Good morning colleagues. This is round two, ding, ding. Welcome back to our witnesses, and let me just put on the record my thanks on behalf of the Committee for making yourselves available for the reopening of session one, as it were. If I may, let me start by just letting you know it is quite unusual for any Committee hearing to start receiving rave notices, if you will, from interested parties. But I have received—through private communication—a large number of correspondences from members of the ONS team and staff. In various ways, they have all amplified and echoed, if you will, the concerns that this Committee was discussing with you last week. It is good that they have felt able to comment through to the Committee via me; of itself that is encouraging. But it is worrying that lots of people across the organisation are saying yes, there were and still are problems: the HR function is wobbly, complaints and suggestions disappear into the ether, and the organisation is not being run professionally. From what we heard last week, that is on the radar of the three of you, but the troops on the ground are reporting in, as it were, and I am told it is the first time a meeting of PACAC has effectively gone viral. I am not suggesting that there are counterfeit copies of the tape being handed round in the cafeteria or whatever, but as you will appreciate, we were worried at the end of the last session, and those worries have continued. I want to pick up in true broadcast fashion, as it were. Viewers who tuned in last week stated that you, Mr Humpherson, were effectively ventriloquising your thoughts or expressions well, so the Committee wants to turn to you first, if we may. Can you add some flesh to those ventriloquised bones that I set out last week?

Yes. When you asked me that question at the end of last week’s session, I had a few things running through my mind. I was thinking that it is the right thing for ONS to have these economic statistics and survey recovery plans, but I did not want anybody to be in any doubt that there is an awful lot to deliver and do. Robert and Emma would probably back me up on that. I also suppose that I am here, in my role as head of the Office for Statistics Regulation, to care about the integrity of evidence in an environment of misinformation. The issues you explored as a Committee probably have the risk of casting a reputational shadow across statistics. In a way, what was going through my mind was that there is that reputational hit, but that disguises the fact that lots of producers of statistics in the UK do really good work. The ONS also does really good work on a whole host of things like census, prices, migration and so on. That was in response to you channelling that anxiety I felt about the reputation. Emma spoke about the culture of internal openness to challenge and the measures she is putting in place to address that. I spoke about ONS’s willingness to take on board, respect and respond to external challenge, and on both those I am probably seeing change. On the points that I made, there is quite a lot of evidence over the last year that ONS is becoming more open, but there is a long way to go; cultures take time to shift. So if you put all that together, what I would have said if I had given a longer answer would have been that there is a lot to do.

That is helpful, and I am grateful to you for that. Before the agenda for change—or the realisation that there needed to be an agenda for change—germinated in a mind somewhere, can you just give us a flavour with regards to your assessment of how it got to where it arrived, operationally and therefore reputationally? Had you and your immediate colleagues flagged and suggested, and how were those flags or suggestions—if indeed made—dealt with and responded to?

Can I take the second part first, on the flags? We did raise flags. We report into a committee of the board called the regulation committee. Sir Robert is on that committee, and it is chaired by another member of the board. They were extremely receptive to the points that we were raising. If anything, they prompted us as OSR—the oversight body—to pursue our points more firmly. I was always aware that, through the committee, the board was taking these findings and concerns that we had on board and then feeding them back to ONS. So the flags were raised and were taken on board by Robert and colleagues. In terms of the first part of your question on my diagnosis of how we got here, I recognise this is just one perspective, and other people have other perspectives.

That are respected.

My 50,000 feet view of this situation is that the ONS has been a very traditional, production-focused organisation producing statistics for many years. Along came the pandemic, and it was required to be very much more immediate and responsive. It produced some new sets of statistics on things like economic activity, called real-time economic indicators, and a world-leading survey of covid infection. In a sense, that agility and responsiveness gave the organisation a sense that it could do more and do differently, and in the national statistician at the time it had a very ambitious, driven and extremely expert leader who wanted to drive the organisation to transform and improve. Once the energy of the pandemic subsided, the gap between the ambition and the capability became increasingly exposed. But unfortunately, with the desire to demonstrate progress, ambition and radicalism—ambitious and radical were two of the values and principles of the ONS’s strategy—it became hard for the organisation to recognise and acknowledge the accumulation of problems until about late ‘23, early ‘24. Through our work and the board’s challenge, it then became increasingly obvious there were growing problems. In a way, there is a human tragedy here that an organisation that thrived and was led with huge dynamism was not able to recognise that it was facing challenges. That is my diagnosis. I recognise Robert and Emma might have different diagnoses, but that is how I would summarise it at a very high level.

Do you think one can attribute trying to successfully resolve the hardness of the post-covid task in any way? It had ramped up its speed and just the dynamism of it had almost kept it going. Then, when the wind came out of the sails, it went back to normal things, but did not quite realise it was. You described it as being hard to address; what is your assessment of why it was hard to address successfully? Without answering my own question, was the heart of the problem that so many people notionally had an iron in this particular fire, but nobody actually had the ultimate directional whip hand? Was it that a lack of that independence versus accountability, ability and the opportunity to instruct had just become so clouded that there was a job that could be done, anybody could have done it, but nobody did?

I am afraid I have got a bit lost in the byways of your question there. It was hard to address because the organisation had a very ambitious programme on economic statistics, labour market statistics and population statistics, and it was trying to do difficult things very quickly. If I could put it succinctly, I suppose how I would summarise it is that it was maybe a little too in the position of believing its own propaganda of transformation.

Do you think there was any resting on the laurels, as in, “We are the ONS; we are the unchallengeable voice of UK statistical data,” and that provided an overly protective cloak?

No, I do not think there was resting on laurels; it definitely did not look like an organisation resting on its laurels. The independent review by Sir Robert Devereux makes this really nice point that amongst the principles of the strategy, the word “realistic” did not appear. It was a very ambitious organisation trying to do difficult things. The tendency that you have hit upon there though is not so much resting on laurels and being cloaked in the authority of independence but more an insularity, which I talked about last week: a sort of slowness to recognise that external challenge may have a point. This is a recurring theme in things that we have looked at and published. Maybe insularity is an unfair word, but it is a good single-word summary.

I have a final question at this juncture. It seems to me that there are some challenges of the board, given how it is constituted and so on and so forth, which we heard last week from Sir Robert. We have this tension between independence and accountability through to Ministers, the Cabinet Office, Cabinet Secretary, whoever. Flags are being waved. You have said that you raised some flags to the board and the board said, “Push harder, drive forward, keep flagging.” Are we right to think that there is a gap to be filled here casting forward as to how those are tracked and responded to, rather than just, “Yes, I’ve seen your flag waving, thank you so much. We’re just going to carry on with what we’re doing?”

That is a fair challenge to me, actually, as opposed to the board. It is my organisation’s responsibility to make sure that the things that we are requiring are being done. “Requiring” is an important word. We do not call our expectations recommendations; we call them requirements. There was actually a sea change in how we approached this in early 2023, where we recognised that maybe the super tanker of ONS was not turning round quickly enough, and we introduced much more specific requirements that were time-bound. I would like to just give an example of that in practice. We looked at ONS’s new population estimates, where it is doing this very difficult, sophisticated, new approach to effectively modelling the population. We published it in July 2024. We had 11 requirements and required an action plan from ONS on those 11 requirements by October, which it did, and in fact then published a further action plan in January 2025. So the lesson of us being more assiduous in setting really sharp deadlines is a lesson that we have taken out of this whole experience. Q95            Marcus Campbell-Savours: I have no experience working in an organisation as large and complex as the ONS, but I understand that there will be project management methodologies used within the organisation. I wondered if someone could talk to me a little about that project management process, and where it comes from. In particular, within the guidance that you follow—The Teal Book—there are requirements around reporting risk and issue management, and I would like to understand a little more about why those particular processes were not raising flags much sooner. Were those risks making their way up to the higher-level corporate risk meetings? Were they brought to the attention of the board?

We follow The Teal Book, as you say. All our large GMPP programmes have a dedicated programme office, and we have a centralised portfolio management office that sits separate and provides professional support. It is part of the professional body across Government of programme and project professionals. It also provides some second-line assurance. The programme boards will oversee the levels of meetings and support the senior responsible owner who will be allocated to the programme in programme board running, as well as all the other meetings and flow of information up to that board. The board will oversee the risk and the issue registers, and those will be funnelled into the strategic risks, as you say. We have a separate risk and assurance team, which will manage the conversation at an executive level about what the strategic risks are for the organisation. In the area of population statistics, we have had the census and now the future of population and migration statistics programme. We have a corporate-level strategic risk on population statistics, and both mitigation and residual risks are reported underneath that strategic risk. There should be line of sight from the very top of the organisation through the programme and project management framework.

Was there line of sight in these situations? Let us take the TLFS and the Integrated Data Service. If we were to look at that process and the people working on those individual projects, was there a clear line of sight? Were the risks identified being recorded correctly?

One learning from TLFS was that it was fragmented. It is only relatively recently that we have created the programme with a dedicated senior responsible owner, which I have just described. It was in part in the census development and transformation programme initially, so an element of transformation was part of that and it was reported through that. But there was an element in the ARIES programme that related to the outputs as well. So I would say that it was not best practice throughout the lifespan of the programme, but something that we have attended to most recently to address those historical issues has been to consolidate it in its totality. Q97            Chair: If it was fragmented and not best practice, the short answer to Mr Campbell-Savours’ question is, “No.”

Correct, yes.

You say that there is now a dedicated individual on the programme board, but was there anybody at that stage who should have had oversight? You mentioned a PMO but you did not call it a PMO. What was the name of the project management office you described before?

Do you mean for the programme or centrally?

Centrally.

Centrally it is the portfolio, programme and project office.

Who heads up that office?

That is a deputy director reporting to the director of finance.

Who is in that role?

That is currently a gentleman on temporary promotion.

Were they in place at the time of these issues that we are describing?

No, they were not.

Who was in the role at the time?

Somebody else, who is now seconded to a different role in the organisation.

You say they have been seconded to something else, but did they have a role in seeing that this did not happen? Are you saying that this somebody has taken responsibility for something there? As the Chair has pointed out, you are saying that the short answer was, “No, it was not best practice.” Have there been some changes in the organisation as a result of best practice not being followed on those occasions? Has someone been held accountable?

No.

Ms Rourke, I would characterise the function as we seem to be hearing it as a series of unaccountable musical chairs, self-styled DGs who are not, interims, temporaries, and people moving within the organisation before the music stops. That might just all be by happenstance, but the mildly cynical observer might just suggest that this was almost a deliberate management ploy to keep the plates spinning and the buck stopping with no one. That is not an eccentric way of describing the organisation in terms of deployment of human resource, is it?

I would say the accountability was dispersed, which is what I have just tried to describe, and that is unacceptable. As a result of the TLFS lessons learnt review that was done and the report that we have published, we agreed that something that we really must do is consolidate the activity and put an SRO in place, which was done. The approach taken last year was to take a hard look at ourselves through the lessons learnt report, be very transparent about the things that went wrong, and hold ourselves collectively accountable for what needed to change. That has initiated a whole range of changes across the organisation.

I understand, but “must do” suggests that these things were in some way novel or innovative; it should be “should have been doing.” I do not think we are hearing at all what the assessment is with regards to why the hell not? Basic things that anybody would expect in an organisation in the 2020s that should be done as a matter of normal best practice just were not being done. Was this deliberate? Did Sir Ian Diamond and others think that the rules somehow or another did not apply to them and that the ONS was something different? Did they think that they could cast a new way of working, hope for the best and nobody would notice? I do not want to speak on behalf of the whole Committee, but these things are just being said as if they are to be accepted without anybody asking why.

On that point, Chair—

I will come to you, Mr Lamb. Let me just ask Ms Rourke to comment on that.

I agree that this is incredibly serious. My response is in no way intended to reflect a degree that this is normalised. If you look at other programmes that the organisation has undertaken, you will see a greater degree of scope and control over here. We learnt a lot from the fact that we had this dispersed oversight of parts of TLFS and it was not adequately resourced at times. We talked previously about the covid boost being taken away. I would not say it was one individual or one choice; it was a systemic failure across a range of different factors that the Lessons Learnt report brought into very sharp relief for us.

So the ONS has sustained a systemic failure?

In that case, there were a series of decisions over time that led to the outcome where we really needed to recover at the rate that we have over the last 12 to 18 months.

From start to finish, everything that we have taken on this issue since this Committee started discussing it has been really frustrating; it has felt like it is everyone’s fault but it is no one’s fault. My line of questioning around the way the project was run was because I know there is someone who has the title of head of planning and portfolio management. My question would hopefully have identified that maybe that person would have had responsibility. I understand that failings can be dispersed, but we are talking about accountability and whether it is defined. So I ask the question again: is there somebody who is accountable for the failure to follow The Teal Book?

It would have been the senior responsible owner; they are accountable for the delivery of the programme.

The answer to that question is, “Yes”.

Yes, that is the purpose of the senior responsible owner.

What happens to them?

In terms of the initial programme that was born from the TLFS, the person who was the senior responsible owner at the time no longer works for the organisation. That role has since changed hands on a number of occasions.

Do they no longer work for the organisation by their choice or the organisation’s choice?

By their choice.

Did they jump before being pushed?

That would be speculation on my part, but I do not believe so.

A pearl-handled revolver and a glass of whisky were left on the desk; you can either write your resignation letter or—

I am sorry, that was not what I was trying to imply. The TLFS programme was initiated in the census 2021 programme.

But in short, somebody who should have done a job did not do it, and left the organisation without a blemish or a stain on their character. Is that correct?

I cannot comment on their reputation at the point at which they left the organisation.

Subliminally, that answers the question.

Following up on that last point, I studied project management quite a while back, and what you appear to have set out is the standard programme management process for dealing with projects. There is no doubt in my mind that the ONS would have originally had that in place across all its projects, because that is the standard way for any large organisation to deliver programmes. Following on from the question around how we got to the position, my question is at what point and for what reason did the ONS stop having such a process in place? That is relevant because if it has fallen out of alignment once, what is to stop it falling out of alignment again in terms of these things? To be completely fair, when I was a local authority leader in my own organisation, the only big problem or cost overrun that we suffered during that time was because someone decided to step outside of standard project management processes, and consequently everything went haywire. So what happened?

I will answer that in two parts. The first is the second and third-line assurance that is in place. The process that I just described is not in isolation; it has oversight from our internal audit. If it is a GMPP programme—which the census and ARIES programmes are—it is also subject to Treasury business case oversight as well as what is now the NISTA—which used to be the IPA—which would come in and do reviews and provide recommendations for improvement.

That is to monitor when things have gone wrong, but the whole point of having programme management processes is to avoid things going wrong in the first place. Who was setting up programmes without having a clear programme management process in place?

I believe that there was a proper programme management process in place, and we had a portfolio investment committee that listened and heard from every SRO on a monthly basis. That was all the directors of the organisation discussing risks and issues with the SROs. So there are a range of opportunities for conversation, surfacing issues and challenging at a senior level in the organisation, in addition to the professional oversight of the planning and portfolio team.

Has the PMO—or the name that you give it—carried out lessons learnt reports on these individual projects that show where the breakdown in communication or failure to act happened?

The TLFS lessons learnt review was specifically undertaken separately, not by the planning and portfolio team because it was felt that a degree of objective oversight was needed in relation to that exercise.

I was not involved in the direct commissioning, but I can certainly describe the senior civil service organisation-wide event where we discussed it as a concept. There was a poll that was done in the session where members of the senior civil service who were present were asked to rate where they were in terms of the axes, which I think were roughly around anxiety and safety. That showed quite a significant proportion of the senior civil service had high levels of anxiety but felt low levels of safety and were also expressing that they felt high levels of accountability, and so, as I understand it, the Labyrinth report was commissioned.

I would agree that the interventions that followed were not sufficient, but if I may I will just talk a little about what happened. To start with the governance, the hypothesis that we took away from the report was that there was a cycle where there was not good enough management information or governance. The senior leaders were not adequately cited on robust concerns and therefore their response was one of surprise, which then disincentivised further candour. You then had less information flowing to the top of the organisation.

There is a leadership issue there too, is there not? Obviously it is the responsibility of people working within projects to report on developments and if there are concerns in terms of progress. But it is also the responsibility of leaders to ensure they are creating a culture whereby that information will be willingly shared, and they also have a responsibility of scrutiny and leadership. Just on this report—which is serious—did the UK Statistics Authority board see the report? Was it presented to them?

No, it was not.

Would you not have expected it to be presented to you, Sir Robert, in terms of these being major organisational and cultural problems within the ONS? I would have thought it was pretty fundamental that it would come to your board for discussion.

Yes, I did make the point that I would like to have seen this at the time that it was produced.

What was the response to that and from whom?

The response was that the report had been undertaken, action had been taken in response to it and things had moved on. I was made aware of the report in the summer of last year and was told by the executive that things had moved on. I said that I felt that this was the sort of thing that should have been brought to me earlier.

You will recall, Sir Robert, that last week we talked about the role of the board, and we learned quite a lot about what the board defined its role as not being. I do not think we ever had a particularly satisfactory definition of what the board was actually for. As Mr Baker has just highlighted, from a managerial or governance oversight point of view, we have had what one could charitably describe as quite devastating. The board asked for some sort of presentation, briefing or report on it and you were rebuffed. Who rebuffed?

As I say, basically, when I was made aware of the report—

No, we have all of that. Who said to you, “Robert, me old friend, no?”

Nobody said, “No,” in that way. When I was aware of the report and asked for it, it was provided to—

How were you made aware of the report?

When I had been talking to members of staff in confidence—obviously I am partly there to provide that sort of safe space—somebody mentioned it to me, partly because I had suggested that maybe commissioning a report of this sort might have been a good idea, and I was told that actually one had already happened.

If you had not had those conversations with any members of staff you would be in the dark, would you not?

I would not have known that the report or this exercise had been undertaken. But as I say, I suggested conducting something of the sort and was told that something had happened but that that had been some time ago, and that action of the type that Emma has described had been taken.

Did you and the board deem that to be satisfactory?

We flagged the concern that I would have expected to have been shown this and the substance of the concerns around culture, which, as I say, were partly echoed by what I was hearing from individual staff in feedback through the performance management process. There was also a subsequent exercise—the My Experience one—which followed and was brought in at the next discussion of people issues at the board. But certainly the picture painted there was partly backed up by what I had heard from individual members of staff, and that was fed back to both the chief executive and the Cabinet Office in terms of the input into the performance management process through last year.

Just to summarise on that, you learn of a report that has been commissioned and have some indication as to what its findings are. As chairman of the board, you seek to explore that further, and that exploration request is refused and rebuffed. I would guess that the appetite of the board to oversight it would still have been there, so what were the board’s next steps?

It partly reflected the content of this, that obviously there was a process in performance management for feedback to the chief executive, so this was roughly around the time that was taking place. The deputy chair at the time pulled together reflections that fed into that, and the culture issues were very much a part of that and were fed in on that basis. As I say, when I was aware of the report I also made sure that it was flagged to the civil service HR function.

What happened when you waved that flag to the civil service?

It reflected part of the conversation that we had around the appraisal process in the summer of last year, and it was obviously something that ultimately fed in to the decision to convene the Devereux review.

If I may I will just come in there, Chair, as we did not necessarily hear the answer on a couple of the points there. Sir Robert, when you approached someone in the organisation and said, “This report’s happened; can the board see it?” you said they did not say no but essentially rebuffed you. Who was that? Who did you approach about this?

I asked the head of people to provide me with the report, and it was provided.

But you said that at some point the board should have discussed it?

I would have expected something of this sort to have been notified either to me individually or to the board, which did not happen, but when I asked for it, it was provided.

Did you make a point to the head of people that this should have been flagged to the board?

I made that point to the chief executive.

What was the chief executive’s response?

It was that he was content I had now seen it. The point was made that this was some time ago and that action had been taken on the basis of the sort that Emma has described. But as I say, the question of what had moved on and that there was still a widespread level of concern across the organisation of the type that this demonstrates was a motivation for setting up the Devereux review in the fashion that it did, with a widespread conversation across different members of staff. So there was an opportunity there to see whether those culture issues were persisting or it was in fact the case that that had been dealt with. But certainly the conversations I had with individual members of staff in confidence suggested to me that there was still an issue to be addressed here, which is why we did it.

Notwithstanding the fact that you took that action and commissioned the Devereux review, I am not quite grasping why the board did not push back and say, “No, we should have seen this report and discussed it. We still need to have an opportunity to discuss this report because it isn’t clear that action has been taken on some of the recommendations.” Why did the board not push further and go, “No, we need to discuss this?”

Essentially because it had been some time since the report had been done, there had been further action and we had a briefing on that. But as I say, in no way did we underestimate or undervalue the importance of the issues that had been brought up in that, and that continued to be brought up in conversations as well.

Just before we leave that point, that is worrying in and of itself and you must correct us if we are not following this rather complex story accurately. It seems that you—as chairman of the board—were in terms effectively told, “Well, yes, but of course, you know, we did this some little while ago now and things have moved on, we’ve addressed some things,” and so on. At what point did the board literally go nuclear and say, “That is not a defence. We should have been kept up to speed with this and shown it in a timely fashion after it had been received, to allow the board to discuss and comment upon it?” It strikes me that either the board was supine, or those who had the responsibility of sending things to the board and making it aware of things decided, in effect, to ignore the board when it suited them to do so to almost allow the musical chairs of accountability, “Oh, somebody’s made a mistake there; well, we’ll move them within the organisation. Somebody’s made a mistake there; we’ll let them leave the organisation,” and so on. I still cannot comprehend how an organisation supposedly focused on serving the public, Government, policymakers and others can have been run like that in this day and age, and the questions of Mr Campbell-Savours, Mr Lamb, Mr Carling and Mr Baker have intimated that as well. We cannot understand how an organisation supposedly populated by bright, clever, engaged people alert to the world around them is running this like some sort of rather abstruse element of a Tudor court.

As I say, I made it very clear I felt that I, at a minimum, and the board should have seen this report. We then obviously wanted continued emphasis on what was being done on the people front, so as they said, discussion around the My Experience report, which was the next stage beyond this. But the board and I were not content that this was sufficient. As I say, following the feedback that we provided to the executive and that the Cabinet Office provided to the executive and through the performance management process, I am sure we decided that the Devereux review was an appropriate way of, as it were, partly following up on—

We have had review after review after review. We had the Lievesley Review and the review that Mr Baker did. How many reviews did this organisation need before somebody rang the big alarm bell, pressed the big red button on the desk and went—I will not quite say what I would say if I had been there—“Dear heaven, something is rotten in the state of Denmark?”

Pressing the button was convening the Devereux review to check whether the culture problems that had been surfaced here—that the executive said that it had dealt with to its best ability—were still present and contributing to the performance of the organisation. That is very much the way in which that was framed, and obviously I am very grateful to the—

So, in essence, you need three people to tell you the building is on fire before you call the fire brigade?

We responded to this as we felt most appropriate and most likely to get good results. There was direct and very clear feedback to the executive about the concerns around culture that we had.

Lievesley should have triggered Devereux. The report that Mr Baker has just referenced should have triggered Devereux or something akin to it.

There is a difference there that the Lievesley review was at a higher level and less involved in talking to a large number of individuals across the organisation about culture.

She came to broadly the same conclusions.

The issues around personal culture here are inevitably starker than Denise was asked to do in that review. But I felt—the board agreed with me—that it was important to conduct the Devereux review in order to again provide an opportunity at a time at which you were getting mixed signals from within the organisation as to what sort of problem there was and whether it had been dealt with. We came back to that and said that it was an important thing to do. Obviously, conducting or convening an external review for publication is not a trivial step. I entirely accept the fact that it would have been better to be able to get to what I now believe is the point of turnaround earlier, but there was a process to go through in terms of giving feedback and the performance management process, which is obviously partly conducted with the Cabinet Office.

It just strikes me as being unbearably sad that an organisation’s reputation has been dragged so low. I go back to this phrase about resting on laurels. That is my takeaway from the whole thing, that there was this, “We are the ONS. Nobody will question us because we are the ONS and we can afford to operate as we will.”

If I may, Chair, on that—

That is the devastating situation in which we find ourselves.

I just wanted to nail down one point from what Sir Robert said, Chair. In discussing the report just here, you mentioned that there was a subsequent exercise on culture. What was that?

My Experience.

There were a few, and My Experience was one of them that consolidated a number of qualitative and quantitative interviews to play back to the organisation about where there were pockets of good practice and opportunities for improvement.

When was this done?

The first one was in 2024, and we will do another one.

In what month, roughly?

I do not know. We can come back to you with that, Mr Carling.

Thanks. As the Chair has said, I am just trying to build up a timeline here of all these reviews that say very similar things and when they happened.

September.

So we have from March 2023 to September 2024. Sir Robert, can you just remind us what you said just now in terms of when you found out about the 2023 report?

It would have been around May 2024.

That was before the subsequent exercise on culture that has just been mentioned. So when you found out about this in May 2024, you took it to the head of people and said, “Why didn’t we see this report?” The response was, “Oh, things have moved on; a board discussion isn’t needed on it now.” But actually things had not really moved on because you did not commission the next review for another several months.

I do not know at what point that exercise was commissioned.

There was a lot of activity in the intervening period. My Experience was a culmination of a range of different work that was played back to the organisation and published internally. There were a lot of other activities; for example, we would have director general and senior leader level engagement with the organisation. There was a view that we would have the corporate level focus on culture, particularly around quality. So quality was our focus and we realised that in order for us to achieve quality, we needed a proper culture where people would speak about their concerns in an open and non-judgmental manner. The two were brought together and we had a range of corporate and local-level open sessions complemented by local group level activities. For example, in my group, we would have in-person and virtual events where we would talk about empowerment. We would invite people who had positive experiences about feeling empowered and feeling like they had the conditions that they could do the best in their work to share with others what those conditions for success were. We also had people speaking about where they had felt less enabled and what they had in terms of support to improve their outcomes. We wanted to inspire and encourage people by listening to where there were pockets of really good practice because the organisation is not homogenous in its experience. To identify those pockets of really good practice was another string to the strategy that was happening in that period directly post the report.

Just for time reasons I will curtail that line there because we have what we need from that.

How many employees are there in UKSA and ONS, would you say?

Five and a half thousand, roughly.

Who on the board is responsible for people?

We have a head of people function who would appear at the board as and when necessary.

Do they not attend the board regularly, at every meeting?

Not at every meeting, but obviously there is regular reporting on people issues and they can be asked when they are needed. But obviously you are getting feedback from the chief executive, director general and executive members of the board as and when necessary.

This is not to catch you out, but would you prefer somebody on the board who is responsible for people, considering the issues we are discussing?

In terms of who the members of the board from the executive are, it is specified that it is essentially two members of the executive. Typically we have the people primarily responsible for population and economic statistics who are standing members, and then other people coming as they do.

We think it is actually three, Sir Robert.

With the chief executive as well, yes. Essentially it has been done in different ways in the past; you have had occasions when there is a sort of rotation. My most recent view is that it has been most important to have the two DGs responsible for economic statistics and populations; that is where most of the programme activity that we have been pursuing has been. Obviously you have other people attending as necessary, for example on communications or other issues.

I am happy to be corrected, but with 5,500 employees, I am a little staggered that there is not somebody who is responsible for people and culture permanently at board level. It worries me, if I am honest. I do not know what the answer to that is.

Obviously one could say the same thing on finance and other important areas of corporate activity as well. Given the fact that this is a board that is not merely a corporate board for ONS but has broader responsibilities—for example for Ed’s operation and much less directly for the GSS—you would end up with something that was very large if you had all those corporate functions represented. So people come at the appropriate point at which it is reporting.

Unless you are going to tell me this is the same issue with finance—which will raise even more issues—people are the biggest part of the organisation. You are not banging rivets into lumps of metal and selling them. Your people are everything. Maybe it is just my idea of culture.

People are everything in most organisations, at the end of the day. As you know, we are conducting a review of board effectiveness and how board oversight happens, and a question that we will look at is what the flow of information needs to be, who it is most useful to be hearing from, and on what periodicity.

In the interests of time, and I know a couple of people have indicated that you will be coming in again, I am going to go to Ms Yang as our guest from the Treasury Committee. Yuan, the floor is yours.

I am going to ask a bit about the Labour Force Survey and the economic reviews that have happened over the last few decades. There have been a series of external expert reviews of economic statistics at the ONS for several decades now. For example, a theme that I saw emerging in the Barker review from the early 2010s is the idea that there had been action on some recommendations from previous reviews—the Pickford and Allsopp review—but some progress had then subsequently gone backwards. It is very easy for us to discuss the problems with the Transformed Labour Force Survey and the underlying Labour Force Survey as pandemic issues. But I wonder if anyone on the panel wants to address the question of whether there have been peaks and troughs or shaky progress made against economic data reviews over a much longer period of time, going back to the 2010s and 2000s.

Ed may want to help on this because the most recent OSR review of economic statistics obviously looked back in part to the Bean review. I am afraid the Barker review is well before my time and out of the scope of my memory. Ed will correct me if I am putting words into OSR’s mouth that are inappropriate. But one conclusion was that, for example, the way in which action was taken around the Bean review was to emphasise putting resource and effort into the areas of relative novelty and new work, rather than the core production of the economic statistics. Clearly, with the benefit of hindsight, you would have said that it would have been better to have had more of that resource there. Obviously, when you have had previous reviews and then you have something like covid coming along, as Ed described, there is quite a significant pivot in the organisation to addressing what was regarded at the time—not unreasonably—as something that was particularly important at the moment, and that a previous review would never have had any contemplation might come across the horizon. You can run a thought experiment where maybe we should simply have said, “No, the organisation is not in a position to be able to set up the covid infection survey and do the other sorts of things that it did in response to that,” and the resources should have continued to be focused on core economic statistics. The organisation would then have come in to critique for a different set of reasons. As I say, on the Bean review, it was the decision to put money into the relatively novel aspects of it that—with the benefit of hindsight—would not have been the best idea.

That is exactly right and I suppose it gets to a systemic question about ONS, which is the focus on the core activity of producing high-quality economic statistics month after month, quarter after quarter, and the capacity to then evolve, develop—I avoid the word transform because it is overused—change and improve those statistics over time. There have been episodes in ONS’s history—one being the period post the Bean review, and the story around the TLFS is also like this—where the separation of transformation—that is the ONS word—from core has meant that both suffer. Transformation is a bit special and gets all the attention; the core is seen as less exciting. Again, this is something that Robert Devereux picked up. Where you see the ONS integrating regular improvements into that core regular process—which is what ONS does on prices—it tends to go much better. Emma, you might want to comment on this; bringing continuous improvement into the core of production is something that I have heard you speak about. That is a perennial of ONS, and the strong message is, “Do not have separate, shiny projects and programmes off to one side; align continuous improvement to the core and do it iteratively.”

For me, this process also seems to outline a pattern, which I will describe and you can tell me if you agree. For example, in the early 2010s, the Bank of England at the time was unhappy with flow of funds data. Since it is in public, there is pressure put on the ONS to improve and there are these reviews by external experts that are brought forward. There is some response to those reviews and we can then say with hindsight whether that was a correct response or not. We have that pattern repeating now, with the Bank of England again being unhappy with the Labour Force Survey data. Public pressure is mounting and there are reviews. To me, this speaks to a fundamental problem with the way that the ONS and its end users of economic data—particularly the bank—are interacting with each other. These issues are being played out in such a way that—rather than being resolved between organisations in a way that gives the bank confidence—there is a process for internal review before it gets to the point of having to call for public accountability and for there to be the scrutiny of the kind that we are now conducting. Do you see there being a systemic issue with the way that end users of economic statistics and the ONS interact?

During the 2010s, I was an end user of ONS statistics at the OBR. Inevitably, each big user—not just government ones but independent institutions as well—will have the particular things that they are concerned about and will flag and say, “We’d like to see improvements in this area or that area.” Obviously, the organisation has to deal with that coming from different entities with different priorities, and that is not a straightforward thing to respond to. What is actually quite helpful—I have had conversations with some end users about this—is obviously if they are able to get a collective view of what are the most important things to tackle at any given moment, that means that you probably have a more orderly process of responding to those sorts of issues.

Does that suggest that you believe that the ONS has too many priorities right now and it cannot sufficiently do them all well with the resource it has?

The economic statistics recovery plan—the survey plan—is showing a clear sense of priority and what are the most important things to deliver there. It is not exclusively that; we also have a census recommendation that—if Ministers accept it—has a set of outcomes and things that need to be done on that. The economic statistics recovery plan is a very good example of basically saying, “Well, what are the key elements?” The quality of economic statistics reflects base quality of things like registers, the collection of the data, the way in which it is processed and the way in which there is bandwidth for quality assurance. I hope that is very much in the space that you described there of seeing what the priorities are at the moment. As we go ahead, it is obviously one that you want to continue to talk to the end users about as to whether it is meeting their needs too.

We set this production of this economic statistics plan as a requirement on ONS because we spoke to those stakeholders. The point that you have made—that the stakeholders are maybe unclear on ONS’s priorities and what ONS regards as the core purpose of its economic statistics—was unclear. We considered that that was a state of affairs that was not only troubling for the users, but also probably for the producers within ONS. That is why we made the requirement. As I said right at the beginning, it is a good plan, but there is a lot to do.

I would like to turn back to the Transformed Labour Force Survey, which was first announced in 2022 to be—among other things—a way of responding to the drop-off rate in response that you were seeing during the pandemic. This survey was meant to be launched by early 2024; we now know that that has been put back by three years. What is concerning to me is that the design that was then developed was a very lengthy design that would not have addressed the issues with attrition. Can you explain why that had to come about?

Shall I start and maybe ask Emma to add anything that she wants? Yes, you are absolutely right. The idea of the TLFS was a long-term view of this being the strategic solution to the problem of declining response rates for household service. If you do them in the household, you have issues—which became even more dramatic after covid—such as people’s reluctance to let people in to have interviews, greater difficulty of getting into secure gated accommodation, a generalised reluctance to engage with public bodies in this sort of way, and so on. That is not unique to the UK, but as you point out, it is more pronounced in the UK, partly around the length of the questionnaire. Obviously, to some extent, the length of the questionnaire reflects the fact that you had different users coming to the ONS and saying, “We’d like information on this,” you end up accumulating more and more barnacles on the ship and the ship moves less quickly, if you like. Correct me if this characterisation is wrong, but the initial attempt with the TLFS tried to do with it what you were doing with the LFS and was getting a large number of questions and therefore too long a survey. As a result, you did not particularly get the response rates lasting through it as much as you would like. That is the rationale for now splitting the TLFS into a core labour market element, a longitudinal one, and a cross-sectional separate survey dealing with a non-labour market issue. So you have the core labour market stuff with basically a shorter, crisper, 15-minute exercise that more people are likely to get through fully, and that is out and being trialled at the moment. There are other changes in design that need to be implemented as well. But as you rightly say, that is the rationale for, first, recognising the merits of a primarily online approach, and secondly, recognising that if you are changing the mode of collection, you need to think carefully about the nature of the survey so you do not throw away the potential gains of the mode of collection.

Can I just clarify? To replay the account that you just made, am I right in thinking that in 2022, the TLFS was being prepared and the initial intention was to start with a short survey that was easier for people to answer. But as the months rolled on, more of your end users—the bank and so on—asked for extra questions and then extra this and extra that. It got so long that by 2024, you were asking an external panel of methodologists how you had managed to end up with this survey that was really not what was intended. Was that the process?

Again, colleagues will correct me. It is more the fact that the LFS was a long survey. It was not that the TLFS started small and accumulated; the TLFS was trying to do what the LFS was trying to do in terms of its length and scope. Particularly when you got to, for example, around May of 2024, when there was the question of whether you should move to the TLFS in the autumn of that year, there was this concern that the quality of the data was not going to be robust enough if you did that, which we can talk about. Part of the response to that was to say that we need to look at a shorter survey. Is that roughly the right picture?

Yes.

So the initial design of the TLFS was not that it was meant to be shorter than the LFS. That is helpful to know.

Maybe we should just redeploy the TLFS to another bit of the organisation and see if we can fare a little better.

I am interested to understand a little more about the board’s input into the TLFS process and at what point the board started to ring alarm bells. Sir Robert, you said last time we met that the board had intervened on the issue. At what point would you say you became aware that the early designs of the TLFS were not really fit for purpose and what was the board doing about it?

As I mentioned to Ms Yang, the key issue is around the points at which there were decisions that needed to be taken about the timescale on which it was realistic to jump horses from the LFS to the TLFS. If you go back to the run-up to the May 2024 board, that was a point at which the published plan was to hope to transition to the TLFS in September 2024. Both the board and I were aware that there were concerns—both externally and internally—about whether the data was going to be robust enough at that stage to make that shift safely enough. The key things that we did at the board were to make sure in the May 2024 meeting that the concerns or views of external independent experts were front and centre in the discussion. There were two key things that were part of that. One was that the Bank of England had set out some concerns about the ability to make the adjustment at that stage and whether it would be better to dual run for a longer period of time. It sent a letter to the director general for economic statistics at the time setting that out; that is now in the public domain. The other important—as we saw it—piece of evidence was that there had been a review commissioned from Professors Chambers and Brown of the structure for TLFS, which was feeding into something called MARP, the Methodological Assurance Review Panel. We felt that it was important that those were both taken aboard. The Chambers and Brown conclusion was that the design was fundamentally robust but that you were not ready to make the transition at that time; the bank was in a similar position. So the board made sure that those were both up for proper discussion. As the minutes of that meeting show, we endorsed the view of those external stakeholders that said that we thought that a longer period of dual running would be appropriate in order to ensure that things were robust. We also emphasised the importance of transparent and candid reporting to the board. At the subsequent meeting, the executive reported that it had indeed decided not to make the transition in September 2024. I am not saying that the executive left to its own devices would not necessarily reach the same conclusion. But we felt that it was important as the board to give a clear position, both on the importance of giving due weight to the external independent analysis and to feel free to give our view that we thought that that was the most sensible way to proceed, and that was the course that was taken.

So at this point, are you saying that the board is not so much concerned about the design of the survey but about the timeline for switching over?

They are related. Obviously, one reason you would be concerned about making the transition is if the length of the survey was such that you were not getting the quality of response that you needed. When you went to the subsequent board meeting at which it was agreed that it would not proceed in September 2024, the point was also made that a key thing to do was to investigate a shorter survey as a way of dealing with that issue. So the two are linked.

Sure, and I know that there is a reference in those minutes to the board requesting that the TLFS became a standing item to provide additional scrutiny, which was certainly the right thing to do. Since then, the board has essentially encouraged the ONS—as far as I can see from the minutes—to be clear that the final date for the delivery of the programme is still uncertain. Is that because you are remaining unconvinced by the delivery capacity?

The general principle is that a decision like that should be based on the evidence and data at the time, not on the fact that we have previously made a commitment to go on this particular date and we should go on that particular date, and whether you think that that is a robust thing to do or not. So yes, there is uncertainty about the point at which you do that, and you wait to see what the quality of the data is at the time. We have an external committee of experts chaired by Jonathan Portes that contributes to this as well. When the decision comes to be made, it should be made on the evidence, the substance, and on the basis of the views of external stakeholders. They cannot all be pleased all the time, but that is the right way to make those sorts of decisions, which was a very clear message from the board along with the importance of flagging concerns and risks around the path. The position as it is at the moment is that you would hope to have a version of the TLFS with all the necessary—or at the moment assumed necessary—design changes in place in the first quarter of 2026. Subject to a successful analysis of that in the summer of the year, you would be able to make the transition in November of 2026. But there is not and should not be a guarantee that that is the date; it should depend on the robustness of the data and what informed external consumers and scrutineers think is appropriate at that stage.

I am glad that you have brought in so much reference to external views on this because that is what is really critical here. Were external stakeholders involved from the start of the process, or was it a case that the board had to point out to those leading on it and essentially say there are not enough stakeholder views from the end users of this data being involved here?

I do not know precisely in the earlier stages. Obviously, there was always regular contact with the bank and the Treasury, which would have a natural interest in this sort of area. I do not know exactly whether there was formalisation at that stage. The Portes group is relatively recent and came in the wake of the discussion around May 2024 that we were talking about. The Chambers and Brown study was one that was commissioned by the ONS to inform the decision around the transition, and that was good and could well happen again if you needed to do similar assurance.

I appreciate that is the later on stuff. What I am really keen to understand is to what extent stakeholders were properly engaged at the start of the process.

That would be before my time. Do you know the answer to that?

Not in any detail. They were being engaged, but I have not seen any evidence that suggests they were engaged in the more structured and purposeful way that they are currently.

It might be helpful if you are able to write to us later with some more details just to the extent to which they were being consulted on the detail of what data needed to be included, what their key objectives are and what they are using the data for so we can understand if they were being involved early on. One more thing on this, and then I would have one other, slightly distinct question. Sir Robert, in the board minutes—essentially since this became a standing item—there are constant references and reminders that more effective reporting for the board is needed on some details. Just the length of time for which those comments were appearing in the minutes suggests to me that the situation was not improving. What were you and other members of the board doing to ramp up the pressure and seek to improve that governance problem?

As you say, whenever we saw that as being an issue, if we felt that we had come to a particular board meeting without the quality—or breadth in some cases—of information that you want, that was emphasised again. One issue was around the surfacing of risks and the ability of those to rise up through the organisation and then get to us. For example, a lesson in the “TLFS lessons learnt review was the importance of ensuring that that flow of information is there. Obviously, to the extent that you might see that as part of the broader cultural and performance questions, it was a motivation for the Devereux review.

But when would the board have gone a bit more nuclear? You had been raising these issues for months and months and nothing was happening. What would your next step have been? What was your next step?

Clear feedback to the executive that this was important and what we expected. As I say, ultimately, to the extent that you think that this is partly a cultural issue, that was part of the motivation for the review.

I am very conscious of time, and we still have a lot of territory to cover. Without curtailing Mr Carling’s line of questioning, I am tempted to say we have—

I was going to move on. I would just ask my final question at that point.

If you do that, and I want to bring in John very quickly before he scampers off, and then I want to bring in Miss Yang. I am very keen that we do not have episode three. We may have to but I am hoping we do not. Questions are proving to be nice and short and succinct; if we could have a mirror in our answers as well, that would be helpful.

My final question is for Mr Humpherson around the overall governance structure at the UKSA. The board structure seems to be pretty unique in terms of the OSR and the ONS reporting to the same board, despite the OSR regulating the ONS. Is that working?

In my view, it is, in two senses. One is the board, Robert as chair and the non-executive directors always completely support me in the judgments that I make. They are very clear—they back this up—that in the event of a disagreement between me and the national statistician, they would support my perspective as a matter of principle. The second and more important sense is that we do not just oversee the ONS but the production of statistics across the UK by the Welsh and Scottish Governments, all the UK Government Departments and the Northern Ireland administration. The test I apply is do we bring the same independence of judgment to statistics produced by, let us say, NHS England, Public Health Scotland and ONS? I would say the record shows that we do.

The reason I ask that question is that you clearly have a lot of deep-seated organisational problems within the ONS that need some work, and your work as the OSI is broader and covers lots of different organisations. I just wonder if there is almost a need for more board time and focus that might benefit from a split.

We have the regulation committee, which is a subset of the board, including Sir Robert, and is where we report in great detail. We get ample attention there, and a lot of that committee’s time—particularly over the last 18 months—has been taken up with us giving our assessments and judgments on ONS’s progress. You would be right if there were not a regulation committee, but the presence of the regulation committee is like my board in a sense.

Ms Rourke, in response to one of my questions last week, you suggested that there was a redeployment of resources away from areas such as health, population and UK-wide data. Who exactly makes that decision about the redeployment of resources?

Ultimately, it is the accounting officer: the national statistician and permanent secretary. Through the last round of business planning, the initial phase and segmentation sat with directors general to look and see what they thought would be lowest risk and more discretionary in terms of removing them from their portfolio to prop up the economic statistics and surveys.

The Devereux review criticised ONS’s budget and planning decisions. How can we be sure that ONS is making sensible decisions when it comes to budget planning?

The answer to that question comes from what do we think went wrong previously? In the last business planning round, something that was missing within the executive team—which is the group that made the final recommendation to the accounting officer—is that we did not set the right parameters at the beginning. There is a standard business planning process that is typically followed by all Government Departments. As Robert Devereux observed, we also did the end-to-end horizontal planning, and we attempted to delegate as much to directors and deputy directors in terms of what your costs and resource model are and what you need in order to be able to deliver high-quality statistics. That did not generate an affordable response. What I believe should have happened at the time was that early on, we should have set some parameters around what we felt should not be considered to be part of our portfolio. Our portfolio was too crowded, we were trying to do too much, and our discretionary pot was too big. We left it to individual directors and then individual directors general to determine what they were and were not going to attempt to fund. In the end, we made some decisions about what we were not going to fund, but I believe that had we made some difficult and inevitably unpopular decisions earlier, it would have been a shorter and more focused set of negotiations.

Do you acknowledge that the previous process failed?

I do not believe the entire process failed. The difficult decisions that needed to happen at the beginning of the year did not happen; they were not taken.

Was there an 80% or 90% failure statistically?

I would not quantify it in those terms.

That is unusual for a statistician.

A lot of this boils down to trust and being willing to have difficult conversations that are then difficult for the organisation to hear. We are not an organisation that likes stopping things. For the executive team to trust one another, agree that we stop certain things, hold the line together and reach a consensus as one is something that is currently a particular focus. We were not able to reach a consensus at the beginning of the year, and a final decision was made that meant we were over-programmed. As a member of that executive committee, I accept that those decisions were not made as they should have been. This year, we have the opportunity to remedy that and stop some work that I mentioned in the previous hearing. I also indicated that I do not believe that that is the end of it yet; we still do more discretionary work than we should be doing. We have to have a continuous conversation about where we stop the work that we do.

I want to go back to something that has been discussed a little already: your human resources capacity within ONS. To the outside world, do you think it is fit for purpose?

The function within the organisation has given up a lot of its resource over a number of years.

In terms of funding or personnel?

In terms of funding; it has been cut by roughly about a third.

Is what you are effectively saying that it has been taken off it and there has been a decision at the top of the organisation to defund the human resources department?

Correct.

So it has not given up; the funding was taken away from it by management.

As part of the negotiations, there were capabilities that were offered up, and we have created an AI-driven portal, so there is more of a self-service model. So it was given up in the sense that the service remains but there is a different way of accessing the information for people in the organisation.

Would you describe it as being robust in terms of how you appoint people? Does the appointment process within ONS meet the standards that any other external organisation would expect its human resources department to be running at, for example, in terms of following due process?

Certainly all the recruitment that I have been involved in—including my own appointment—has been done through standard processes. I was appointed as director general with a commissioner, and all the recruitment that I have been involved in has been according to civil service standards.

So you have no concerns about any of the recruitment process. In terms of the disciplinary processes that are in place and those individuals within the organisation who are perhaps not meeting the standard that you would expect, are you confident that it is meeting that standard that the outside world might expect from an HR department?

There will be performance-related conversations that suitably I am not privy to. As an organisation, I agree that our performance management approach is not as robust as it needs to be; you need a much stronger mechanism and a willingness to address performance problems. It is not just about the process; it is the willingness to invoke them and have difficult conversations. That completes the approach in its entirety and we have more work to do in that space.

Something that alarmed me greatly earlier in this session was that you made reference three or four times to pockets of good practice within the organisation. Does that suggest that the majority of those within the organisation are not meeting that standard of good practice?

As I understand it, there is not a consistent level of performance or assurance across the organisation that has been observed to me recently as well as over time, which is why having localised, local ownership of problem solving is really critical to raise up the performance and ownership of change. We have done two proofs of concept that have been very successful and place continuous improvement, and that is the blend of both delivery of high-quality statistics with people, communication and breaking out of silos. The ambition is that that is now rolled out across the organisation, such has been the success realised by those two pilots. I said that purposefully: we have too much variation in the organisation, but I believe that we have a range of mechanisms that can be driven and owned by colleagues in the organisation that will have a significant effect.

My last question is do you think the ONS is recoverable? Is it an organisation that is failing to such a serious extent that we should just scrap it and start fresh, start again?

No. However, the organisation is deeply challenged at the moment. The Devereux review has surfaced a lot of emotion and concern that have been expressed. We have talked about some mechanisms by which people have been enabled to express their concern, but fundamentally I believe it is recoverable. When I talk about some activities that we have done to address quality and continuous improvement, we see genuine signs of success and people who are wholly committed to meeting the standards that we set for ourselves. I accept that the leadership made significant errors, and I believe that we are in the process of being very honest with where we have made mistakes. At the moment, it is my job to support the organisation—in advance of a new permanent secretary coming in—to work through that period of what is quite an emotional conversation. It is that conversation where people can speak about their experiences without judgment but to be met with a series of activities and a plan that we have confidence in and can deliver against, and give the opportunity to the organisation to demonstrate that it has the capability at its core to achieve what we know it can.

I do not know how to break this to you, Ms Rourke. Let me be clear, nobody can doubt your empathy for and understanding of the issues. Picking up on Mr Lamont’s point, I am afraid my takeaway—which I take no pleasure in saying—is that the ONS is now in the junk bond market of Gerald Ratner’s jewellery stores, and I am showing my age when I reference that. I remain disappointingly unpersuaded of an energetic appetite for understanding there is no one straw that breaks the camel’s back. But the accumulation of the straws—the catastrophic failure of HR, the wacky style of leadership of Ian Diamond, the nebulous definition of precisely what the board can and cannot do and the tensions created by independence of operation—seem to have resulted in a laissez-faire attitude towards accountability, and utterly and totally divorced from the evolution of the corporate world—whether it is within the public or the private sector—operating in some sort of rather peculiar vacuum of whatever. All these things that you are pointing to—which I know you are trying to give us with heartiness—will be done. The obvious rebuttal point that we as a Committee and Mr and Mrs Smith on the Clapham omnibus would make is that these are the things we had an expectation that you would have been doing—not you personally, but the organisation—in any event. We are not talking about things that are novel or innovative, but about basic table stakes of operation. I am disappointedly unconvinced that the scale of the challenge when set against the scale of the accumulation of failures—both big and small—is readily understood. I may go to Mr Taylor because a lot of this speaks to Cabinet Office. Ms Yang, I will come back to your question, but in the time that we have available to us, I want us to go over the ground that Mr Taylor and others will wish to go over.

I would like to move on to some impacts on MPs in our dealings with the ONS. There was a suggestion that recent parliamentary questions appear to have been written by Cabinet Office SpAds rather than statisticians. I guess the question is whether there is a lack of confidence from the Cabinet Office in the data that the ONS is providing or it is interfering in the data for political reasons. Are Ministerial Departments censoring data that the ONS is providing? Is that something you are aware of?

Ministerial Departments are not able to interfere with the numbers; there is a robust basis for that. It is core to the independence of the institution that it is for the organisation to determine the methodology, how particular statistical outputs are produced and the way in which they are disseminated. Obviously, in terms of parliamentary questions, you have questions to a Minister. Typically, if it relates to an output from the authority, there will be a PDF letter from the national statistician attached to it on that basis. There is a long-standing convention that special advisers would not request changes to that sort of letter unless there is a concern that it is incorrect or it could be misinterpreted and those comments will be taken on their merits. I think that is how it works.

Just specifically on that then, I do not know whether you are aware of the example here, but we have an example of an MP recently asking for figures from the ONS on public sector headcount. The Cabinet Office discarded the analysis prepared by the ONS and instead drafted its own figures response. In response to a follow-up PQ, Cabinet Office Ministers did not deny the involvement of special advisers. So we are just really trying to explore what the Cabinet Office is doing relating to the provision of ONS data.

That is not something I can comment on as the Cabinet Office decision. I can certainly speak to the independence of the statistics and the drafting of the letter that goes to the Cabinet Office. Beyond that, that is a matter for them.

Just back to organisation, looking at your organigram and your annual report, you have staff turnover running at 16% compared to the civil service rate of 10%, 24 people on off-payroll employment, as it were, for over a year—six of whom have been off-payroll employment for four years or more—four people in your director and above level doing double roles, and then 11 on temporary placements; I assume that is what TP means on the organigram. Can you build a turnaround on that basis, and do you think the Cabinet Office is taking you seriously as an organisation?

Our turnover has dropped dramatically; it is currently at just over 10%, which was the last figure earlier this year. I am certainly happy to give you the current figure outside this hearing if that is helpful. I cannot give you the number but we have fewer people on temporary promotion, particularly at senior civil service level. I accept the challenge around the executive committee. I am acting national statistician and am still deputy national statistician for Health, Population and Methods. We currently have a temporary promotion for the deputy national statistician for economic statistics but we have an appointment who is coming into post on 4 August and is already engaging with the organisation in a really positive and constructive way. I feel that that will be a very positive appointment to that part of the organisation. We have previously talked about the split of the role, and with a new permanent secretary and national statistician coming in, we will have a degree of stability that is required and will bring energy and objectivity into the turnaround.

To the last point, do you think the Cabinet Office is taking you seriously?

Yes, I do. I have had some very positive conversations, offering support in whatever form that might take, that I personally need and what the organisation requires. Q183       Lauren Edwards: My questions are for Sir Robert. Can you tell us briefly about the relationship between UKSA and the Cabinet Office, and in particular how often you have met with Ministers and senior officials since the new Government were formed?

It is a good working relationship that necessarily reflects and respects the independence of the organisation and the importance of it not being seen to be politically driven in the outputs that it is producing but ensures that areas of mutual importance are dealt with appropriately. There is a memorandum of understanding that recognises the need for us to be in close touch on areas of common interest. There are regular meetings between myself and Ministers at the point at which they or I request. I met or spoke with Baroness Neville-Rolfe—the Minister in the previous Government—in the first half of last year on three or four occasions. Along with the then national statistician, I had an introductory meeting with the new Minister, Minister Gould. He had a separate meeting with her particularly around the need to discuss the census recommendation, which is a very important one and will be very important for Parliament to be engaged with, as ultimately it is going to be your decision how to proceed on that if Ministers agree with it. I would also see senior officials at the Cabinet Office on a regular basis, most frequently the permanent secretary to the Cabinet Office, Cat Little. As you can imagine given what has been going on—the performance management process and setting up of the Devereux review—over the past half year, there have been more of those meetings than there would be normally.

So you have only had an introductory meeting with the current Minister. How regularly do you anticipate meeting that Minister?

I am waiting to have my next meeting. The expectation is that there will be one soon, and as soon as her office comes back with a date, I will be very happy to have that.

You have previously expressed concerns about the way in which the Cabinet Office fails to appoint non-executive directors. Am I right in thinking that there has not been a risk and assurance NED on the board since May 2024?

On the substance of my concerns around that, obviously one of the issues was that we should have eight NEDs at full complement. If you look back over various points over the past three years, we have had up to four vacancies. Most obviously, the main reason for that is we had three and then very briefly four because we had an appointment process that was under way, then the general election came along so that had to be stopped. You then had to have new Ministers put in place and obviously let them settle in. So clearly, that was a longer period than would have been ideal, but not the fault of the Cabinet Office or Ministers at that point. Given that this was the point that we were engaging on the TLFS—as we were discussing earlier—we showed that with a smaller number of people, we were still able to engage on that point effectively. The other issue I had raised—which was with the previous Government—was less around vacancies, but the fact that there was a view under the last Government that you should basically not renew NEDs after the first three years; no doubt a desire to refresh the makeup. That happened in late 2022 and did not result in a long period of vacancies, but arguably you lose some institutional memory and skill that would have been quite valued. At the time, I said publicly that it would be better if that were not to be the case. So those were the two issues around there. Obviously, we have the audit and risk committee of the board, the chair of which reports regularly to the main board meeting on the most recent audit and risk committee, assuming there have been ones since the previous board. Jacob Abboud currently chairs the audit and risk committee. He was interim during the period in which we had the last NED recruitment round, which as I say we were waiting to see what came out of that. He was then made permanent and is an excellent chair of that. In the most recent NED recruitment round, we have also bolstered a new NED who sits on that committee as well, with excellent expertise of having been on audit and risk committees previously. So that has been a very welcome strengthening of that part of the board.

So there is a properly recruited chair for that risk and audit committee; it was somebody who was interim and then you have made them permanent.

Yes, that is right. It was because it coincided with a recruitment round where we explicitly said that we were also looking for more experience around audit and risk. So you did not know quite what it was but the current incumbent was doing very well. We obviously discussed it with both the new arrival and him and are very content with the quality and robustness of challenge that comes from the audit and risk committee and gets fed through to the board.

More broadly, what impact do you think some vacancies have had in terms of the oversight of the ONS and the other parts of the organisation?

As I say, there are two elements to this, one of which is if you are just short on numbers, the other is if you unnecessarily lose some skill and expertise. In her review, Denise Lievesley pointed out that a consequence of people not staying on beyond an initial term was that we did not have someone with specific, strategic communication skills on the board. She rightly said that was important. We had tried to find someone in that space in a recruitment round and did not find anyone, but we now have an excellent person in that space who has come in through the most recent round. So that has been filled on that basis.

In terms of the Cabinet Office, do you find that they have a good grasp of the organisation and the requirements of operating in that statistic space compared to, say, some of your other stakeholders, such as Treasury or the Bank of England?

They do, partly because they will understandably talk regularly to the Treasury around issues of common interest. Given the nature of the authority and the fact that it spans far beyond economic statistics, that is a perfectly reasonable reason for the Cabinet Office to be the sponsor Department. But it is obviously alert to what it is getting from other parts of Whitehall and concerns, and the Treasury is naturally a very important stakeholder directly for ONS. I would obviously make myself available to talk to people at the Treasury as and when they want to do that, if they feel it is important at the board level. But there is an awful lot of working interaction between the Treasury, the Bank, the OBR, the Government and independent stakeholders in the economic statistics space, and the Cabinet Office is obviously able to plug into that.

A final question from me. Hypothetically, if you had concerns about the performance of the national statistician, is it your understanding that you would have the power as chair to remove them?

The national statistician is a Crown employee and a permanent secretary. At the end of the day, it is the Cabinet Secretary who is the line manager. If there are matters in that space, it is ultimately a decision for the Cabinet Office and the Cabinet Secretary, but it is clear and written down that that should be after consultation with the chair of the authority.

Just to return to a previous topic, I wanted to get something straight about the design of the TLFS. In the answers that you gave to my questions earlier as a panel, you described that the original intent of it was not necessarily to produce a shorter survey. In its concluding remarks, the Methodological Assurance Review Panel has written that the original concept of the TLFS was to produce a short online survey, collecting data on core LFS variables. What was the intention behind the original TLFS? Is the unclarity of that intention something that has led to the vagueness in the execution of it?

I would say yes. As we discussed earlier, there have been several SROs over the period of its development. I do not think it has always been consistent and there has always been an aspiration to have a shorter online response. Sir Robert was talking about the pressure to add more and more to it, and it appears it lost its way.

So the original intention was to have the shorter, more compact survey, and then it lost its way through the process?

That was the advice at the time. What we have now in terms of the two 15-minute surveys is certainly different to what its initial conception was. What we launched yesterday was the core 15-minute survey. I do not think that is what was initially suggested and is almost at the extreme of a condensed core set of variables that focus very specifically on the labour market rather than a broader range of socioeconomic variables, for example. What we launched yesterday is not necessarily what the vision was in the mind’s eye; it was something that was shorter, but I could not speak to the precision of how short it was originally intended to be.

Sir Robert, did you want to add anything?

No. The early stages of this were before my time, but that sounds very accurate.

We will have to move on in terms of time, but just to put on the record that if I had been a statistician working for the ONS, I would still remain a bit confused as to what the original plan was for this TLFS, which of course would impact the execution of it.

I do not think you would be the only one to be confused; I will join you in the confusion lobby. Mr Baker, a man who is never confused.

It was intended to be a platform and a trusted research environment that would enable what was described as friction-free access for accredited researchers to a range of indexed data from surveys, administrative sources and other public sources.

Since 2020, the cost to date has been £224 million.

There are two parts to my answer, if I may. The first is what are we retaining in terms of the platform? What the money has been spent on is a set of pipelines and a platform with a tool bench. As an analyst in ONS, you can utilise the platform and it is highly secure. The three indices—demographic, address and business—are sufficiently mature to start informing a lot of our analysis. We have multiple datasets that are published on our websites and 29 projects. From ingestion through to dissemination, we have something that is stable and holds individual-level data safely.

There are a number of projects that are live on there at the moment and we are just working through the future of them. We have closed the system to any new external. It is quite a resource-intensive process to onboard new datasets, go through the process of accreditation for the researchers, and get permission to use the data. We have a Five Safes model and quite a lot of weight is borne by the accreditation process to assure the public and other users that we are using data in a responsible fashion. That is the bit that is stopping. We are not enabling any further government analysts to be onboarded; it is being internalised. The second part of the answer is that we are having conversations with the Department for Science, Innovation and Technology about the National Data Library. As it is developing its thinking around the capability that is required to support access to the data, including the indexing capability, are there further benefits that can be realised from this investment that support that vision and aspiration? We have not concluded those conversations, but in addition to realising the benefits for analytical development—not least potentially for the census and administrative data solutions to support economic statistics—we also have the possibility that it supports a broader government strategic intention in the longer term.

No, that is continuing.

That is entirely possible. It is currently live. We have about 900 projects that are live on the secure research service currently, with nearly 2,000 accredited researchers engaging on those projects. It has a slightly different set of tools and a slightly different tool bench to the IDS, and that is welcomed by the academic community in particular.

I do not know the answer to that.

Yes. We have projects that support the missions, for example, and that is why I was hesitating to say we have not yet determined the length of time that those will be in train for before we wrap them up and how we manage that portfolio of projects. But in terms of new ones that might be waiting to be onboarded, if I may write to you after this session.

That would be helpful, thank you.

Just one quick question. Ms Rourke, you mentioned that the Cabinet Office has been very supportive, which is very good news. Sir Robert, you told us you spoke directly with the Cabinet Office about the creation of a new director of operations post, which sounds eminently sensible. Who did you speak with in the Cabinet Office?

The issue that had been raised with me here was whether the salary attached to that role was consistent with civil service pay rules. So I asked the chief executive to provide me with written assurance on that, which I received from the people function within ONS, that ultimately the necessary requests and approvals have been reached for that. For completeness, I was having a conversation with the director of senior talent at the Cabinet Office on some other area and asked if it would be possible to simply confirm that that was the case, and she was able to confirm that with the relative team for me.

Interestingly and unfortunately, you have led me to another question. What else did you discuss around the issues and the reason you need a director of operations?

As I say, this was a very limited question around a particular query that had been placed around the pay attached to this or the change in pay as the role was broadened. The broadening of the role is something that happens in existing roles across Whitehall an awful lot. It is within the accounting officer’s powers to do that, and I was just checking on that specific issue. There was no specific meeting set up to discuss a range of issues.

You will be delighted to know I am not going to sing to you. I had to check so that I got the precise wording because somebody would say, “Oh, Chairman, you haven’t got the right words,” but I am reminded of Johnny Nash in 1972. I was three at the time, but I vaguely remember my mother playing it. His lyrics ran, “There are more questions than answers; pictures in my mind that will not show. There are more questions than answers, and the more I find out, the less I know.” The more we have found out over these last two sessions, the more we want to know. There is less time for questions but between the three of you over these last two sessions, you have painted an honest picture of the ONS. Both colleagues and I still find it—as I hope has been intimated in the tone and range of questions—an utter befuddlement as to how any organisation in today’s world can have slipped or slid into the hole that it has. Picking up on earlier questions, I do not know whether there are enough grappling hooks, ladders and energy to get it out. We shall watch this space and will obviously be hearing or discussing the issues with the Cabinet Office, Ministers and others. My takeaway—I rather get the impression that this is yours as well—is that you can dress it up in all sorts of language, but it is not just a fundamental review and recalibration of operational approach. It is a fundamental orthopaedic review of the skeleton that is the ONS, what links it all together and to whom it is accountable in order to ensure—I do not seek to scapegoat in any way—the character and/or characteristics of one or two people in senior leadership can so percolate down through an organisation that has clearly both destabilised it in terms of reputation. As we have been hearing, it has fundamentally discombobulated the men and women who work within and for it, who are probably getting towards the end of their tether. If you are happy to take this message back, I want them to hear that we as a Committee will be their champion to make sure that if there is a turnaround plan, it will be adhered to, and woe betide anybody who stands in the way of it. Again, I reiterate our thanks for the three of you finding time to clear your diaries to appear for a second time. I am tempted to say we could fill a third slot, but I am not going to tempt my luck. Thank you very much indeed.