Good morning to colleagues and our witnesses. We are looking at the ONS and the serious issues that face it. Just to put our witnesses on notice, as it were, what the Committee has heard hitherto has left it frustrated and underwhelmed, so today we are looking for some focused clarity of answers and hopefully some hope and/or light at the end of the tunnel. My first question has to be, in your assessment, is the ONS in any way aware of the depth of the reputational hole into which it has sunk?

Thank you very much indeed for asking us. The authority and the ONS do not in any way underestimate the importance of improving the quality of core statistics and ensuring that there is public confidence in the robustness of the numbers that it can produce. The plans that have been published by the ONS recently on economic statistics and surveys demonstrate the appreciation and importance of that. I know ONS colleagues are going to be pursuing those plans promptly, with determination and the support and oversight of the board. As you say, Chair, the value of those reports is recognising the depth and nature of the challenge that has to be addressed. There are multiple elements to this, which include dealing with—at base—the quality of data collection, the registers, the foundations as it were upon which the system needs to operate, the way in which information is collected, for example, and the importance of increasing the number of people who are undertaking surveys, so strengthening the field force in that respect. Crucially reflecting the diagnosis that Robert Devereux helpfully provided is the importance of essentially strengthening the production teams within ONS across the range of core statistics. The plan is aiming to invest £10 million, partly on systems but also on getting 150 people into the core teams, which will provide scope for continuous improvement, greater bandwidth for quality assurance and will put things on to a better course. In no way does the system as a whole or ONS underestimate the challenge, and it is determined to address it as vigorously as possible. I am very glad that Robert Devereux has commended the plan that the ONS has put together at pace, partly in response to the critique of economic statistics of Ed and colleagues and through the regulation committee. He commended it for focus, comprehensiveness and honesty. We reflect that, and I know ONS colleagues will be pursuing it with determination, so there is absolutely a recognition of what needs to be done and the importance of doing it.

We are all familiar with the phrase, “There’s lies, damn lies and statistics,” and while what you have just set out may very well be welcomed, none of that could be described in any way as being revolutionary. That is surely just the core competence of any public body, particularly one dealing with and producing data and statistics upon which so many policy formulators and the Bank of England rely. There is a bit of a distinct juxtaposition between the question and how you have energised the answer, which was, “Well, 80% seems to have been working very well; 20% not as well as it might have been. We have appointed five jolly good people to reinvigorate all this. The caravan must move on; there is nothing to be seen other than draft strategy reports that talk about ‘within the lifetime of the comprehensive spending review’, ‘in the fullness of time’ and ‘at pace’.” I do not think this Committee is persuaded with regard to the fact that the ONS collectively has an understanding of the size and the depth of the hole that it is in. Surely there has to be an expectation among parliamentarians and the public that these so-called remedial correctives that are being put in place should have been in place in the first place. So our next question has to be what the hell went wrong?

If I can go back to the earlier part of your question, you essentially asked whether there was a need for revolution as distinct from the bread and butter. Actually, if you look at some critiques that Robert Devereux put forward, in the past there has been a bit too much revolution and a bit too little attention to the bread and butter. I do not think the report really apologises for the fact that it is a good idea to actually get people into the core teams to provide greater bandwidth to deal with the underinvestment that there has been in the past in some aspects of data collection and systems.

But how would that circumstance have come to be? There are core competencies. I might ask Mr Humpherson, who was nodding vigorously at one of the remarks that you have just said. It is very nice to see you again, Mr Humpherson. What is your assessment as to why or how?

Should I start or go straight to Ed?

No, let me ask Ed.

Helpfully, we did our review of economic statistics, which we published in April, and we found three problems. First, a fundamental decline in the quality of data sources; secondly, a lack of transparency on the core purpose and prioritisation of economic statistics, and then those two things combined to make it a challenge for ONS to make sufficient resources available to do the core work. What we then said is that ONS needs to restore confidence urgently, which very much goes to your point, Chair, about the need for urgency here. We then said that the first stage in that recovery has to be the open acknowledgement of quality problems. If I can then take a step back from that and answer the causal question that you are asking.

Just before you do, this may be a supplementary to the causal question: did those two things that you have identified happen overnight or had they been a long time in the gestation? The one thing we have yet to find anybody who is prepared to speak definitively and authoritatively on is when did the rot set in?

These things did not happen overnight. We publish a regular cycle of assessments and if you look back, it was in March 2020 that we first highlighted the problem with declining response rates. In 2021, I wrote to the director of national accounts about the problems of a lack of investment in the core data collection actually on business surveys, but that is a key input into business statistics. So these things were a long time in gestation. But if you were to ask me what the moment was when the straw became too heavy on the camel’s back, it was in October 2023 when the ONS was unable to produce reliable employment statistics and withdrew them. That was following a decision in July 2023 to withdraw an extra boost to increase the survey responses as a direct consequence by October 2023. That was the moment, and it was the moment when the outside world started to see these issues beginning to emerge. That is my honest answer.

So the genesis is five years, and the peak of the decline—if there was such a thing—set in two years ago?

That is your construction on what I have said. I would just say these things were a long time in gestation, but the key moment was that October 2023 Labour Force Survey. I do not know if you want to come in on that.

Robert Devereux makes a compelling narrative on this. Essentially, having performed and being seen to perform impressively externally—in particular through the census, its response to the covid pandemic, stepping up the Coronavirus (COVID-19) Infection Survey and so on—in a sense, the ONS bit off more than it could chew in the 2021 spending review by seeking the big ringfenced programmes around IDS population and across economic statistics. It also tended to spread itself too thinly thereafter by responding—no doubt with the best of intentions—to new demands and requests from users, which were well received in many cases. The consequence of that was that there was a deprioritisation of the bread and butter work on economic statistics. It left too many people reliant on legacy systems and there was underinvestment in surveys and data collections. As I say, in some respects, that reflected a well-meaning desire to innovate and meet new user needs. But as Robert points out, it caused greater problems than would have been ideal in the organisation because of weaknesses in planning and budgeting, for example, how you deal with bringing spending back down again after it has gone up, the way in which prioritisation took place, and the fact that you were in a situation where bad news and concerns were not rising up through the executive and to the board as they would do. Robert’s summary sentence is that it is the consequences of choices taken and not taken over several years. As Ed points out, there was a particular moment around the LFS and the problems that happened in the summer and into the autumn of 2023, but that is the longer-term narrative that Robert puts around that, which I have no reason to dispute.

To put that in layman’s terms, there was a catastrophic failure of planning and analysis, and effectively doing an MOT on the organisation and making sure that the core foundational expectations were not just being met but meetable, notwithstanding what is a perpetual thing for innovation and improvement.

Yes. There was not that attention over this period. There was a desire to respond to other requests from users, and obviously this was in an environment in which there was much greater awareness of what impact data could bring, but that had consequences for the underpinning foundations. The value of the plan that Emma and colleagues have been putting together—both on surveys and economic statistics—is that it is trying to identify the different elements of that. It is trying to recognise that you do not solve all those things overnight but that you can have a progressive period of improvement based on getting the underlying foundations right, putting more resource and better ways of doing data collection, and having more resource placed into the core production teams to give greater scope for continuous improvement and greater capacity for quality assessment. If you look at some errors that arose in ONS outputs, some are down to coding changes that were made a very long time ago but have been unearthed and corrected as they were found. That is obviously a good thing, not a bad thing; you want to see errors found and corrected rather than just leaving them there under the soil. But there were also issues around errors in the data that was supplied to ONS, and if you do not have enough capacity in those teams, then you are less likely to spot that and be able to correct it. That fits very much with that sense of strengthening the foundations, and the report recognises both the substance and the seriousness of the need to do that. As I say, I was pleased that Robert has commended ONS colleagues for the work they put into that and the approach that they are now taking.

Mrs Rourke, you have been part of a poisoned chalice, have you not?

No. It is a real privilege to be leading the organisation at this time. While I am in this acting position, I have the opportunity to surface some deep-seated issues that Sir Robert talks about in his report. We are in the process of listening to colleagues across the organisation and ensuring we attend in a very broad but also very deep way to some systemic issues that have led to what has just been very eloquently described.

For perfectly legitimate reasons, most interims will say that they are there to maintain a holding pattern or tread water until the permanent comes in because one does not want to start a whole process of work that a successor may think is wrong or wish to alter. Given the scale of the task and the size of the hole as we assess it to be, can you assure the Committee that that is not the case in your interimship?

If I may, I would like to respond by speaking about some things that I am doing now. The plan that Sir Robert Devereux has indicated has been published alongside the Survey Improvement and Enhancement Plan, and that is our due north. There has been criticism legitimately levelled at us for not having that clarity of purpose. So the plan and the sequence of change—making sure that we do not overload ourselves and giving colleagues confidence that what they have put forward as a valid set of deliverables is going to be supported by the executive committee—is the starting point. More broadly, when the report speaks to culture, for example, that is not going to be solved in the few months that I am in this position. However, what I can do is build on the work that started some time ago in establishing the right mechanisms for the right conversations so that when we do change, we do it better, and when there are opportunities for people to speak truth to power, we enable that. That will all take time. As has already been noted, there are parts of the organisation that function in a healthy way, and we have to understand and learn from that. For example, we understand that the knowledge transfer is not as optimal as it should be and the lessons learned are not always as internalised as they should be. So my plan is to start setting out a plan and a set of options for cultural change that the new incoming permanent secretary can consider in partnership with the organisation.

I will direct a question to Sir Robert Chote. Given the problems that we have been discussing this morning already—some of which have been described as systemic—do you think that you as chair and the UKSA board have been effective in your oversight of ONS?

Inevitably, if you commission or co-commission a report that finds the sorts of relatively longstanding issues that Sir Robert did in his report, with the benefit of hindsight, you wish that you had got to this position earlier so that the turnaround—which I genuinely believe is underway—could have got to this point to begin with. There are different dimensions to this but the key parts of the task of myself and the board are to provide support and challenge to the ONS and the OSR on an ongoing basis. At the same time, as a board with predominantly non-executive members, we cannot and should not try to do the executive’s job for it, blur lines of accountability, and cause a different set of problems by essentially wading in and trying to run the organisation ourselves when we see difficulties. What you essentially see is a necessarily staged response, intensifying some support and challenge, and as you would see from the minutes, requesting and making the point that it is important to see a good flow of candid, balanced, transparent information flowing up through the organisation itself and then to the board to fulfil its oversight functions. Obviously the fact that we have had the review has not come out of thin air. The board had concerns; it heard concerns internally and externally about performance and culture, hence the scope of the Devereux Review when we set it up. Obviously I speak to staff and former staff and you get that sense from them as well. In addition, there is engagement with the Cabinet Office as our sponsor Department, making sure there are consistent messages going to the executive about the concerns we hear and the importance of what to do. In the end, we got to the point of feeling that the review was necessary; the Cabinet Office and us were agreed on that, so I proposed it to the board and it went forward. That was an appropriate step to take. It was also clearly an unusual step to take, but Robert ended up—singly or with others in various groups—talking to about 200 members of staff, bringing that outside perspective. If we—as members of the board—had tried to engage in that sort of exercise, it would have horribly blurred the appropriate lines of accountability with the executive and non-executive. At the end of the day, the board has responded to the concerns that it had and heard. Over time, it has moved to a relatively high-profile intervention of calling this review. The key thing now is to get on with the turnaround that it called for and is endorsing in terms of the plans for economic statistics recovery and surveys and the pending changes in the leadership. Subject to the Prime Minister’s approval, we will have a new senior leader—permanent secretary—to take on the task of providing greater operational leadership to the ONS. That allows you to build on initial steps that have already included the broadening of the executive team that Emma leads. You are bringing more people from the core production teams to the top table so it will be less of a pyramidal structure. Also very importantly, Emma—with the support of the board—is very much working on improving the quality of the management information that the executive has to judge on what is working well, what is not, and where the risks are, and that coming to the board for us to be able to take an informed view of rather than an inadequate and partial flow of information. That is not just about bringing a whole lot more data to us but better data that sends better signals. We are quite an unusual board because we are clearly not just a corporate board for ONS in a straightforward corporate sense; we have broader responsibilities, obviously working very closely with Ed and the OSR. I am on the regulation committee, so the reports that OSR has put forward over time on what issues it sees in economic statistics, what needs to be done and following that up is part of the board’s activity as well. I do not shy away from the fact that in an ideal world, having commissioned a review that found what it did, we would have got to that position more quickly.

Just on the management information point that you made, something that we have heard is that the papers that were going to the board were quite thin and repetitive. Would you agree with that assessment?

This is something you would hear from many non-executives in many board situations. Ideally, what you want is a focus on the issues of concern, and where the board can assist with providing advice, you want transparency and clarity around that. For example, if you look at what the board said at its meeting after the failure of the need to withdraw the employment numbers in October 2023, and around June 2024 when it was clear that the TLFS could not transfer, the minutes are showing that the board is saying, “We need to have good, accurate, balanced reporting here rather than a rosy view and too much of that sort.” You are absolutely right in the sense that what you want is information that comes to you with where the board can be of assistance, what it needs to worry about, and what is worrying the executive at the moment. It is not straightforward because you will have colleagues who have a whole different variety of views on the things that they are working on, and at the end of the day, you have to funnel that up to a view, distillation for the board and decision making. But for the board to do its job effectively, you need to have a balanced, transparent flow of information coming to you and something that is focusing on the key questions. One thing we did a while back was basically, when board papers were coming up, we got other people in the secretariat who were not directly producing them to ask what the key questions were that we actually wanted to answer out of this rather than something that simply said, “Board is invited to note A, B or C.”

Over the last five or six years that we are looking at here, would you say that the management information that you were receiving was painting an overly rosy picture and that that did not enable you as a board to tackle some early warning signs that things were starting to go wrong?

A couple of examples I gave you there align with that. Certainly those board members who were around earlier during this period—certainly before I got there—got that sense that maybe for 2023 onwards, you were not seeing the same degree of breadth of view and surfacing of concern that you would want. Obviously this is an organisation that is working in a challenging, not straightforward environment, and the wish to press on with that is an important one. The importance of clarity, balance and candour in the reporting of information and the explanation of what is or is not a source of worry is absolutely vital to the board’s ability to do its job properly.

Some things that you have been discussing raised concerns about people in the senior leadership team at the time and perhaps the effectiveness of the board and the sub-committees—particularly the risk committee—underneath it. We have had a couple of reviews—the Devereux Review is one—of the ONS recently, but neither has really looked in depth at the governance of UKSA and the ONS and whether that is appropriate and working effectively. Do you think it is time for that review to happen? It seems to be the missing piece in the jigsaw.

The third recommendation of the Devereux Review was that—in addition to the near-term splitting of the roles and the opportunity that that provides for a more focused perspective on operational leadership of ONS—you look to the longer-term structure of the governance of the system and ONS. We will work with the Cabinet Office and are obviously very interested to hear the Committee’s views—it has expressed views on this in the past—as to how that should best proceed. If you look at the sort of issues that have been talked about at the high-level governance level before, there are obviously questions around exactly how the role of the National Statistician is defined and normally interpreted. There are also questions about whether it is basically bringing too wide a range of responsibilities for one person, which is addressed in the near-term, and there is obviously a question mark about whether you do that in the longer-term. There have been questions raised in the past about whether it is appropriate to have the regulatory and production bodies under the same governance umbrella. Again, the Committee has expressed views on that in previous iterations. In the past, that debate has often been characterised about whether it puts OSR into a difficult and conflicted space by being too close and under that umbrella. As I explained in my appointment hearing—this view was backed up by Denise Lievesley in her review—there is no evidence that there is a cosy relationship, that OSR ever pulls its punches or that the regulation committee—which Denise Lievesley came and attended—instructs or encourages the OSR to pull its punches. Ed would probably agree with me that if anything, it tends to nudge in the other direction.

Yes.

What people may look at now is less as to whether this structure serves OSR most effectively, than whether it provides appropriate governance for ONS. Rather than having a dedicated board for ONS, obviously at the moment, we have one with much broader, system-wide responsibilities. Those are issues on which I know people—who have had views on them for many years—have given you evidence, and it is perfectly appropriate for that to be on the table now.

That is helpful. Obviously it is appropriate to have somebody independent come in, but what was your view as chair? It would certainly be my understanding that you would be able to gauge cultural and performance issues yourself through your discussions, particularly in board. What were your views around those particular issues as they were emerging?

That we had concerns and heard concerns. Obviously, in one sense, it manifests itself. It is not a homogeneous organisation; there are different aspects to culture in different parts of the organisation, and Emma will give you a much more informed view on that. In terms of the way it is most pertinent and live to the board, it is these issues of the quality of debate, reporting, candour and breadth of what gets up to you. Certainly, from conversations with staff colleagues, you hear particular concerns that people have. But Robert made the point that there are inevitably mixed signals here. You hear some people saying, “I’ve got worries in this particular area; I’m not sure the decision-making is working as effectively as it could be.” There are other people who take a different view, either that they do not see that problem or they recognise there has been a problem of that sort but that it is being dealt with. One reason for convening the review was to cut through that thicket of mixed signals, as it were. It would have been impossible and inappropriate for me or some subset of the board to start basically going around and talking to a huge number of staff, so bringing somebody from outside in to do that seemed a sensible and appropriate response. The particular advantage with Robert Devereux is that he started his civil service career as a statistician—so he has a sense of what the bread and butter aspects of the operation are—but he has also had experience as a permanent secretary of running a couple of really quite complicated Whitehall Departments. So bringing that perspective, the ability to step back and a fresh eye from outside was the appropriate way for the board to exercise its responsibilities most effectively by knowing that it should not be encroaching too much personally but finding someone who could come in. I hope you agree that Robert’s report is clear-sighted and not a huge, waffly document. He has been pretty clear on that and has both diagnosis and prescription.

Just a final question from me if I may. Obviously I want to look forward, but would you agree that those sustained failures that we have seen under the ONS also represent a failure in the board’s oversight?

Having commissioned a review that found the conclusions that it did, I recognise that it would be better to have been in this position and started the turnaround earlier. It is, though, easier to make that judgment with hindsight than it is in the moment. As I say, there is always the balance of if you go in gung-ho too early and confuse the boundary between the executives and the non-executives, you create different—

Let us just pause there because you said “too early”. I just want to bring Mr Carling in because it would be enormously helpful to us to try to get a clearer handle on something that we are having difficulty nailing, which is timeline and timeframe.

Could you bring it up now? We started, witnesses, with a large hole of concern. I cannot speak for all my colleagues, but looking around and seeing some faces, the hole seems to be getting bigger, which is always worrying, although there is no fault in your answers. I just want to try to see if we can nail this down.

That is in a published assessment, so it is available publicly, to the board and ONS. We put all our assessments into the public domain.

Is it March 2020?

It is March 2020: our review of employment and job statistics.

The fact that response rates were a problem—not just in the UK but globally—shows it has clearly been around for ages. That was the rationale for commencing the work on what has ended up to be the TLFS in the long term anyway. The fact that response rates for household surveys have been declining is a relatively long-standing phenomenon. It seems to be more pronounced in the UK than elsewhere, and there can essentially be issues there around people’s trust in public institutions and their willingness to let people into their homes and talk to them. The ONS has been more vulnerable to this than some others because some of its surveys tend to be relatively long and you are having people in your homes. The fact that there was an issue about declining response rates was not unremarkable, and clearly the way in which you were also looking at the global picture mattered. In terms of what happened specifically in the summer to autumn of 2023, you had the decline in response rates accelerating in the wake of the covid pandemic for the intensification of the reasons that I have just described: people reluctant to let people into their homes and so on. But the removal of the covid boost—which seems to have been the proximate trigger of the problems in October 2023—was not something that the board was cited on. Clearly the decision around that would have been one that was operational, and it would have been much better for the board to have had a sense that a decision was being taken that had considerable risk attached to it.

Immediately on that point, not being informed of the fact that the decline was happening in the near term but being well aware that there was a challenge across the system and the world of having declining response rates.

That was lying behind the need to do the work.

It is a yes or no question, Sir Robert.

Yes, we did. You are heading back to 2021 or thereabouts.

It would have been before my time, but there would have been discussion of the need to do that and around the need for the development of the TLFS.

Obviously, I do not know exactly what he knew and at what point. As soon as it became clear that the numbers were going to have to be withdrawn for October 2023, he certainly informed me of that and noted in particular, as I recall, that there was an issue around the response rates of men aged 18 to 24, where that was showing up most concretely. The board was briefed more fully later in October. As you would see from the minutes, they said one of the issues that we were concerned about was not having been cited on the decline of response rates in the run-up to that point. Something that was consequent on that was making sure that there was then regular reporting on response rates, not just in that survey but across the other surveys at the time.

Indeed, one of the conclusions was that you wanted information on the response rates on other surveys as well. As Ed has pointed out, you see less of an issue on business surveys, although you cannot be complacent in that area. There is obviously a difference there in the sense that business surveys are mandatory in a way that household surveys are not, but none the less it is still important. A consequence of that was a better flow of information to the board on response rates across the board.

I have a couple of thoughts. I mentioned our review of economic statistics—a systematic review that we published earlier this year—in which we outlined two really pertinent requirements. One is a requirement on ONS to publish a survey recovery plan, which it did last week and we are very pleased to see that it responded quickly to that. The second is an outstanding requirement that we hope it starts to implement very soon, which is a rolling programme of quality reviews. In other words that ONS does what you have just described actually: a process of data source by data source, assuring itself on the quality of the collection, actually the exposure to the legacy IT challenges that have popped up in different places, and then reports those. That is a really important part of restoring stakeholder confidence: not just doing the work but making it publicly available.

The board certainly keenly endorsed that view through the regulation committee.

I was director general and deputy national statistician.

I was aware of a number of mechanisms. Part of my group is methods and quality, so I was aware of the efforts in methodology to counter some declining response rates. That is normal and is a part of statistical practice in making sure that where there is missingness, for example, the methodologists respond appropriately. So I was aware that there was a team of methodologists who were working very closely, some of whom were embedded in the survey team to do that. As an organisation, as we were talking about the census and the delivery of it, and whether that was an opportunity as a step change in terms of the familiarity of our brand, the purpose of surveys, the way in which we can engage with the citizen, and whether that would have a catalytic effect in improving the way in which citizens engaged with our surveys.

Mr Quigley has indicated he wants to come in, and Charlotte Cane wanted to as well.

It has been covered now.

Luke Taylor and then Ms Yang. Mr Quigley first please.

I will be very brief. You mentioned too much revolution and not enough evolution, and you have told us at quite some length what should not be happening and what not to do. Imagine I was a five-year-old—it is actually probably not a bad idea to speak to all MPs as if they are five-year-olds—

Speak for yourself, Quigley.

Explain to me what you are going to do.

Put more people with the right skills in the right place and sort out problems in data collection and underlying systems. If I were a five-year-old, I am not sure that would—

No, that will do.

Apologies for reopening a conversation that we have had at length. Ed, you spoke about how you first raised concerns in March 2020, but that must have been an accumulation of problems that you had seen occurring up until that point. So that was not the first point at which problems were occurring; that was when the first flag was raised. Sir Robert, you obviously mentioned the challenges happening at the time—obviously covid, which we are all aware of—but that was after the problems in the status quo. I would just like to understand and make sure that we are not blaming or attributing problems that had already accumulated and had been, it seems, broken by that point to an obviously incredibly impactful global event. And then actually add to this the time lag that the board did not seem to be fully aware of until late 2023, when it was already obvious by 2020 that things were going wrong.

The concern we raised in 2020 was about declining response rates on the Labour Force Survey. At that point, we reported that the survey response rate was 38%. Of course, it subsequently declined very significantly in the pandemic. The mode of collecting responses from citizens changed effectively from face to face to telephone so there were mode effects. If you do not mind, I can go into a little detail of the story. What the ONS then did—I hope this is not too methodological—was implement some re-weighting, which enabled it to draw sound statistical conclusions about employment in the economy, albeit from a lower response rate. At the same time, it started to plan for a post-pandemic world in which it would not return to a full face-to-face model but move online, and that is what has become known as the Transformed Labour Force Survey. It did that explicitly and indeed reported to the board that it was doing that because of the decline in response rates. We tracked both those developments: first, the weighting process and whether it could effectively substitute for the declining response rate, and secondly, whether the transformed, online approach was likely to deliver in time. Is the cavalry arriving over the hill likely to arrive in time? I hope that cavalry metaphor is okay for a five-year-old. In July 2023, we actually reported to say that again we were concerned that the declining response rates meant there was this increasing reliance on the weighting, and we were wondering whether that was sustainable. So it was a long process through that period where we were reporting publicly on the progress, and I suppose it is our job to be sceptical about progress because that is what we are here for. As I say, the key moment was the withdrawal of this boost in July 2023, which is when the weighting process ran out of road. ONS could no longer do credible weights to correct for the declining response rates. At that point, there were about 16% response rates and we withdrew the accreditation for those statistics immediately after that.

I just wanted to explore that a little. Yes, it is the role to be sceptical and ensure that the quality is there but there seems to be a bit of optimism bias that actually we figured this out, we can recover this. But by the point it dropped to 16% and the period of being able to weight the data to fill in the gaps ran out, you were actually too far gone to recover.

It is not us; it is ONS.

Sorry, yes. By that point, it was too far gone to recover. Is that a fair assessment?

At that high aggregate level, I would say that is a fair assessment.

Colleagues will correct me if I mischaracterise, but there are three layers of a really long-term acceleration: the UK decline in response to household survey rates, the wake of covid, and then the particular problem of the withdrawal of the covid boost, so you have those elements. As you say, that clearly resulted in the fact that the statistics had to be withdrawn and the OSR took its view. Is it too late to do anything about it? No. There has been an awful lot of effort to recover the quality of the LFS by essentially reversing some things and getting into more people asking questions, and so increasing those response rates back up. As I understand it—Emma will correct me if I am wrong—you are now back up to wave one of the LFS being almost back to its pre-pandemic levels. There is a bit further to go on waves two to five—this is a five-quarter longitudinal survey—but plans to increase the size of the field force will hopefully bring those up as well. There was an event at that point that required the numbers to be withdrawn and OSR to take the action as it did, but then resource has been pushed back in. So it is not that that was the end of the story; there is now more resource going back in, improving the quality of the LFS. The organisation has been very candid and straightforward with the stakeholders saying, “You’ve now got pretty reasonable, headline employment figures and less variability than you had had previously, but you should be careful about doing too much drilling down into those numbers and make the most of alternative sources of information as well, notably the information from real-time information from HMRC and PAYE,” which is a data source we did not have when I began a forecasting job at the OBR. It is actually an additional part of the toolkit that policymakers and forecasters have to look at. So yes, there was a clear and chronic problem turned into an acute problem, but surgery has been applied.

Mrs Rourke, first can I say that I very much admire your stepping into the breach, because it is a tremendously significant responsibility and task that you are taking on right now? I want to return to Mr Carling’s questions about when people became aware of issues and the testimony of the former National Statistician, Sir Ian Diamond, to the Treasury Committee. In February, when the Treasury Committee was asking him about when he became aware of the problems with response rates in the Labour Force Survey in 2023 and whether in hindsight he should have been able to know these problems earlier, he responded with, “I cannot really answer that question; you would need to ask Darren Morgan when he first became aware.” Darren Morgan was the director of economic statistics at the time. I can only conclude from his answer that either there is a culture at the ONS of bad news not being fed up through the organisation, a problem with the senior leadership taking full responsibility for the failings of the organisation, or potentially both those things. Which do you think it is?

If I may start by referring to Sir Robert Devereux’s report, he observes there being limitations within the workforce of people speaking up when there are concerns. Through the TLFS lessons learned, that was something that we had explored, discussed and put a number of things in place, which I am very happy to share with the Committee if that is of interest to you.

Yes, please.

May I just launch in to discuss it?

Launch away. We have hours to go. We can send out for sandwiches and cream teas if necessary. Yes, crack on.

There is a sequence of signals, if you like, emerging from the workforce. I might hear that from my direct reports or through some other mechanisms that I instilled into my working practice, for example, offering face-to-face time with colleagues where they could book a time to talk to me about something that mattered to them. So from quite an ovalised formal virtual opportunity all the way through to something a bit more personal, designed to feel safe, and an opportunity for somebody to speak in a very open and candid way. We were also hearing signals through some grades 6 and 7—the layer below the senior civil service, our most senior managers—about some frustrations. As a result of that, we created a number of programmes where we reached out very deliberately and consistently to that community. We traditionally had a particular focus on the senior civil service. The leadership of the system starts from the top, so we were having away days and spending time together. Certainly in my own area, I brought in external facilitation, making sure that I was challenged as well as other people. But for some time, we had not attended to that community, which is really fundamental in building that culture that we know that we need. We have had a range of opt-in sessions that are face-to-face across all our geographies, inviting those senior managers to come and talk to us about some issues and how they lead and manage change. I and other directors general would lead series and talk about some concerns that people were facing and how we could best support them.

In your experience of the ONS’s work, do you feel that culture of challenge once existed and has deteriorated over time or that you are building it up from scratch now?

I do not feel that we are building it up from scratch. A lot of work was done locally: we had people action groups and a number of different mechanisms to really respond to the data that we were getting. Our challenge is consistency and sufficiency. One of my observations and learnings is about the evaluation. In some cases, we would deliver some interventions and get very positive feedback, but that positive feedback was not often followed up several months later about how that had been internalised. They had had a really positive experience with myself or another director general, and I would speak very candidly about mistakes I have made, but we did not follow that up about internalisation. That is something that I would very much like us to see because it is integral to our future success.

Are you saying that there was the ability for your senior civil servants to make representations and that that would be heard but not necessarily acted on at the senior level, and that is what you are trying to change?

I do not think it is just in the senior civil service, if I may say. We are an organisation that should strive to see leadership at all levels, but I recognise that the civil service and ONS are inherently very hierarchical organisations. In order for the senior civil service to role model challenge and successful speaking truth to power, that needs to be in place first.

Can I just come in on that? From my perspective as somebody who oversees and reviews the statistical outputs of ONS, I definitely recognise everything that Emma has said. Indeed, in the review that we published, we reported that early warnings of problems are not always welcomed within ONS. But there is another aspect to this that I would really like to land with you as a Committee, and that is openness to external challenge, feedback and users. It has been a consistent theme of our reviews that ONS sometimes tends to be defensive or downplay external concerns, and that is a very significant weakness in the way ONS has engaged with the world in the past. What is really so striking is that this year, there have been two things that have emerged from ONS that have broken that cycle. The first is all the things that Emma has overseen, for example, publishing; their survey recovery plan and overall plan for economic statistics are clearly more open. I just think if you read the foreword that has been written, you will see that sense of openness. But there is also something that has not come up in our discussions so far this morning: the Statistics Assembly, which exemplified an openness to external users so much so that ONS did not really do much presenting at it. It adheres to something that one of our board members, David Spiegelhalter, always says: that the first rule of communication is to shut up and listen. The Statistics Assembly exemplified that. So I am seeing more openness, but—we have not talked about it so far today—part of this cultural backstory is not just the internal openness that Emma outlined; what I have seen in my work is a tendency to be slightly dismissive of external input and challenge.

Can we just pause while I get out my water paints and attempt to draw a five-year-old’s picture of ONS in modern Britain, with some slightly random observations but hopefully connected? A huge amount of weight has been placed on the efficacy and usefulness of Devereux, but Devereux is a signal of the failure of the organisation. It is much better if we see it as a response to problems rather than a heralder to better ways of doing things, because better suggests things being done were being done well, when clearly they were not. Some of this is legislative knots, which can be undone, but quite a lot of the problems seem to suggest to me an organisation which is independent of Government but dependent upon the mindset of the leadership, which can also make it unaccountable to the court of public opinion. A few people have mentioned Sir Ian Diamond. He was perhaps a great innovator so he took his eye off the basic plumbing of the organisation to pursue whizzy data-led things; but his inability to face into, report or think innovatively about how to deal with a changed scenario and landscape of getting data upon which the ONS could come up with some statistics, does not suggest that he was a particularly good innovator. He demonstrated no innovation in that sphere from issues going back, as we have heard, quite a considerable period of time. Sir Robert, you have told us very clearly and with good justification what the board was not there to do. Frankly I do not have the foggiest idea of what the board was supposed to be doing, and whether or not it did it. What I find slightly peculiar to understand is that in a modern, social media, sunlight-is-the-best-disinfectant type approach to life by which Ian Diamond appeared to have run this organisation, he was a sort of hybrid of a Medici prince and Blofeld, and seemed unwilling or uninterested in anything anybody had to say. He managed to pull wool over the eyes of Ministers and the board, saying that everything was hunky-dory and innovation was going to be the great salvation to all our prayers. That probably explains why—I know I am mixing my metaphors horribly—when the emperor was found to be naked and not clothed in regal purple, ill health meant that he had to leave the ONS PDQ. Mrs Rourke, you have indicated changes and they are welcome, but basic organisational change is now needed, is it not? A change to the relationship between the ONS and Government, better clarity on what the board should be doing and a fundamental dismantling and rebuilding of the HR function. If somebody cannot blow a whistle within an organisation and have it dealt with in an appropriate way, heaven help us. It appears not to have been run in any way, shape or form on modern expectations of governance, but a sort of a slightly boosterish chumocracy, which has not served the organisation very well. I may have used non-statistical language in that analysis, but in broad terms have I said anything upon which any of the three of you could disagree?

In terms of the what the board is supposed to be there to do, as you know it is a complicated and unusual structure as set out in legislation. But a key part of it is providing support and challenge to the institutions in delivering what they are supposed to.

But it did not provide challenge, did it? It took what it was given, sucked it up and accepted it.

We had concerns, we heard concerns, we responded to those progressively. We ended up commissioning an unusual review to look at this in a drains-up way and to put in place the plans that are now going to be pursued looking forward. To recognise your issue, part of that—

But not with any appetite of urgency: Mr Carling has managed to identify these problems going back over five years. The Lievesley Review was undertaken, and I hear what Mr Humpherson—or it might have been you, Sir Robert, forgive me—said: the best thing about communication is to stop talking and listen. Nobody seemed to do anything with the Lievesley Review. It was a tick-box exercise, “Oh, thank you very much indeed, that is marvellous, and then we will crack on in our usual way.” This requires Cabinet Office to go back and look at the legislation, the reporting lines and this issue in a proper way between, in effect, independence not meaning unaccountability, does it not? Because unless or until that is done, my five-year-old’s take on the situation is that a new Blofeld—Mrs Rourke is not a new Blofeld or anything else—could take the organisation in exactly the same direction that Sir Ian Diamond was allowed to, completely unfettered and without any check or balance upon his boosterish desire for very expensive, cost-ineffective expenditure on pretty pointless IT and data solutions.

The Cabinet Office and the board both accept the third of Sir Robert’s recommendations, that it is appropriate to look at the appropriate governance of the ONS and across the system. As I say, having the system-wide responsibilities and the different elements is an unusual structure; whether that is the best way of providing support and challenge to the ONS specifically will obviously be an important part. I am sure the Cabinet Office, and certainly we, will be interested in the Committee’s views on how best to approach that. At the end of the day you need to get to a situation in which you have a structure, as you say, that has appropriate lines of accountability so that it is not a situation where you overcorrect and have a group of non-executives trying to run an organisation from the side.

But there is a difference between trying to run and robustly scrutinising and questioning. That is not trying to take over the administration of an organisation; it is having people confident in their understanding and who are able to ask the right questions and to keep on asking them until they get a satisfactory answer. That is not running the organisation, that is doing the job of scrutinising.

No. Take what happened with the TLFS, for example, and the way in which the board interacted with the Executive in the consideration of whether we were ready to jump horses to the new way of doing the labour market statistics in September of last year. If you look at what the board said in the minutes, we absolutely emphasised that the decision should be based on data and evidence, not on a particular person’s desire to get to a particular point at a particular moment. We wanted to make sure that we were getting a full set of information and that the board and the Executive were taking full notice of what the Bank of England and the Treasury’s concerns may have been, what we were hearing from independent sources of statistical advice and the expert group that was set up around that. That is an example of the board saying that we needed to make robust decisions based on evidence. The consequence of that is that we are now at a more sensible timeframe for the TLFS move. But as you say, it is absolutely right that you have that willingness to ask questions in a situation where the flow of information is not as good as it can be, and in which there are mixed signals from within the organisation. It is not straightforward, but I do not for a minute step back from your critique of what the proper role of the body should be.

Just pausing there and moving to Mrs Rourke: Devereux suggests a splitting of the roles, but temporarily. Could you please give us a couple of bullet points on the merits or otherwise of splitting the roles? If there is merit in splitting the roles, should it be temporary? My understanding is that it would require legislative change: apart from the logistics, could you give the headlines first? I am dealing in the sort of bullet points.

I will perhaps respond in the context both as a member of the Executive Committee but also having done the job for six weeks, and I agree with the recommendation.

Should it be temporary?

I am going to reserve judgment on that on the basis that the decision to separate the roles now is in response to the context in which the organisation has found itself, which may be very different in two to three years.

Yes, I agree with you on that. However my understanding is—I do not know if you have taken legal advice on this or not—the splitting of the roles will require legislative changes. But we do not want to be doing the legislation hokey-cokey of it is one, it is two, it is one again.

Indeed, yes. The merits of the model in its revised form with splitting it out means that you have the benefit of the operational leader. At its core, ONS is an operational organisation; it is about the production of high quality statistics and the delivery of high quality surveys. There is a lot of merit in having somebody who knows how to deliver, assure that delivery and engage with users in a way that protects both of those things. In terms of the national statistician role, in the separation of the role you would see real technical expertise. For example, you referred to Denise Lievesley’s review: she has a critique on the leadership of methodology and suggests a senior methodology leader. Having the bandwidth at that level to attend not only to leadership of methodology domestically but internationally, across the Government’s statistical service and currently as leader of the analysis function, would be welcome; however—

If there is a “however” we had better have it. I thought we had reached a full stop not a semi-colon, but I am always interested in a “however”.

The thing that I really support in it is why I am cautious about endorsing it in the longer term; recognising the legislative issues and that we have some very talented statisticians in ONS. I agree with Sir Robert Devereaux that with the right career paths and support, seconding them out of the organisation and giving them the right input, somebody who can balance both those things is entirely possible but only if we have addressed some of those systemic issues we have discussed here. But, based on the talent that I observe in the organisation currently, it is possible.

Have you sought legal advice to confirm whether or not legislative change will be required in order to split the roles, as we understand?

That has been done through Sir Robert, as the recruiting—

So has legal advice been taken?

The Cabinet Office is in charge of the appointment processes; for example the National Statistician is a Crown appointment, and if the Prime Minister approves the idea of having a permanent secretary, then that is a process there as well. The fact that the Cabinet Office has accepted the recommendation and is proceeding to those appointments clearly indicates that it believes that is legally possible in the current circumstances. I am not wishing to put words into Sir Robert’s mouth, but basically I think he was saying that he would not exclude the possibility that you could reunite the roles in the longer term, not necessarily that that was his first expectation.

It was the other way: his starting point was that it should be temporary.

Yes. In the past, the filling of this role has always been a challenge on the grounds that, as he rightly says, you are looking for a really very wide set of skills: somebody who can be the adviser to Government and the board on matters of statistical importance, somebody who is a chief executive of an organisation with thousands of people in it, and somebody who is a permanent secretary in the Whitehall structures as well. The chances of finding all those things in one person is not straightforward and, arguably, has become more difficult over time. I entirely support the recommendation to do this in the short-term and, certainly on the operational side, the sooner that can be done the better.

Returning to our questions, I know you have a Treasury question, John, so I am going to bring you in.

Good morning to the witnesses. I suspect my question is temporarily taking this discussion on a slightly different angle. Sir Robert, you are chair of the UK Statistics Authority. Does the authority have a consistent approach in all parts of the UK in terms of Scotland, Wales and Northern Ireland?

We have regular conversations and there are mechanisms, for example for the Inter Administration Committee—I think I have the title right—which is a regular discussion between the national statisticians and the lead statisticians in the nations. I am obviously in regular contact with the heads of the statistical functions in the different parts of the UK as and when they want to be. For example, if you look at the census recommendation that the board endorsed relatively recently, that is a recommendation for England and Wales, but obviously it was discussed a great deal with Scotland and Northern Ireland and they put out recommendations at the same time. That is quite a good example of making sure that the UK-wide perspective and the perspectives and particular needs of the nations are taken into account as and when the board may have to make a decision, because ONS’s function is more around England and Wales. So yes, that is an important part of the role.

Can I just step in there? By far the greater bulk of the work that I lead is covering the assessment of statistics not produced by ONS but by ministerially led Departments at the UK level in Scotland, Wales and Northern Ireland. We have accredited 800 statistics, of which only 15% are ONS; the remainder are from the rest of the public sector. We do a lot of work with Scottish producers, the Scottish Government, Public Health Scotland and so on, similarly for Wales and Northern Ireland. I am therefore confident that data is being collected on a consistent basis across all parts of the UK.

Are you therefore confident that the data is being collected on a consistent basis across all parts of the UK?

There is a separate issue there around the comparability and coherence of data across the UK.

That is my point. When I was trying to compare the levels of child poverty in Liverpool to Dundee, for example, that became an almost impossible task because of the lack of availability of data from Scotland. Do you agree that that is a problem?

I may come to that. Ed has recently published an excellent report on this subject. There are two concepts it is useful to distinguish: ideally you would like data that is materially comparable so you are essentially looking at like for like, and that you see something where the difference in the numbers is giving you a good reflection of the difference in reality. That is obviously a lot easier to do if you have a survey that is set up nationally, where you have the same questions being asked of the same people in all places. You will quite properly have different parts of the country, different parts of the UK, with different policies and approaches as a result of devolution, and collecting the data that is appropriate to those things is more difficult. So the issue there is coherence, which is basically being able to understand the differences between the two. One of the things that the Inter Administration Committee is about is identifying areas where you want to try to improve coherence and understanding across those, and there is an active programme of that. There are things you could do to make that better and Ed has some suggestions of that around, for example, reviewing the legal frameworks. With administrative data, the crucial thing is that if you have different parts of the country that are collecting data in a way that delivers value for money for the way in which they are delivering those services. If you think that there is an additional value in collecting data that allows a more straightforward comparison, who pays for that? That is another point that Ed makes: if you want to have data collected in a way that goes above and beyond what suits the individual policymakers and service deliverers in particular parts of the country, that is an additional cost.

But surely it should not be a matter of what suits the policymakers and the decision makers. You rightly make the point that we have devolved Parliaments in this nation, but we are still part of the United Kingdom. When it comes to assessing how different parts of the United Kingdom are performing, surely we need a consistency of data to allow that comparison to be made. It should not be a matter that the policymakers, for whatever reason, decide they do not want their policies to be exposed for failing, and not to have that consistency of data.

Yes, but obviously if you are the Scottish Government with a budget to manage, you are looking to collect the data that you think you need to deliver the policies that you are pursuing to the best of your ability. If there is an additional cross-UK externality to having better information, then there is a question of, on what basis do you collect and who pays for that?

Just to go back to where I started, surely you would agree it is wrong that as a UK Member of Parliament I cannot get data that allows me to compare levels of child poverty between Liverpool and Dundee?

It is clearly not desirable for that to be the case, but we are in a devolution framework in which people are making their own decisions in different parts of the UK. There is a recognition of the importance of pursuing this in different areas and I do not know whether child poverty is a specific one at the moment.

That was the example I used, but there are other examples that could easily be pulled out.

I might add to what Sir Robert was saying in terms of the work we are doing with the chief statisticians in the devolved Governments. We have done an intensive piece of work on health and have collectively published a series of comparative measures around health; our next phase is population statistics. Unfortunately, within ONS we have had to reduce our investment in coherence and supporting devolved Governments in order to support economic statistics. It is regrettable, but these are the sort of hard decisions that we have had to make. However, we have had some very constructive conversations with the chief statisticians in the devolved Governments about the demographic index and the future of population statistics. Using the resources that we have had for our programmes, for example, this has allowed us to develop a mechanism by which devolved Governments can take their own data, their methods with facilitated access to some of their data, so that there is that ownership and that consistency, where we have had the resources and the mechanisms to pursue coherence. This is important in the context of child poverty because you need to be able to access that granular data and really quite precisely understand where there are problems, both in geographical and other socioeconomic terms. Our focus has been on where we can leverage the investment that we have had to overcome some of the barriers that Sir Robert was talking about. I have already spoken with the chief statisticians about Ed’s report last week, and we have agreed that the Inter Administration Committee, which I chair, will focus very much on those lessons and what we can do practically with what we have available to us.

In terms of the last census we had, there was a separate census in Scotland which was delayed due to covid, so that presented challenges. Can you reassure me that the next census will be done on a UK-wide basis, and there will be no opt outs of any of the devolved nations?

Our recommendation, which is currently being considered by Ministers, is that we run the census in 2031 for England and Wales, with separate agreements with Scotland and Northern Ireland. Subject to finances, we will have one shared platform where we can share our data, and we will develop a model that is consistent in terms of timing but allows for local discretion for particular questions that that Government might be particularly interested in pursuing.

Ms Rourke, this is a question for you initially. The annual report for 2023-24 indicates that a new post—director of operations—was created a couple of years ago. Why was it felt necessary to create that post?

That was a judgment made by the chief executive at the time, following the departure of the second permanent secretary. There was a view that the oversight of operations at that point needed a focus on people, getting the workforce ready, and the operational oversight of surveys. So, as is civil service process, there was a current role that was expanded to accommodate that.

If I might just add briefly: it was explained to us there was a particular sense that improving the performance of the survey field force was a people issue, and therefore the chief executive saw synergy in that area in particular.

You said normal civil service procedure, so would it not normally be necessary to run a recruitment process for a new role?

The decisions on pay and the structure of that were consistent with what the Cabinet Office said was appropriate for the accounting officer to be willing to do, and I specifically checked this.

Did you not feel it would be useful to do a recruitment process to make sure you got the best person for that role?

In an ideal world you have a competition with a clearly set out job description on that basis, justifying and explaining the nature of the new role. It was more—I do not know what the jargon would be—a level transfer in the organisation. That issue was in my mind, as I say, and I checked with the Cabinet Office that the appropriate approvals had been sought; the conclusion was that it was appropriate for the accounting officer to take that decision.

You said that in an ideal world it would have been good to have a recruitment process, and then went on to say “and have a job description.” Are you saying there was not a job description?

Not a job description; a justification for the nature of the role and the associated package that goes with that.

But you would have had to have that justification regardless if you were having a recruitment process, would you not? Are you saying there was not a full justification package done?

I do not know the precise details. Clearly there was a full discussion at the executive committee of the appropriateness of this role and the suggested salary package that went with it, as I was told. I checked with the Cabinet Office ahead of the appropriate discussion at our level as to whether the appropriate approvals and process had been followed.

So a new senior role can be created without the board seeing the full justification.

The answer to that question is yes, from what I have heard.

With a role of that sort, yes. This was an operational decision as to structuring the team, so we did not see a full description of that role before it was put in place. But as I say—

Sorry; I have never worked anywhere, public sector or anywhere else, where a new senior role can be created without the board’s approval. I am amazed and shocked.

Or when somebody is appointed and one does not know what the skill requirements are.

As I say, the chief executive explained to us that he thought the role was appropriately structured in this way and that there had been careful discussion that this was appropriate at the executive committee. When I was told, I felt it was important to check directly with the Cabinet Office that the appropriate approval had been given for that. In an ideal world you would have had a fuller process around that, but I was told that, at the end of the day, this was a decision that the accounting officer could take.

This brings us back to what the role of the board is.

This brings us back to that fundamental question of who guards the guards. Something tells me that the ONS is now falling into the realm of Schleswig-Holstein, where only three people knew the answer to the question.

Can I just ask one further question on that?

Please do.

You gave this extra responsibility to someone who was responsible for HR because you felt it was to do with people. But we heard earlier that there seems to have been a culture of not wanting, or not feeling able, to pass problems up the chain, or maybe people up the chain not hearing it. So there clearly was a really systemic HR issue. Did you not feel your HR person should have been consulting on that rather than giving them extra workload?

That is a perfectly legitimate question to raise, but we asked the chief executive at the time why he thought this was the appropriate thing to do. He said that that was the right thing to do, particularly given the synergies between the importance of people policy and surveys. As I say, I did take this to check out and personally pursue that the appropriate process had been followed. But any incoming permanent secretary or senior leader, depending on how you define it, is obviously going to want to look across the scope of the senior roles and see what makes sense for the longer term.

I would have thought people in a statistics body would be very keen to make sure the statistics were accurate. Mrs Rourke, have you managed to get any sense of why people were so wary of passing on concerns?

We have a number of reports from some of the deeper dives that we had done prior to Sir Robert Devereux’s review.

May we see them?

Yes, I would be delighted to write to you with that detail. People reported feeling very hurried in their workload, and there was a focus on delivery rather than a pause and reflection. I had a very interesting conversation with somebody who provides oversight to some of our quality assurance work and this speaks in part to what Ed talked about on defensiveness: when there is an error there is quite often a posture adopted which is, “Why is this right?” Rather than, “Why might it be wrong?” So the framing of the questions has been flawed at times. There is also something we have discussed which speaks to it being rushed: the fact that quite often the final challenge session—called the curiosity sessions in some teams—is often at the end of the process and not built in all the way through. So that sense of being rushed and of, “Well, I do not really have enough time to talk about this in a deep and meaningful way” is part of it. There have been times when people have reported getting a very robust response, which also speaks to the MI issue that we have to address. The management information and inadequacy of the warning signs meant that quite often when there was an issue, senior leaders were not warned and so the response was often one of surprise. That came across as quite a robust, “Why did I not know?” That was clearly inappropriate at that time. So through our discussions and some of our work, particularly around the Grade 6 and Grade 7 conversations that I referred to earlier, there are a number of factors giving us a sense of what needs to change .

Are you using “robust” where you should be using “aggressive” in terms of the response?

Some people have interpreted that as being aggressive. I cannot speak to the totality of the experience, but I certainly have witnessed very robust challenges.

What have you been able to put in place to address these people feeling rushed and hurried to make sure there is time for checking and reviewing throughout the system?

A lot of this speaks to the investment, the economic statistics plan, which is providing additional resource into those teams that provide the bandwidth. Part of the governance work is also ensuring that our assurance framework is proportionate and asks the right questions in the right tone and place. So yes, it is about bandwidth and pace. We see people putting their hands up now and see evidence of productionisation stopping as a result of concern where we might withdraw or delay an output, and we report near misses as well. There are a range of mechanisms in place that we need to ensure the consistency and sufficiency of those measures.

You have pulled in more resource, which sounds good, but you have presumably taken that from elsewhere. We heard earlier that some has come from looking at getting comparative statistics between the devolved Administrations. Where else have you had to take resource from, and what are you doing to make sure that we do not get new problems popping up in those other places?

We have thought very carefully about this. The coherence point is something that I had to reduce drastically in my area as a part of business planning for this year. That was done at the beginning of the year, and I stopped a range of other activities including work on longitudinal data and productionisation of some newer models for population statistics in order to be able to support surveys and economic statistics. We also cut back quite drastically in the health space as well. So that was done at the beginning. What we have done more recently has been to revisit some of those decisions. The integrated data system programme is one where we have dramatically reduced the scope in order to liberate and deploy skill and assets to support economic statistics. We are also redeploying really high-quality analytical capability that is in our analytical hub. This hub was created several years ago to face into the Cabinet Office and support a lot of cross-cutting analytical questions. I do not think we have finished cutting back yet. My personal view is this is just the first phase of within year changes. We will update our strategic business plan and reflect the changes that we have made so far so that people can see the totality of the impact and we have complete transparency in that space. It is important because people are disappointed in what we are stopping, but that is part of the course in terms of shoring up those economic statistics.

I am conscious of time and conscious of the temperature. I am rather regretting that I did not take my own advice and remove my coat at the start of the meeting, but there we are. We still have a lot of territory to cover. I am giving some thought, but look to the Committee for are nod or a not, to adjourning. Sam, you have one set of questions that we could get through by about 12 pm, but I feel we may have to adjourn this session and reconvene next week to pick up where we left off. I am afraid, witnesses, that may mean that you have to rejig your diaries.

I am supposed to be appearing before you again next week anyway with another hat on, so we will avoid one diary clash. Chair You can keep the same hat that you have on now, but that might be better. I am entirely in your hands whether we turn to Mr Carling for his questions or whether we adjourn now and pick up—

Let us do Sam Carling’s question first, and we may very well do yours, Michelle, because I know yours is quite a quick one, but again it is on staffing appointments and processes.

On the specifics of the appropriateness of the defining of that role and the salary that went with it, I checked with the Cabinet Office that this was something that was appropriate and possible.

As I say, it would have been better had that been done in the more formal way that you have described.

The view from the Cabinet Office was that it would be ideal to have more competition and more of the specific, detailed justification of the nature of the new role.

Yes.

That said, it would have shared the view that that would have been the better thing to do, but at the end of the day it was within the rights of the accounting officer to proceed on that basis and therefore it was not something that was to be blocked.

In these circumstances you need to apply for approval from the departmental permanent secretary and the cross-Whitehall head of profession; this was initially a job that was defined temporarily and then permanently. In both cases the notification was made; the Cabinet Office concluded that in an ideal world you would have had that fuller process, but at the end of the day it was something that was in the powers of the accounting officer to do, and therefore not something that would have been appropriate for us to block.

I will confirm the exact date of his appointment to you, if I may, but it is about nine months.

That was in 2024, so roughly a year ago.

It was at the beginning of the year, in the spring.

Yes, there was a gap.

I apologise; I was not having conversations with the chief executive at the time about why certain decisions were made.

My understanding is that the person holding that role is on leave of absence, and that Peter was asked to cover the role of data capability in March 2024.

As I understand it, she remains on leave of absence at the moment.

That was before my time. I assumed that that seemed to be appropriate for the role or for the senior leadership team at that time. When you get to 2020—

Let us pause there. Mrs Rourke, maybe you could very kindly get somebody to dig out the relevant paperwork which led to that, and let the Committee have it?

Yes.

That happened in 2023, so you will be aware this was obviously at a time when you had quite a proliferation in second permanent secretary roles across Whitehall, post-Brexit and through covid. So in general there was a desire that when these roles became vacant or somebody moved out of them, you would review this and remove the roles at that level, if possible. At a more local level, the chief executive at the time took the view that not replacing this role would allow resource to be deployed elsewhere. There were restrictions on the ability of other parts of the organisation to hire people, and it was important that the senior team not be immune from that need for economy.

I have a number. I want to say eight, but will you allow me to go and just double-check that outside, please?

It is certainly not a majority.

Sir Robert, your board minutes for April 2024 suggest that Mr Nigel Green was appointed as director general in charge of IDS, but we cannot find any record of that appointment.

As I understand it, he was brought in, as you say, essentially to have a key role in the IDS. This is contractual and not a formal role that line manages people. So it is at a director general level, but it would be helpful to be more transparent about that in the organogram.

Was he a director general or not?

He is a short-term contractor at director general level. I am not quite sure what the formal legal term is.

He was brought in by the chief exec to strengthen the delivery leadership of the IDS delivery.

Did he have all the powers, influence and functions of a director general?

He reported in to a director general, but I understand that he was given the title in order for him to have effective bilaterals with other government departments at a similar level.

Either he was a director general or he was not.

He did not have the line management responsibility that a director general would do.

But he had the perks of being able to access all the different departments, so he was half a director general, not a full one?

I would not describe it as perks: that was in order to facilitate the conversations that he would need to have across the rest of Government.

Are you saying he only had half a role of a director general? Because you are saying he did not have the managerial role a director general would normally have, but he was called a director general?

Was he remunerated as a director general?

The reason he was brought in was because he had the relevant digital programme knowledge across Whitehall so, given the specific issues that the Integrated Data Service was going through at that stage, he was the right person.

What was his title?

It was director general.

So he was director general?

Yes.

But he did not have all the responsibilities that a director general would have. Was he remunerated at a director general level?

We can supply you with the salary; I have not been privy to that. He is with us on a fixed-term part-time contract specifically in the area of technical delivery.

When a department or an organisation, an arm’s length body, call it what you will, temporarily brings in some expert, skilled resource to assist with a specific project, is it customary that they accrue the title director general? Surely they are just consultants

I do not know how widespread this practice was, but both internally and from the Cabinet Office’s point of view, this was an opportunity to make sure that there was appropriate input into a project that needed more expert input of that sort at that level.

I am going to ask the final question and then if the Committee could just stay behind. I want to finish at about 11.55 am and then I will just ask people to clear the room as quickly as possible so we can just have a few moments before we depart. I am going to remind all witnesses that you are protected by Parliamentary privilege when evidence is given. Mr Humpherson, you and I have met on a few occasions and you strike me as a man of integrity and the utmost probity; I am not saying that Sir Robert and Emma are not, so nobody should draw that erroneous conclusion. Your facial expression and your body language this morning has given me a very clear message of unease, disquiet, discomfort and concern from a professional analysis capacity that a variety of people were ringing alarm bells or raising concerns at various times and at various volumes. Please feel free to correct me if my impression is wrong. I think you have concerns about how this organisation was run and administered; that you have concerns of the damage that it does to the reputation of statistical gathering in and across the United Kingdom and our international reputation in that arena. Is that a fair assessment?

You ventriloquise me very well.

That is an encouraging but also worrying place to pause this session. We may pick up on that in the second session. I would just say for the record, as Chairman, this Committee does not expect any witnesses to be lent on over the next few days with regards to what they may or may not say. We will adjourn; colleagues and witnesses, thank you for your time this morning. I apologise for the blistering airless heat but we have all remained cool, calm and collected none the less. Thank you very much indeed. We are adjourned.