and the back office systems that perform the other processing. Many different programme directors, operational leaders and digital leaders are involved in this service delivery. It goes on to say, “This is a major impediment to service redesign. A single service director-level owner for each service, with accountability for its end-to-end operation and decision-making authority for continually improving the service, would enable Departments to form a complete and joined-up view of their services. It is saying, I think, that there is too much churn of people. What can be done about that?

Megan Lee Devlin: This is a concept that is very much based on best practice in the private sector, which is product ownership or service ownership. The key difference between that and the organisational models we use at the moment in Government is end-to-end visibility and accountability for the performance of that service. As the NAO Report highlights, there are a number of handovers at the moment between teams, whether they be policy teams, service operations teams, delivery teams, transformation teams or digital teams. The single service owner construct seeks to tie those together so that we can better enable cross-functional working.

Alex Chisholm: Just to give a specific example, which might be helpful to the Committee, you have heard over the years about universal credit, which does take that view. That is the way that it is organised. The SRO programme directors for that are able to reach across all parts of DWP to help deliver the overall solution. As a result, you can ask them, as I do, “What is your cost to serve? How many people have switched from this form of benefit to the new one? How did you deal with this surge when you changed those things? They always have all that data at their fingertips and can give you forecasts for what is next going to drive the improvement to that, because they have that single service philosophy.

We know that we can do it in Government. It is just that, at the moment, it is relatively rare. The NAO Report, which you have just referred to, rightly says that it would make a big change to have that accountability for single service owners. That is a key recommendation in the Report, which we take very seriously.

Chair: It sounds like not only would it be useful for digital purposes, but it is important generally.

Q38            Sir Geoffrey Clifton-Brown: My second theme is on numbers and buy-in by senior people in each Department, which I still have questions on. I take you to paragraph 3.16 on page 31, which says, “Despite some early steps to improve Governments data, CDDOs latest progress assessment is that significant technical and cultural barriers remain. It highlighted a risk to the digital and data board in December 2022 that sustained effort may fail to overcome the barriers to standardising definitions and sharing critical data assets. Our experience suggests that Government is starting from a low base given the condition of its existing data, and the scale of the task is huge based on its own analysis.

That paragraph seems to indicate that there is a problem with all permanent secretaries, and all senior directors and other senior individuals in each Department, buying in. I ask again, Mr Chisholm, whether it is variable between Departments. Are they all buying in, in the way that they should?

Alex Chisholm: I would interpret that differently. I have a long experience of dealing with these challenges. I do not think that there is any resistance at the top. People at the top say, “Why should we not be sharing data? It is going to save us money. It is going to deal with fraud. It is going to improve services, etc. Why is it, then, that we find it difficult to do? Sometimes it is for genuine legal or regulatory reasons. Sometimes it is due to perceived legal or regulatory barriers. Sometimes it is because the expertise is not available to be able to put in place the necessary connections between data.

Other times—and this is a more subtle one that has not come up enoughit just is not a priority for the person to do the piece of work that benefits another organisation. Who has valuable datasets? HMRC and DWP are examples. It might be that another public service delivery organisation would love to know the piece of data that they have, but they have their own stack of priority projects to do.

That is why we have been trying to create this data marketplace-type idea, and to have some expertise in putting in place these APIs in order to make it easier to address those issues for the people who control the data, in a safe and respectful way, of course, with all the legal and regulatory protections.

On the one hand, you could say that we have made incredible progress. One of the recommendations from a previous report was that we should put in place a catalogue of the most useful data to hold. That catalogue now has 230 APIs—application programming interfaces—available. You might ask whether that is a large or a small number. That is supporting billions of transactions every month.

Again, to give a specific example of that, because it helps to bring it alive, if you go into a pharmacy for a prescription, one of the things that they need to know is whether you have to pay for it or whether you are entitled to get it free. That is now available as an API look-up in pharmacies to the NHS data in order to tell you what your status is. It just gives you an idea of how, instead of looking at an out-of-date list that was given to you, sending an email or making a phone call, that is done automatically through that API-type interface. That is great.

Q39            Chair: Is that universally rolled out? We have had problems when we have looked at NHS charging.

Alex Chisholm: I cannot tell you whether it is universal, but it is certainly in place.

Q40            Chair: It exists but may not be being used by—

Alex Chisholm: I do not know whether it is universal.

Megan Lee Devlin: It is used over a half a million times a month, so it is certainly used quite extensively by pharmacists.

Q41            Chair: Half a million is a very small percentage of the number of people. It sounded very positive. It is a good step.

Alex Chisholm: It is very positive.

Chair: But it is not yet universal.

Alex Chisholm: Not everyone is using it yet, so that would be better. I was just giving that as an eye-opener to the world of opportunity that comes. So much time is wasted in the administrative processing of public services, which is just not rewarding for anyone to be involved in.

In order to accelerate this, we have to, in some cases, fund the digitalisation of the data. I mentioned, in spending round 2021, the £1 billion that was on data. A lot of that is to put data into digital form, because lots of public records were not digitised, and then make it easier for people to share legally through the Digital Economy Act and through regulatory easements where necessary, as well as training. There is also being able to make it a priority to do things of wider public value, which is where some of these transformation exercises say, “If you had this piece of information, would you be able to provide a much better service to the end customer?

Q42            Sir Geoffrey Clifton-Brown: We have the message. We are in danger of running out of time. I have a piece of paper in front of me that is a standard appointment letter for a second permanent secretary. It mentions all sorts of things about managing public money and everything else. Should one of the standard paragraphs in this type of letter be that any permanent secretary, on their appointment, has to adhere to certain digital improvements in their Department?

Alex Chisholm: That is an interesting possibility.

Q43            Chair: That is something to take away and think about. Who would make the decision on thatyou?

Alex Chisholm: The appointing Secretary of State and the head of the civil service would have views on that, but I take your point. Especially for those Departments that very much offer public services, that could well be a relevant factor.

Sir Geoffrey Clifton-Brown: Thank you for that. It is rare to hear that from a permanent secretary. Usually what happens is that they say, “No, no, no, and, a few months later, they do what we suggest, so thank you for that answer. I have one final question, but I have been put off my stride now.

Chair: How to knock out the Public Accounts Committee: flatter a member.

Q44            Sir Geoffrey Clifton-Brown: I take you to paragraph 3.23 on page 33, which says, “CDDO has asked the digital and data board to promote digital training for all senior civil servants. This comes back to the question that I had earlier. Digital, data and technology essentials is a recent course for non-digital senior civil servants aimed at educating more than 90% of senior civil servants over time. This training is absolutely critical. Ms Lee, perhaps you would just address that. You did answer the question before, and I accept that, but is that a realistic target and, if so, how long will it take to get there?

Megan Lee Devlin: Is the target realistic? Absolutely, it is achievable, but it is a stretch target. As I mentioned earlier, the roadmap commitments were deliberately ambitious. We have just completed our pilot programme of the digital excellence programme that you alluded to, with 300 civil senior civil servants who undertook that training over the course of five weeks or so. It was very in-depth and designed to reach the standard that we set out last year, which covers topics such as legacy risk, end-to-end service design, user-centred Government and so forth.

We have then set ourselves a series of targets over the next two years as a glide path to reaching the target of 90%, the first of which is a 15% target for September this year. It is ambitious and it will take sustained focus, but we will be reporting regularly to the digital and data board on progress and working closely with Departments through the quarterly business review process to measure and monitor progress over time.

Q45            Sir Geoffrey Clifton-Brown: I have one other question for you, please, Mr Chisholm. It is about this engagement of senior people in Departments on the digital and data board and, indeed, on the new Government transformation steering group. Do permanent secretaries sit on those two committees and do they attend?

Alex Chisholm: The answer is not every one, and that is the right answer because otherwise you would have a committee of 40 or so people. We do have regular attendance from the permanent secretaries for the Home Office, DWP and HMRC, which are the big delivery Departments.

Paul Willmott: There is DEFRA as well. We have also included the second permanent secretary from the MoD on the digital and data board.

Q46            Sir Geoffrey Clifton-Brown: To my second point, do they all attend?

Paul Willmott: Not all of them attend every meeting, but in general they do attend.

Q47            Sir Geoffrey Clifton-Brown: I have heard that quite a lot of them are sending deputies.

Paul Willmott: That does happen.

Q48            Sir Geoffrey Clifton-Brown: Is that true that a warning went out that they were expected to attend?

Paul Willmott: We have certainly reminded their offices that they are expected to attend in person, yes.

Q49            Sir Geoffrey Clifton-Brown: You will be maintaining that in the future.

Paul Willmott: Yes, absolutely.

Chair: Too often on this Committee we have seen cases where people do not attend or they send deputies. Having a committee itself is not, on its own, delivering anything, with no disrespect to committees,

Q50            Mr Djanogly: Good afternoon. A point came out of Sir Geoffreys questions that I was interested to know a little more about. Sir Geoffrey was talking about the training of executives, and Ms Lee was talking about having a glide path for management training. Sir Geoffrey did mention some figures, but they were not really addressed.

Paragraph 26 of the NAO Report says, “To date, only around 200 non-specialist executives have received digital awareness training against the target of 6,500 set in the roadmap.” Ms Lee, could you explain why that is the case?

Megan Lee Devlin: The digital excellence programme, which is the programme we are delivering to meet the standard we expect of senior leaders when it comes to digital, data and technology essentials, has just completed its pilot phase. That programme is under development. It has been built in partnership with the LSE and Apolitical. It is a really strong programme that we have received very positive feedback on. Now that the programme is live, we expect to accelerate the number of senior civil servants who are taking up that training that is available for them to take.

Q51            Mr Djanogly: You are saying that you are not diverting from the target you had set.

Megan Lee Devlin: No, we are not diverting from the target.

Q52            Mr Djanogly: Okay, I get it, thank you. Departments will often have costings on new digital projects, but I just want to talk a little more about the transforming of legacy services. At paragraph 1.5, the NAO Report says, “Government has poor data on the full costs of its existing services. It estimated that nearly half of all technology expenditure across Government in 2019 was dedicated to keeping the lights on activity on outdated legacy systems.

The first question, Ms Lee, is perhaps for you. Why do some Departments seem to resist measuring their costs?

Megan Lee Devlin: We have quite a lot of visibility in terms of digital, data and technology costs across Government Departments. We measure that through the digital dashboard, which is the core artefact used for discussions at the quarterly business reviews. That looks at spend across Government Departments in terms of their technology estate, their digital services, their data programmes and so on.

In terms of looking at services end to end, to which you are alluding, that is a core part of the top 75 service performance analysis that we are undertaking. We look at the costs involved from the very start of a user journey with Government through to the point at which they receive an outcome. We add up all those costs to understand cost per transaction, which we measure over time to assess performance.

It is challenging to assess the full extent of a technology estate because technology can be created in different parts of the organisation, but we also look at that through technology maturity assessments as part of our quarterly business review to ensure that Departments have the right sight, across their estate, of new solutions, where they are and whether they align to their technical architectures.

Q53            Mr Djanogly: I am still unsure as to whether we are talking here about Departments resisting measurement or being unable to measure.

Megan Lee Devlin: Is that specifically with respect to their services or their technology estate?

Mr Djanogly: I mean the technology estate.

Megan Lee Devlin: Most Departments have a good assessment of their total technology spend. You may be alluding to their spend on legacy technology, which is a core theme in the NAO Report. Legacy is a very difficult thing to measure. Indeed, not all legacy is bad because many older systems do serve us quite well in terms of operational delivery and are fit for purpose and delivering what they need to.

However, we have taken a risk-based approach to assessing legacy across Government. So far, 16 Departments have adopted that legacy framework. That enables them to look across their estate and identify the highest-risk legacy systems, which we can then monitor over time.

Q54            Mr Djanogly: You are just looking at the highest-risk legacy systems. If you do not know what the costs are, how can you manage those costs or plan ahead?

Megan Lee Devlin: Your point is absolutely spot on. We need to have visibility of cost to be able to gauge performance over time. Through our quarterly business reviews, we are making sure that Departments are measuring their technology estates and they have sight of their highest-risk legacy systems so we can do that.

Alex Chisholm: If I could just add a couple of words to that, first of all, that assessment was done as part of Organising for Digital Delivery. At the time, as far as I recall, we thought that something like £2.6 billion out of the £4.3 billion of annual spend went on this maintenance activity. We now have much better visibility through the quarterly business review process my colleague was just describing.

When I look across Government, I see quite big variations. If you were talking to DWP, for example, as I have done with our executive committee, in many cases it can tell you its cost to serve different customer categories for different services. In some other Departments, that would not be the case. They do not have that much expertise and visibility at the top level. So there are different levels of maturity. Part of what we are trying to do here is bring everyone up to the standard of the best.

The second factor to bring out is the point we made before: if you do not have a single service owner, it is very hard to be able to answer those questions about what your actual cost to serve is.

The third is just a technical point. If the Committee is trying to measure over time whether we get better or worse at this—I am sorry to introduce financial accounting here—there is this issue about capital expenditure versus resource expenditure. It is brought out very nicely in the NAO Report, but it is an important point. As we reduce the amount of new capital expenditure and more of that is expensed, as it were, as resource expenditure—you pay as you go to access a cloud system—the actual amount of “keeping the lights on” activity spend might be going up and that might not be a bad thing.

Chair: Yes, we are very aware of that, but nevertheless it is important to get it on the record.

Q55            Mr Djanogly: You are talking about bringing people up to a certain level. Ms Lee was talking about certain levers you are using to try to get to that. Do you have adequate levers to do what you need to do?

Alex Chisholm: One always wants more levers. We have been trying to work with the strategy the NAO has recommended. It does not use this language, but it is about having both a carrot and a sticknot in the sense that the Cabinet Office is telling other Departments what to do, but it is about recognising that most Departments want to deliver great public services and the people who work absolutely have that mindset.

How can we support and facilitate that activity? We can train people up in good practices; we can help them to access valuable resources; we can fund, between ourselves and the Treasury, the best programmes that need to be done. Where people are going off track, we can use this quarterly assessment to say what steps might be taken to get back on track. We can apply the common rigour that is necessary around data standards, the single legacy framework and the pay framework.

This is about working as an enlightened digital centre with Departments to make sure that their own digital activities are as good as they could be. That is the philosophy of partnership and collaboration that has been recommended by the NAO and PAC on previous occasions, which we entirely embrace.

Q56            Mr Djanogly: Mr Willmott, are there any lessons coming out of the private sector that we could or should be applying?

Paul Willmott: Yes, absolutely. In the private sector we would expect to see consistent, reliable and timely metrics reported both on end-to-end services and on the systems underlying the technology. To some extent, we would consider that foundational. If you do not have good numbers on your starting point, it is very hard to make a business case for investment and to track the improvements you are making.

As for the metrics, you would expect to see unit costs: “What is the cost of applying for a passport?” You would also expect to see service levels: “What is the average time taken? What is the average error rate?” and so on. We have that for some services in Government, but for others the data is not of the quality I would like to see because it does not provide that baseline for improvement.

Q57            Mr Djanogly: Do you take advice from the private sector on this sort of stuff?

Paul Willmott: Yes. Part of the reason why I was lured into this role by Mr Chisholm was exactly to build that bridge to the private sector. For example, I convene a digital advisory board where we have a distinguished group who between them have many decades of experience of this activity in banks, retailers, utilities firms and so on.

Q58            Mr Djanogly: Is that helpful?

Paul Willmott: I think so. Ms Lee might like to comment, because she and her team experience that on a day-to-day basis. We are trying to get the advisers not just to sit around a table, but to work with some of the teams in the Departments. I would say we learn something new every day from the private sector.

Q59            Chair: When you say “work with some of the teams in the Departments, do you mean as non-executive directors or just to pop in for a chat with them? How does that work?

Paul Willmott: There are also non-executive directors within the Departments we are connected to. For example, I attended the non-executive director day a couple of weeks ago. There is a strong digital capability in that group: in fact, nearly half claim some digital experience.

The digital advisory board, which I convene, works in a hands-on way with the CDDO team and with some working teams in Departments. The delivery of the expertise to the real point of need is one way of moving us forward well.

Q60            Chair: Are they remunerated for that?

Paul Willmott: No. They all do it out of the goodness of their heart.

Q61            Mr Djanogly: I am looking at paragraph 1.7 in the NAO Report, which is looking at the sorts of problems. It says, “Our assessment of Governments operational management capability showed low levels of maturity, suggesting that Government lacks an understanding of why an end-to-end view is necessary.

Ms Lee, how do you support Departments to understand and appoint a single service owner of suitable experience and seniority to be responsible for the end-to-end running of a major service?

Megan Lee Devlin: In terms of understanding, we have spoken at length about the digital excellence programme, which is about ensuring that all senior civil servants have the right skills and experience to be able to lead digital teams.

In terms of your more specific point around single service owners, in CDDO we set the strategy, the policy, the standards, the guidance and the frameworks. We are very much there to support and enable Departments to meet those standards. It will be no different with a single service owner. We are working to define and iterate that standard at the moment to ensure that it is fit for the Government context and purpose, and that it applies in different situations.

Different services look quite different. They might be transactional; they might be informational. Therefore, it needs to work for a number of purposes. We will then be rolling that out with a specific focus on the top 75 and enabling the teams to meet that standard.

Alex Chisholm: We are always in a multi-lever environment. One way in which we can encourage that to happen is through the top 75 programme and our review of that, which has been endorsed by Ministers and Departments. It is also a recommendation of this Report. That gives us an additional point that we will need to respond to formally. Indeed, you can hold us to account for that, which is good.

More valuable than even either of those things is for Departments to understand that this is the best way to do things. That is an educational experience type of thing. You mentioned one of those bodies that we have created, the transformation forum, which is to try to enable the people who have done this to show and explain to the others why it is a better way of doing things. We are pulling on all the levers we can.

Q62            Mr Djanogly: We touched before on capital investment in legacy. I am just looking here at paragraph 25 of the NAO Report. It says, “Departments have also told us that it is easier to bid for capital funding for new developments than resource funding to maintain existing services and keep them up to date.” How is that dealt with? Is that right, to start with, Mr Chisholm?

Alex Chisholm: It is a fair observation. I suspect the Committee probably has quite a few reminders of that in your hearings from week to week. In some ways, it tells us something about human nature that we are more excited by the brand-new thing than dealing with the past.

What have we done to try to address this? Again, I would really emphasise the importance of the single framework for looking at legacy. We have now identified the top 20 risks, and funded and programmatised—apologies for that word—remediation plans for those. We wrote to you in December with those, and you should ask us to write again in December to say whether those are all done.

His Majestys Courts and Tribunals Service, for example, has so far decommissioned six of its legacy systems, and it has seven more due in this financial year. It is very important to hold people to that and make sure they use resources to deal with these things. Otherwise, you accumulate unacceptable risks and poor performance. It is very hard to create transformational systems if there are poor old legacy systems behind that. That is a big part of it.

There is another area where we have worked incredibly hard. Again, this was embedded in SR21 for the first time, and we need to do it in all of our spending round processes in future. People should be able to say not only, “This is the cost of this particular project, this is how long it will take, and these are the risks I am managing, but “These are the benefits.

We went through 150 or so—I cannot remember the total—business plans and propositions for that. The best were outstanding; they were able to say exactly what the return on investment was, when you would get it, and how that would reduce their operating costs going forward. The worst had no such figures. They were not able to describe the benefit realisation from their programmes due to a lack of expertise and unfamiliarity. We helped to ameliorate that the best we could in the time available.

That is the other thing that I would suggest the Committee focuses on. People will under-invest if they do not see the rewards. Being very clear on the rewards and holding people to those benefits is the best way to ensure we invest at the efficient level.

Q63            Mr Djanogly: Have we got to the stage where we have a date when we will be able to get a full picture of the extent of the legacy systems across Government, the level of risk this poses and the likely costs and timescales for addressing all of this? Is there a map?

Megan Lee Devlin: We are due to roll out the legacy risk framework for all Government Departments by the end of this financial year, March 2024.

Alex Chisholm: There are 16 Departments already using it, and 153 legacy assets have been charged already. That is the basis on which we extracted the top 20, which we have never had before. It is a big leap forward. Some of these are very old systems with a very high level of risk associated with them.

Q64            Mr Djanogly: Would year-end be an appropriate time for you to give us an update on this?

Alex Chisholm: Yes, or possibly just after year-end. One year on is July, and two years on is next July. This is a point of detail for your scheduling.

Q65            Mr Djanogly: Finally, how can Government support senior leaders to understand the difference between tweaking legacy systems and the fundamental redesign of their services?

Alex Chisholm: Again, there are a number of zingers in the Report, with which I totally agree. It is partly a point about skills and experience. The way to look at it best—hence the top 75 programme—is to say, “Dont just put a nice new front end on five existing, very poorly performing systems and assume that that will look good and be good.” It will look good, but it will not be good. It will be maddening for people, because they will not understand why they are getting such poor data results from it.

Business discipline—it used to be called business process re-engineering, and it is now called various things by consultancies—is very important. This is about looking, from a pure process control perspective, at what you are trying to achieve and the different steps within that process. At the moment, when you apply that to a lot of our things, you find a massive level of complexity and huge numbers of exception powers, policy overrides, legal checks and lots of different people. It is no surprise that what you get at the end of that is often too little, too late.

If you take the personal independence payment—I know you have looked at that recently—it required a pretty fundamental look at the processes that were being followed in the health assessments. When you build wholly new ones like Future Farming, it is a chance to build in really good process practice from the beginning. The transformation comes in the whole of the process. In a way, the digitalisation is the easier bit of it.

Q66            Nick Smith: Mr Willmott, in recent months we have had some really interesting sessions in which we have found out about the DVLA and the Passport Office. These places have large data centres and much complexity, with industrial relations negotiations to achieve the necessary systemic changes. How good are the Departments at negotiating the business re-engineering that Mr Chisholm has just talked about in those places to get a better service for customers?

Paul Willmott: This is a really important thread to pull on, and I am glad that you have returned to it. The NAO has returned to it in multiple reports. As Mr Chisholm said, it is about having discipline around business process re-engineering, but somehow it is that-plus. You need people who have skills in technology, skills in understanding end-to-end process and the ability and willingness to look at organisational structures. If you have a single process that dips through 10 organisations, there is always friction.

It is hard to find people with the right capability and credibility to make the level of change required, but this is the level of change that we need to make if we are serious about transforming services. We are looking at how to make that a little easier by de-risking it. For example, having stage-gated procurement processes allows what is called a sandbox approach, where you can trial a new process in a limited scope. You might do it in one region before deploying it more broadly, for example. That is one way.

Again, this is part of what we are doing with the digital academy. We are training people on how to capture these opportunities.

Alex Chisholm: If I can add one additional point there, from a whole-system perspective, you are very familiar with the role of the IPA and the major projects review group, which I chair. All major projects go on to the Government major projects portfolio.

Within the IPA, the way in which it assesses digital projects is a bit different from other projects. One of its conditions is that it has somebody who is expert in so-called agile delivery as one of the assessors for all digital projects. Again, it is just a way in which we are trying to make sure this is systematic thinking that is embedded every time. That is the process you go through in the gateway review to get your project approved.

Q67            Chair: I just want to cover one quick point, which we have skirted around, about the culture of the different organisations. When we looked at the Rural Payments Agency, I remember there was an extraordinary moment in the room where the digital people and the people in DEFRA were clearly in different camps. One of them said, “We work on different floors, and we wear different types of clothing.” There was absolutely no cultural collision, really. Well, there was a cultural collision between them—in fact, it was just that.

How does it feel now, Ms Lee? Do you feel you are welcomed into Departments, or do they think, “Oh no, here come the digital people, trying to tell us what to do”? How do you overcome those cultural barriers? Let us face it: you have a civil service that is full of policy generalists. It is specialising more and more, more than it was 10 years ago when I was first on this Committee, but it is still quite a generalist system.

Megan Lee Devlin: I have been in Government for two years now, and even in that time I have seen really significant improvements in digital engagement at the most senior levels of Government. We have a huge support from the permanent secretary community, and many of the permanent secretaries state that digital transformation is one of their key focus areas. That is hugely encouraging.

Q68            Chair: There must be some barriers that still remain, though. What are they? How are you overcoming them?

Megan Lee Devlin: Culture can be a bit of a nebulous construct. Rather than speak about that, I would instead point to the capabilities and processes that we use in Government and how we are looking at ensuring they enable digital teams as effectively as they can.

We have spoken about capability, particularly focusing on SCS. On the processes, mission 6 of the roadmap looks at system reform. That is about partnering with other functions to ensure the way we work in Government is appropriate for digital delivery. That includes working closely with the commercial function and working closely with the IPA to ensure its guidance is continuously enabling digital teams, as well as working with the Treasury on challenges such as the RDEL/CDEL challenge that was mentioned earlier. Indeed, the Treasury recently updated its consolidated budgetary guidance to reflect a willingness to switch between CDEL and RDEL in certain situations.

We are absolutely working on some of these broader process challenges that we hope to address effectively by 2025.

Q69            Chair: You have anticipated my question about the RDEL capital spending issue. One of the reasons it is easier to bid for capital budgets, in broad terms, is that it is one lump of money that is spent, and that is it. That is putting it very simplistically, obviously. There is a risk. Governments of all colours come in and then do the headcount thing and the salami-slice savings. I understand that this is not entirely down to you; it is often down to Treasury and, indeed, ultimately to Ministers.

When we improve these big systems, they have to run, and run well. A little bit of a cut one year can have a devastating impact and create the same problems we are trying to deal with now. What levers are you pulling to try to make sure that does not happen?

Megan Lee Devlin: There are three levers, which the NAO Report suggested were much stronger than previous strategies that were in place. The first is our governance forum. We have spoken at length about the digital and data board at the permanent secretary level. We also have a functional leaders group, which is the chief digital and information officers across Government.

We also have delivery groups of the most senior technologists in different fields—chief technology officers, chief data officers and so forth—that are working very hard on some of these difficult challenges. Ultimately, any issues in large programmes tend to come through those governance structures.

Secondly, we have the digital dashboard, which I spoke about earlier. That gives us quarterly sight of the progress against the roadmap commitments so we can keep abreast of developments and any challenges Departments raise in their quarterly business reviews. As Mr Chisholm pointed out earlier, those are very much two-way conversations. We ensure we are offering the support and enablement we can from the centre as well.

The third lever is really our processes, which have both that quarterly business review element and the spending controls, which are delegated authorities from Treasury. If a programme does not meet the bar, or it does not meet our red lines, which have been aligned to the digital and data roadmap, we can use that harder lever.

Having said that, we have had fantastic collaboration from colleagues right across Government. We are therefore very positive in terms of developments over the last year and progress against the roadmap.

Q70            Chair: Mr Chisholm, if the spending taps are cut off or there is a salami-slice percentage cut, there is still a risk to programmes like this.

Alex Chisholm: I generally feel that the Treasury shares the view we have that this is incredibly important. It partners with this programme of funding transformation, legacy remediation and data.

Q71            Chair: Is it effectively ringfenced, then?

Alex Chisholm: The legacy remediation is, yes. We have come a long way, which is great. I have a couple of things to add.

Painful as it has been, dealing with covid has greatly improved Governments understanding of how important it is to use data to target services and analyse the nature of problems. It used to be that not every meeting started with “What does the data show?”, but that is very often the case now.

Indeed, thinking about how we put in place key services, you might remember the one for vulnerable persons who were trying to get food parcels. That involved getting data from five NHS systems, building a new platform on the Government Digital Service and then using that to support people to provide local support, food parcels, etc. All of that had to be integrated with the private sector, the local authorities and volunteer groups.

That was a digital and data challenge that was put in place. It took about eight weeks. When you think about where Government used to be, that would have taken eight months and it would not have got anywhere near the coverage it did. We have got a lot better through that.

In order to get even better, a lot of the elements that we need to do are embedded in this Report from the NAO, in particular encouraging people—of course, the PAC can help with this—to pay attention to the end user. What is the end users experience?

Chair: We have some thoughts about how we might encourage good behaviour.

Alex Chisholm: What is more, as a follow-up question, what is your cost to serve the end user? If you are met by blank faces, I suggest you ask again.

Q72            Chair: We never mind sensible thoughts about how we might focus our questioning. From what you have said, do you have enough disincentives to look for quick wins? It is easy to say, “We have a parliamentary cycle.” There are all sorts of fast-cycle things that happen in Government, where a permanent secretary might want to prove their Department has achieved X or Y, but those quick wins may not be good for the longer-term nature of the project. Is what you and Ms Lee have described enough to stop people trying to game it?

Alex Chisholm: It is going to be much harder to game it, if anyone were tempted to do so. On things like legacy remediation, you have a common framework, committed funding for specific projects and specific milestones, and a tough assurance process. It is not only our good selves but this Committee as well. That is why we have made our commitments to you. I think it will greatly improve, I really do.

It is harder to see the missed opportunity around the future transformation. You might be going at a particular level where everyone says, “That is great,” but maybe there was a chance to do even better and you could not really see it. That is an area where Government will have to continue to get better.

Q73            Chair: The danger, as we touched on right at the beginning, is that a reduction in funding just means going slower. Going very slow could mean that by the time you achieve the first goal, it is already out of date. How are you guarding against that?

Alex Chisholm: If you look at the overall pressures on Government spending, they are going to require us to get much better at the delivery of digital services. That is the only way to make the numbers add up.

Q74            Chair: On the data, which we have touched on many times, it is a really big challenge. When we looked at this in one Department—this was probably a number of years ago now—there were 13 ways of recording the country you live in, using different data systems, because people put it in differently.

When the Home Office did the Magee review back in 2008 or 2009, there was an organogram of what data was held by the various bits of the criminal justice system and how it was recorded. It was not just that it did not map exactly; it was way off. People were having to fill in myriad forms asking the same question but in different ways.

That is a mammoth task. We have skated over data a bit, I feel. We need to codify that and deal with the fact we have all this old data that was collected in different ways. All of us around this table will have data with HMRC that is different from data with DWP, which will be different from the data with other bits of the system. How do you go about getting that in line?

Alex Chisholm: As you imply in your question, it is partly a standardisation challenge. We have the Data Standards Authority to do that. The only slightly more optimistic way to look at it is to compare it with when we tried to deal with those problems 10 or 20 years ago.

The vast improvements in technology and so on have made it much easier to get over imperfections in data and the way it is recorded, and to make intelligent connections between differently formatted datasets.

Chair: Mr Willmott, you are nodding.

Paul Willmott: Yes, I am nodding for two reasons. First, you are absolutely right that it is a huge barrier. We are already moving towards common standards, but cleaning up past datasets is a substantial task. Secondly, I am nodding to Mr Chisholm because the technology to do this has vastly improved.

The One Login programme will help because it will provide a common key to which other data can be attached. Over time, subject to opt-in requirements, privacy restrictions and so on, that should provide a much better view of the citizen.

Q75            Chair: Are there any legal issues, Mr Chisholm, about putting data together? I have the scars on my back from identity cards.

Alex Chisholm: Yes, absolutely. There are all the requirements of the UKs GDPR regime. When we make changes in this area, we do so in a very open and consultative manner. We get advice from experts and we do public consultations. Indeed, we have just done one on digital identity. We had many thousands of responses.

Q76            Chair: Is there anything you want to add, Ms Lee?

Megan Lee Devlin: The scale of the challenge here is quite enormous, and I really welcome this Committees focus on data in particular.

Chair: I have been on the Committee for over a decade, and we talk about data at every meeting. Other colleagues have been around a while longer. We seem to have made little dent in it over a decade.

Megan Lee Devlin: It is very valuable. CDDO is thinking about this from two perspectives. First, we are thinking about near-term interventions and practical support we can offer Departments to enable them to share data more effectively. We are developing the data marketplace, which is a part of the commitments under mission 3 in the roadmap. That will improve the discoverability of data for colleagues across Departments and help us to align around a single ownership model.

We are also doubling down on our critical data assets, coming back to the recommendation of previous reports to focus on the highest-value data assets across Government. We have a really good community of chief data officers across Government, who are currently in the process of defining those.

Q77            Chair: What are the most important data assets in Government?

Megan Lee Devlin: They are the data assets that underpin a lot of operational delivery services, for instance those that help with eligibility checks.

Chair: So HMRC and DWP would come in.

Megan Lee Devlin: Yes, of course.

Alex Chisholm: It is personal dataaddress data, property data and number of children. These are the things you get over and over again.

Megan Lee Devlin: We are rolling out some practical near-term interventions, and then there is the longer-term challenge of broader data maturity across Government. We recently launched our data maturity index for Government Departments. All Departments have signed up to a commitment to roll that out in part of their organisation in the next 12 months. That was piloted quite extensively and is a really robust assessment of how each Department compares with best practice and, therefore, where they need to focus their efforts to improve their data maturity.

Q78            Chair: You are not getting any pushback from the Departments, where they are not anywhere near best practice and they are not willing to acknowledge it?

Megan Lee Devlin: No. Our digital and data board agreed to roll it out across all Departments in our meeting last month.

Q79            Chair: We need honesty on this, or we are never going to get results.

Finally from me, we have not talked about the NHS. You are dealing with Whitehall Government Departments, but there is an awful lot of talk about what is going on in the NHS. Any one of us who uses anything from a telephone system to a GP service or any out-patient system knows that there is an awful lot of opportunity for digital change in the NHS.

I do not know who would be answering that, but, Mr Willmott, you chair the CDDO. Are you having any conversations with the NHS about the learning you are having in Whitehall? What could you be doing to share that with the NHS?

Paul Willmott: Personally, I have not been involved in any conversations with the NHS, at least not for the last couple of years. There is an opportunity to share. I understand that the NHS is open to that.

Chair: You need more resources, I suppose.

Paul Willmott: Yes.

Megan Lee Devlin: We provide support and guidance to the NHS, and we are actively doing so at the moment around our DDAT pay provisions. We are supporting them as they think through their strategy in terms of people and digital skills.

We are incorporating the NHS into our quarterly business reviews from the next round of reviews, which will start next week, and therefore we will have greater sight on what is happening there. We certainly work closely with colleagues there from a collaborative and communities-of-practice perspective. A number of colleagues in the broader health ecosystem join our governance forums and our communities of practice. We absolutely do share knowledge and skills.

Chair: They are baby steps, but there is some conversation.

Alex Chisholm: There is huge potential. Just to manage the Committees expectations, we have about 200 people working in the Central Digital and Data Office. We are minnows compared with the NHS, which, as you know, is over twice the size of the civil service. It is a massive area of opportunity with a lot of specific requirements and tens of thousands of people working in digital domains. We offer a little bit of assistance from the side, but it is not the prime focus of our work, just to be absolutely clear.

Chair: No, we are aware that you cannot take everything on. Maybe it is an opportunity to expand your empire. I will leave you to discuss that with the Treasury.

Q80            Sir Geoffrey Clifton-Brown: Mr Chisholm, this Committee believes wholeheartedly in transparency. It is what we do in these hearings. We shine a light on an aspect of Government work and we ask for that information to be followed up. We now, because of our persistence, have a twice-yearly report on HS2.

I would like to see whether we could get a follow-up to this hearing, on a regular basis, to Parliament. I am just wondering whether the mechanism is not contained in paragraph 2.7, which says, “CDDO provides six-monthly progress reports on the roadmap, starting in December 2022, which include additional actions for the boards endorsement that are needed to maintain progress.” Could we not publish those every six months?

Alex Chisholm: Quite possibly, yes. We certainly wrote to you in December with our first six-month one. We are doing another one for the one year on. I am not giving you an absolutely categorical answer, only because I am respectful of the decisions of Ministers who are not present.

Chair: We get the point. We all have an interestthe taxpayer, this Committee, Parliament and every bit of Governmentin seeing what is happening and shining a light on it. As Sir Geoffrey says, that is part of what we are all here to do.

Thank you very much indeed for your time. This is a fascinating issue that is vital to citizen service and good value for the taxpayer. It is a very long project, enough to see us all through to retirement and beyond.

The transcript of this session will be available on the website uncorrected in the next couple of days, and we will be producing a report, we hope, before the summer recess. Thank you very much.