Public Administration and Constitutional Affairs Committee

Oral evidence: UK Statistics Authority, HC 978

Tuesday 26 April 2016

Ordered by the House of Commons to be published on 26 April 2016

Watch the meeting

Members present: Mr Bernard Jenkin (Chair), Ronnie Cowan, Paul Flynn, Mrs Cheryl Gillan, Kate Hoey, Kelvin Hopkins, Mr David Jones, Mr Andrew Turner

Questions 1 - 82

Examination of Witnesses

Witnesses: Sir Andrew Dilnot CBE, Chair, UK Statistics Authority, and John Pullinger, National Statistician, gave evidence.

Q1   Chair: Can I welcome our two witnesses this morning to this session on the accountability of UKSA, the UK Statistics Authority, and could I ask each of our witnesses to identify themselves for the record, please?

Sir Andrew Dilnot: I am Andrew Dilnot. I am the Chairman of the UK Statistics Authority board.

John Pullinger: I am John Pullinger, National Statistician.

Q2   Chair: Our first few questions are very much directed at you, Chairman, rather than the National Statistician. We have a lot of questions to get through. We are going to ask quite crisp, short questions and if you can keep the answers crisp I will be very grateful. If either my colleagues or you wander off, I will pull you back. We have had the Bean review and you have established a management team, your newly established management team, and you are facing a new Select Committee. In the light of all this, what kind of relationship do you think should exist between this Committee and UKSA to serve the public interest?

Sir Andrew Dilnot: I think the short answer to that is as close a one as this Committee can manage. It is very important to the authority and its board—and I think perhaps particularly its non-executive members—that the board is independent of Government but that it is answerable to Parliament. We would hope to continue to have a close relationship. I am aware that in the last Parliament the predecessor Committee to this Committee was able to spend quite a lot of time on statistics. This Committee now has a wider set of responsibilities, but I would have thought that one obvious area for focus in the next year or two, which will be a focus for the board, is the implementation of the Bean review. We welcomed the Bean review. We thought it was encouraging us to go in directions that were consistent with the strategy we had previously published, but it seems to us that that is going to be a core part of our activities for the coming years.

Q3   Chair: How much of the Bean review do you feel you had heard before, perhaps from this Committee?

Sir Andrew Dilnot: I will not attempt a statistical estimate, but some of the core points we had heard before from this Committee, from the strategy that we ourselves published in October 2014 and from other commentators. That does not mean in any way that we think that the Bean review was redundant. Far from it, I think building up a sense of momentum and a sense of urgency here is crucial.

Q4   Chair: It is that sense of urgency and momentum, which I think the Bean review rather exposes, which has not been as evident as it should be. I recall one witness saying that you have had a number of high-level discussions about statistics and stuff, but it does not seem to get carried through into the execution throughout the organisation. How fair do you think that is?

Sir Andrew Dilnot: I am not sure which particular reference that is, but it would certainly be fair to say that the board as a whole would love to have seen even more rapid progress than there has been. The Bean review itself recognised there had been progress in a number of areas but wanted to see further progress. That is certainly our position as a board, and I think as an executive team, that we want to keep the speed up.

Q5   Chair: Just to be clear, Professor Bean says, “One former board member said that at least in the past actions had sometimes been agreed to at board level but not followed through on”.

Sir Andrew Dilnot: I think that is true. But it is important to note that that witness—who I think is anonymous—said, “at least in the past”. Our view is that with the governance reforms that we have implemented in the last two or three years, which in part flowed from the predecessor Committee to this Committee’s report in 2013, and with the new management structure that we have put in place in the last year or two, we are now much better placed to make sure that the decisions made by the organisation’s board do get followed through on.

Q6   Chair: When do you think we will be able to stop making these points?

Sir Andrew Dilnot: I hope very soon. I think it is possible to point to a number of changes that have occurred in the last little while that are promising. There is still a lot to do. There is still a great deal to be done on the use of administrative data, on modernising technology, and on shifting the organisation from being an organisation that has focused on the collection and production of statistics to being an organisation that focuses on analysis and added value. I don’t think that we will reach nirvana soon, but I think in process terms we would be disappointed if it were the case that decisions made now were not being followed through on.

Q7   Chair: Without going too negative on you, as some Committees do, how can we assist the pace of your improvement?

Sir Andrew Dilnot: I think knowing that there is external scrutiny is always a wonderful form of keeping one’s feet to the fire, so I would have thought a natural thing to think about is our business plan—which we published just last week—in which we describe the ways in which we plan to transform the business over the next few years and the ways in which that relates to the Bean review. It seems to me that it would be an absolutely natural thing for this Committee to hold us accountable for that.

Q8   Chair: It is easier to hold you accountable if you are as transparent as you can be about what is happening, and that requires you to be internally transparent and to encourage internal transparency. For example, we recommended in our report that the non-executive directors should publish the minutes of the non-executive meetings where I guess you discuss these things and monitor progress. Why have you not done that?

Sir Andrew Dilnot: I am pretty sure that the minutes of our board do include, at the beginning of each minute, a description of the subjects that have been discussed.

Q9   Chair: Would it help if you published more comprehensive minutes that gave the non-executives’ views about whether things were being done or not? Then we could hold you to account on this information. It is difficult to hold you accountable without more information.

Sir Andrew Dilnot: My view on that is the non-executive scrutiny, of whether things are being done or not, ought to occur in the full open session and be fully minuted. To take a case in point, the website—which Sir Charlie to my great delight describes as a massive improvement or greatly improved—about two years ago the board as a whole took the view that where we were was simply not adequate and we needed to stop and start again afresh. I am pretty sure that those views were expressed and minuted.

Q10   Chair: What other role could your non-executives play in the scrutiny role?

Sir Andrew Dilnot: In the last year or so, one of the things that the non-execs have reflected on quite a lot is the need to have very clear milestones and measurable performance targets. You will notice in the business plan, which we published last week, there are significantly more quantitative targets for levels of achievement of certain things and dates. I would say that, in the last nine months or so, we have also moved to requiring at each board meeting a single page—which we are referring to, possibly slightly inelegantly, as transformation on a page—which lists the various things that we are expecting to happen in the next year or two, next to specific dates for when they should be happening. We are trying to build that up and I think that is something that we need to do more and more.

Q11   Chair: Can I suggest that transparency is probably more about us understanding how you talk about things internally, how you discuss frustrations and things going wrong and how you learn from what is going wrong, rather than just meeting numerical targets? That is why I am suggesting that there should be much more openness about those things that cannot be quantified, but we know when they feel right and we know when things are feeling better. We want to see evidence that you are feeling things are going better, not just presenting us with numerical targets. Does that make sense to you?

Sir Andrew Dilnot: I think it does. I might turn to John in a moment just to—

Chair: No, I do not want to talk to John at the moment.

Sir Andrew Dilnot: Well, because John is responsible for delivering the secretariat that produces my minutes and one of the things that I—

Chair: No, you are the Chairman. You are responsible.

Sir Andrew Dilnot: I will tell you what I will do. I have a board meeting on Friday. We will discuss this at the board meeting on Friday and see if we can come up with some suggestions for ways of describing how we are feeling, which might help the Committee get a sense of that. Something else that I am quite keen to see, if we might do, is to invite members of the Committee to come along to one of our board meetings.

Q12   Chair: Yes, but if you just wrote to us every now and then, about things you are finding difficult and how you are addressing them, that would be open and transparent. It is not in order to create an adversarial conversation with us. We are not worried about hearing about your problems. What we are worried about is hearing that you have problems but you cannot talk about them and you are not dealing with them.

Sir Andrew Dilnot: I would certainly be happy to commit to do that. Perhaps we should aim—where are we now, late April—to do that before the summer recess.

Chair: Yes, very good. Thank you very much.

Q13   Mr Jones: Sir Andrew, could you outline briefly for the Committee how you see your role as Chairman of UKSA and your relationship with the National Statistician?

Sir Andrew Dilnot: I see my role as Chairman of UKSA as chairman of the governing body. The National Statistician is our chief executive. As Chairman of the stats authority, there are two or possibly three activities that I think I and the board as a whole needs to be concerned about. One is the performance of our executive arm, the Office for National Statistics. A second is the performance of our regulatory function in the assessment of official and national statistics and in monitoring the performance of the statistical system as a whole. Then a third—slightly separate, which covers both of these on some occasions—is intervention about the use and misuse of statistics.

On the delivery of the work of our executive function, I think of John, the National Statistician, as being our chief executive and having clear responsibility for the functioning of the ONS as well as responsibility for the wider GSS. On the regulatory side, the assessment of statistics, John is not involved at all and that is a statutory requirement. My principal adviser there is Ed Humpherson, the Director General for Regulation. When it comes to the use and misuse of statistics, I will primarily seek advice from the regulatory function, from Mr Humpherson as the Director General for Regulation, but also on occasions will seek advice from the National Statistician as the person looking after the production of statistics and their understanding.

Q14   Mr Jones: Do you consider that there is a tension between the first of the two roles you outlined, one the oversight of the production of statistics and the other the regulatory role? If so, how do you practically manage that tension?

Sir Andrew Dilnot: I think there is inevitably a set of issues to be dealt with there. The way that we manage them is by having Mr Humpherson and his team completely separate in line management. Ed reports directly to me; he does not report to John. We rely heavily both on the code of practice and on the statutory protections that that function has. With that regulatory function in mind, we have been pretty tough on a number of people, including the Office for National Statistics, on a number of occasions over the last several years. I think both sides of that relationship understand that that is what is necessary.

It is worth saying that in his review Charlie Bean suggested greater separation between these two functions, and it is also worth noting that your predecessor Committee said the same in 2013. When your predecessor Committee said that, in 2013, we made a number of changes to further entrench the separation. We are entirely persuaded that we need to go further. In particular, our sense is that the separation between these two functions, which on the inside of the organisation seems very clear indeed, does not seem so clear from the outside. As we describe in our business plan, we will be working over the next couple of months to come up with new and different arrangements that will further entrench that separation so that it is clear not just from the inside, which is important, but also from the outside, which is very important for the—

Q15   Mr Jones: Could you tell us what arrangements you have in mind?

Sir Andrew Dilnot: There are four ways in which we want to develop the regulatory work, at least in part, in response to Sir Charlie’s review. The first is that, while we recognise that the assessment programme to begin with had to cover all of the official statistics—and I think it has persuaded most people of the trustworthiness of those statistics—we now want to focus even more on quality rather than on some of the process issues. We will be looking at focusing much more on quality.

A second is creating a clearer, separate voice for the regulatory function. I think some of that will be what seemed like rather instrumental issues about who the press office is and having separate communications professionals working just for the regulatory function.

A third is that I think we can do more on a standard setting. One of the pieces of work that I have been most pleased with on the regulatory side, is some work that we have done over the last two years emphasising the importance of standards in assurance of administrative data. Some of this came out of our interest—shared with your predecessor committee—in crime statistics, where it was clear that administrative data and police recorded crime had some issues to face.

Finally, our sense is that we need to do more work that is system wide. We have recently started a review of statistics in the area of health. We think there is scope to do much more of that. We are persuaded that all of that should be held within an entity of its own, which Sir Charlie calls an independent regulatory evaluation office. We are now working on plans to set up such a thing, which will give a much clearer actual and representation of separation between the regulatory function and the production side.

Q16   Mr Jones: Do you approve of Sir Charlie Bean’s suggestion that an independent office be created?

Sir Andrew Dilnot: Yes. Although what I would also say is that we are very clear that that function is already independent. It is independent by virtue of statute and by virtue of the fact that, if anyone were to ever try to subvert its independence I am quite confident—and the Chairman and I have had this conversation—that I could come to the Chairman of this Committee and this Committee would absolutely support the independence of that function. But we do see that further bolstering of that in fact and in representation is important.

Q17   Mr Jones: It does look very odd from the outside that a regulator can be independent if that regulator is collocated with the entity it is regulating. You share offices with UKSA. What would be a response to that?

Sir Andrew Dilnot: That it does not feel like that on the inside. Of course, I have to wear both these hats. Sometimes I wear a hat where I am going to the National Statistician saying, “Come on, what is going on with X on the ONS production side?” Sometimes I am wearing a hat where I am sitting down with the Director General for Regulation and we are saying, “These statistics on police recorded crime or trade are not good enough. We are going to take away the national statistics designation”, which we have done on several occasions in the last few years. It requires good behaviour on all sides. Honestly, it has not felt like a problem to me operating it, but I am aware that it is not well understood by those outside, which is why I think we need to take further action.

Q18   Mr Jones: Do you think it is possible to create that level of understanding, because you acknowledge that there is—at least in presentational terms—a problem there?

Sir Andrew Dilnot: We will wait and see. I think we can try very hard and I hope that we can, yes.

Q19   Mr Jones: UKSA resources that have been devoted to monitoring and assessment have fallen in recent years, yet there appears to be still more work to do. Can you explain why the resources devoted to monitoring have fallen?

Sir Andrew Dilnot: Not through our taking it less seriously. I think what Mr Humpherson would say would be remarkable efficiency savings. There has been some of that and we are getting more efficient at doing these things. The way in which we do some of the assessments has changed, but it is also the case that the board thinks that is the wrong direction of travel. Already, before the Bean review’s report had agreed to an increase in the resources that will be allocated to monitoring and assessment work, regulation—

Q20   Mr Jones: You can afford that?

Sir Andrew Dilnot: We are reconsidering even that over the next couple of months. One of the points I have made persistently to both Mr Humpherson and his predecessor is that, given the relatively small scale of the regulatory function in the statistics authority’s overall budget, I do not want a budget constraint there to affect the work that we do. So we have asked the Director General for Regulation to come forward with still more ambitious plans and we will be working on those over the next couple of months.

Q21   Mr Jones: You believe they are affordable?

Sir Andrew Dilnot: Yes, because the regulation activity is very cheap relative to the production activity because of the scale of it.

Mr Jones: Thank you.

Q22   Mrs Gillan: Sir Andrew, how many business plans do you have going at the moment?

Sir Andrew Dilnot: One.

Q23   Mrs Gillan: One. It is easy to confuse me, I am sure, but you published a business plan last year for the period 2015 to 2018. You also have the original “Better Statistics, Better Decisions”, and you have just told us you have published one last week, which I have been looking up on your webpage. You have three documents running on this at the moment. Which supersedes which, the business plan April 2015 to March 2018 or this latest one?

Sir Andrew Dilnot: The latest one.

Q24   Mrs Gillan: Why is there a latest one?

Sir Andrew Dilnot: Because I think we are intending to publish a new business plan each year, which—

Q25   Mrs Gillan: Is it made clear in the document that this is an annual refresher of the original business plan? I don’t think it is, but I am only scan reading it rapidly.

Sir Andrew Dilnot: I don’t know the answer to that. Let me pass that question, which is an operational question, to—

John Pullinger: If you are happy for me to answer that, there are three documents you refer to. The first one is a strategy, which covers the period 2015 to 2020. That remains and is still the strategy that guides our planning. We published a business plan 2015 to 2018 last year, which described itself as transitional. The particular reason it was transitional is because of last year’s general election and the fact we knew there was going to be a spending review. This year’s plan should be pretty resilient, assuming the spending review holds. We have now been able to match our ambitions with the level of resources we are going to be able to get from the Treasury for the period between now and 2019. Andrew is right that the current business plan sets out—and that is the one that supersedes the other plan—what we are going to be doing from now through until 2019. That is what the board is going to use to hold me to account for delivering each of those elements over the next three or four years.

Q26   Mrs Gillan: I have just read your foreword. It is not clear, is it, from your foreword that is signed off by you?

John Pullinger: It is not clear what, that it supersedes the previous one?

Mrs Gillan: Yes. Your explanation to me just now would have been very helpful if it was in this document because, at the moment, there are three documents out there that anybody can look at.

John Pullinger: Okay, that is fair.

Mrs Gillan: I come from a business background. That is confusing to anybody unless it is very clear.

John Pullinger: No, I can see that. It is less unusual within a Government context where plans do tend to get updated each year, with the first year being quite solid and the future years being less solid.

Mrs Gillan: But you could say, “Building on the last plan and now because we are after the general election period” and it would flow.

John Pullinger: I could have done that, I agree.

Q27   Mrs Gillan: Is it this latest business plan that you published just last week that we should judge you against? It is against that that you are going to be accountable?

John Pullinger: Yes.

Q28   Mrs Gillan: What are the key changes for the organisation since the business plan you published before the election and this business plan that you have now issued?

John Pullinger: The main change is that, on the back of the spending review, we now have confidence that we can take forward what I see as the main elements of what I think we need to accomplish to deliver the strategy, which certainly in my book has urgency to it and it is, I hope, going to be a radical change and a radical improvement in the statistics that the country has.

The first element, reflecting what the whole of the public sector is doing, is efficiency. We now have a concrete plan to deliver 19% efficiency savings on our current operations over this period and various ways of doing that. The second element is that I feel that our technology is woefully behind where it needs to be. We need to be ahead of it if we are to be a successful statistical organisation. We now have a technology transformation element that will be primarily over the next two or three years, and that will move us from a legacy state into a much more modern operational state. That is the second element.

The third element, which I hope is the bit that will interest this Committee the most, is the better statistics, the various elements in the first sections of Charlie Bean’s report. It is not just in economics. It is in areas like migration—which I guess we will come to—and a whole range of other topics where we need to have statistics that are much faster, much more fine-grained and much more attuned to the policy needs of the country. Those are the three core elements you will see in the plan.

Q29   Mrs Gillan: What are your milestones for measuring this new business plan?

John Pullinger: Some of the milestones are in the plan. As you go through each of the elements, some of them describe precisely when they will happen. At a very high level—the plan does not put it this way but it may help the Committee if I describe it—some of what I would describe as technical things, which are fundamentally about technology, I hope the next year will see a very significant improvement. The website has been the start of it. This week we are ramping up our move to more modern electronic forms of collection of data—particularly from businesses—and that will move up very quickly over the next year. We are also trying to wind out several of our very old systems.

There are some quite significant technology milestones over the coming year, but for me the real transformation will come probably next year with our efforts on learning and development of staff, improving staff engagement, getting staff at all levels much more curious—to use Charlie’s word—so that we are improving our statistics continuously at all levels in the organisation. That will start coming through very strongly in the next year.

Q30   Mrs Gillan: In this latest business plan, what have you dropped from the business plan prior to the election?

John Pullinger: The things that we have dropped: we had a consultation before Christmas on a whole range of areas where we felt we could either reduce the amount of data to be collected, reduce the frequency or the style of publications. As a result of that consultation, we have been able to bear down on our data collection costs. That is the first point. The second point is some of our publications will stop. They are outlined in a companion document to the business plan, which is much more detailed and technical.

Q31   Mrs Gillan: Did you publish that at the same time as the business plan?

John Pullinger: If it was not at the same time, it was the following day. There was a little bit of a rush, just organisationally getting things out.

Mrs Gillan: It is very recent, obviously, because we did not have access to that.

John Pullinger: It is very recent, yes. That is the response to our consultation on specific outputs.

Q32   Mrs Gillan: Just to understand, how do you arrive at your business plan? Who puts this together? Which elements of the organisation do you talk to and consult with? Is it driven from the bottom up or is it driven from the top down?

John Pullinger: What I hope it is is driven from the outside in. That is where I think the fundamental shift that the organisation needs to make should come from. Let’s use the Charlie Bean element as a starting point because that works across other topics, too. Many of our business plans talk, for example, about improvements to the national accounts, improvements to prices statistics. What Charlie talks about is better understanding of the modern economy. Your plan drives in a different way if you start from what is the modern economy and what do we need to know about it if we are to measure it. It takes you to a different place than if you start from the national accounts and build out.

Some concrete examples that I guess you are looking for: our discussions with the Bank of England, for example, have demonstrated there is a missing piece in our understanding of the flow of funds around the finance system. We now have a joint project with the bank, which is very much front and centre in the plan, to improve flow of fund statistics. Productivity: all of my discussions with the Treasury and various economic commentators are saying we have a productivity puzzle. We do not understand exactly what the elements of productivity are. What can we do, both at a macro and a micro level, to improve our understanding of that? The plan is driven by a myriad of conversations that we are having, including with this Committee and including the reports that you made just before the election. Your report on economic statistics is very valuable to us. Outside in would be my answer to your question.

Sir Andrew Dilnot: Can I add one important point? That is, that the business plan for the regulatory function is constructed in a wholly different way. It is not constructed by John and his colleagues as the National Statistician at the ONS. It is constructed by the Director General for Regulation and then discussed at the regulation committee, which is comprised entirely of people who are not involved in production, and then it comes up to the board. At the moment it is published at the end of but not completely separately from our other business plan. We need to think about exactly how we do that in the years to come as we move to more independent—

Q33   Mrs Gillan: Can I also clarify how much ministerial diktat there is?

Sir Andrew Dilnot: None.

Mrs Gillan: None at all?

Sir Andrew Dilnot: No.

Mrs Gillan: That is good.

Sir Andrew Dilnot: Absolutely none. There is one point in the cycle when there is no escaping Government involvement and that is in the setting of the budget. Apart from that, there is no ministerial diktat, certainly as far as what we are doing at the ONS is concerned. No Minister has that role. It is also worth saying that, in my time, no Minister has ever tried to influence in any way. Ministers may explain statistics that they really care about, and we take that into account as part of our user input, but there is no ministerial diktat. Were there ever to be such an attempt we would be straight on the phone.

Mrs Gillan: Just checking. Thank you.

John Pullinger: Nonetheless, I still feel free to talk to Ministers. I feel free to talk to members of all parties and both Houses. I want to be as open-minded as possible. When I talk about “outside in”, what guides me is that the things we need better statistics on are the things that are the most important decisions facing the country. Understanding where Ministers are coming from is an important part of that, but certainly no diktat.

Chair: We are going to press on more quickly. Thank you very much.

Q34   Mr Turner: The section of your business plan concerning regulation says it was produced before Sir Charles Bean’s report into statistics appeared. Which parts of your business plan are now obsolete given his conclusions?

Sir Andrew Dilnot: The point that is being made there, in particular, is that Sir Charlie produced two reports. The interim report late last year set out much of his thinking on the production side of the authority’s work. Therefore, we were able to take that into account when we produced the business plan for the production side. Our view is that, on the production side, most of what Sir Charlie said in his final report has been integrated into our business plan because we had a very clear understanding in advance about it.

Q35   Mr Turner: You said “most of”. What does “most” mean?

Sir Andrew Dilnot: To all intents and purposes, all I think. I will pass to John for confirmation of that, because we were very clear—as was Sir Charlie—for some time before his final report, what he was going to be saying on the production side.

On the regulatory side, he did not address that at all in his interim report. Our first conversations with him about what he was thinking of proposing came only relatively briefly before he produced his final report, so we had had to produce an interim business plan on the regulatory side before the final report. As Chairman, my view is that it is only on the regulatory side that our business plan needs further development. On the production side I think we were able to integrate all of Sir Charlie’s insights. John, is that your sense?

John Pullinger: That is correct.

Q36   Mr Turner: Yes. The work that is still to be done is not work by John Pullinger—

Sir Andrew Dilnot: No.

Mr Turner: But by whom?

Sir Andrew Dilnot: By Ed Humpherson, who is the Director General for Regulation, Sir Adrian Smith, who is the deputy chair of the authority responsible for that regulatory function, and the regulation committee.

Q37   Mr Turner: In their absence, you will have to answer for these?

Sir Andrew Dilnot: Absolutely.

Q38   Mr Turner: Right. It sounded quite a large piece of work for these other people.

Sir Andrew Dilnot: Yes. I think there is scope to do quite a lot. I don’t think the development of the business plan itself will be too large a piece of work. Doing what it is that we set out will be a significant amount of work.

Q39   Mr Turner: When will it be done by?

Sir Andrew Dilnot: We hope to have the business plan—our plan is for regulation resolved over the next couple of months—by the summer recess, so that we can implement the structural changes where they are necessary in the autumn.

Q40   Ronnie Cowan: I am looking at your business report and it contains phrases like, “will be proactive in bidding to deliver third party surveys”, “will offer more paid-for expert advice”, “maximise the number of customers paying for bespoke statistical analysis” and I wonder what the driver is behind this.

John Pullinger: That is for me. I think the driver behind that is, in a time of fiscal challenge when our central budgets are going to be constrained, if it is possible to find ways of raising extra revenue, then we can make the organisation more successful. Coupled with that, I think within the Government system we have a unique capability. For example, in recent times we have taken on responsibility for the household survey for Wales. That is extra revenue for us but it is also using our competence in social surveys. We have done some work for the Department for Business on the Teaching Excellence Framework in universities. We have a capability in the organisation. If we can get work from other parts of the system, which adds to our revenue so we can employ more people and build our capability, that helps everything else and it is beneficial all round. It makes the money we get from the Exchequer go further.

Q41   Ronnie Cowan: Is there not a danger of conflicts?

John Pullinger: There should not be. We will not take on any work that is not entirely consistent and supporting our public good mission. The legislation is very clear. I have no discretion to go beyond that. We have to be careful—and of course we will seek to be careful—but this is where I do need guidance from the Chair and the board as to where the boundaries are. This is a topic that we have discussed in the board and I very much welcome the widest possible advice because, even if I am confident, if it is perceived to be something that is straying beyond our remit I need to be cautious.

Q42   Ronnie Cowan: Will you keep two different sets of accounts then? The public are giving you money to run the service and the danger is it is subsidising your private enterprise.

John Pullinger: Our accounting should mean that does not happen. We have clarity over the estimate. I think the Committee has seen our estimate, which describes precisely what we are getting from the Exchequer. Our accounts also show the amounts that we are getting from other income-generating activities.

Q43   Ronnie Cowan: But there is no need here to set up a separate organisation that would be seen as a completely different entity?

John Pullinger: No.

Sir Andrew Dilnot: No.

Q44   Ronnie Cowan: That would make it more transparent, would it not?

Sir Andrew Dilnot: I think these are concerns that certainly deserve to be aired. They are concerns that we have discussed on at least two and possibly three occasions at the board in some detail. Our view is that we are transparent and it is very clear in our accounts that we have some income that comes in as a form of trading activity.

Q45   Ronnie Cowan: Does it reflect the money you are spending on that? If you are using personnel who are paid for through a public fund to do private projects that is not reflected in your accounts, is it?

John Pullinger: Let me respond. I use the example of the household survey for Wales. We have a contract with the Welsh Government to deliver that survey. They and we demand transparency over the money that is being spent on that. In practice, the team delivering that survey is the social survey team inside the Office for National Statistics, which is also delivering the labour force survey and the various other surveys that are the prime responsibility of ONS. Separating them into separate entities would be unhelpful to me. Having clarity about separate accounting flows is very important to me because I need to know where the money is going, and you need to know where the money is going to be able to scrutinise me. I do not believe it is necessary—and indeed it could potentially be unhelpful—to have a separate trading arm over here, separate from what we are doing as our day job, because the teams are integrated.

Q46   Ronnie Cowan: Maybe I am not making myself clear here. You have employees working for you who are paid for by public funds. You can use those people—their time—to do private work for yourselves, for the organisation. You can hive off five or six people and say, “Here is a job. Go and do that” and that generates funds for you, yes?

John Pullinger: Yes.

Ronnie Cowan: But the money you have put into their time, their man time, is that set off against that separate contract so we know if that separate contract is generating a genuine profit?

John Pullinger: There are established principles of public accounting. It would not be generating a profit. You would be looking at the full economic cost of their time and we would be charging it. What we are doing here is very common across the public sector and is certainly not a new thing for ONS. I am pushing it for the reasons I have said. I think it can improve our capability and enable us to be more effective, and we have things that are potentially beneficial to the rest of the public sector, in particular. There is nothing unusual here, but you are right to observe in the plan it is something I have given priority to.

Q47   Kelvin Hopkins: The Bean review recommended a high-level group of stakeholders—including the Treasury and the Bank of England—allegedly to assist with economic statistics. I have known you both well. I trust your objectivity, but I certainly would not trust the objectivity of the Treasury or the Bank of England. They will be there to spike things that the Government would not like, might be my guess. Don’t you think that, even though this might be a good idea, you should have some more independent figures on there as well?

Sir Andrew Dilnot: It is very important that any such group is not composed simply of representatives of Government Departments. I think Sir Charlie shares that view. On Friday of this week, at our next board meeting, we have invited Mark Carney, the Governor of the Bank of England, Dave Ramsden, the head of the Government Economic Service, senior civil servant at the Treasury, Dame Kate Barker, who did our national statistics quality review on economic statistics, and Paul Johnson, the director of the IFS, to come along to discuss the possibility that a group—which I think would meet relatively infrequently—could give the board feedback on what high-level stakeholders thought about what we were doing and what they thought the priorities were. As you will note, the people who we have invited to come along on Friday are the Governor of the Bank of England, the senior economist at the Treasury and two entirely external and independent figures. I think such a balance would be an important part of any such committee.

Q48   Kelvin Hopkins: You mentioned some other well-known establishment figures, but don’t you want some grit in the oyster, someone who is going to be a bit more challenging? For example, I think the National Audit Office ought at least to be around.

Sir Andrew Dilnot: We certainly think about that. We do not yet have firm plans. It is also the case that we think that, as well as a rather high-level group that would not meet very frequently, we probably should establish a committee of economic statistical experts that will be much larger and have many more people involved in it.

Q49   Kelvin Hopkins: A second question: do you think there should be a similar group or groups, indeed, to scrutinise non-economic statistics? There are some areas that are controversial: crime, immigration and so on.

Sir Andrew Dilnot: We have a number of such groups. I might ask John to talk a bit about one of them. For example, we have recently reformed a National Statistician’s Crime Statistics Advisory Committee. We think these can have some value. Some of them might want to be standing committees, some of them might want to be more ad hoc. In the last six or nine months we have begun doing work on the overall strategy for health statistics, where we have some concern. I can imagine that there may well be an argument for having a group to help us think about that. We have already informally formed such a group. That might not want to be a standing committee. John, do you want to say anything about the crime—

John Pullinger: Yes, if I may. Those committees are useful and good, but the main thing I want to say is: I want people to tell me the truth and I want people to say what they really think about us. If I am being completely candid, my observation is often groups do not do that particularly if you have mixed groups of government and non-government people. I want to know what senior people in the Government think of us. I also want to know what senior people not in the Government think of us. You asked at the beginning, Chairman, about ways you could help us. Any help you can give on how I can get accurate feedback about what we are doing well and what we are doing badly would be greatly appreciated. Particularly this Committee, you have a wide ranging remit across the whole sector. If you could tell me someone who does this well, I will go and meet them and talk to them. It is the thing that I struggle most with, to be absolutely honest.

ONS has had a long tradition of getting reports each year from the Bank of England, and the OBR, and all these people, and it is all very soft. This Committee has been quite sharp in some of its reports on the Office, but that is what I need. I need to know how we are really doing, and your first set of questions was about that. Any advice is gratefully received. I am not sure that a group is going to be the answer, valuable though it will be. I think Sir Andrew has described the nature of the group we are trying to put together, and if we can get a bit of grit in it, excellent. But I am not sure, just by adding one or two extra people, whether that will help the candour that I need to come our way so that we are really getting the truth as it is. Then we can learn, and then we can develop, which I think was the sentiment you started with.

Q50   Kelvin Hopkins: One more question, if I may: if I was on one of these committees or these groups, not dealing with economics, but immigration—and it is really for the Government, but you may comment on it—I would ask why is it that at least a decade on since we wanted all those who leave the country to be recorded, and all those who come in to be recorded, and for everyone to have biometric passports so we really get some grip on immigration, and what is happening? That is not happening.

John Pullinger: Is that a question for us? We are humble statisticians here. We measure the numbers that we can find, and we will be very open-minded about which numbers we get. We may well come a bit more to immigration later, but we do work very actively with the Home Office to get such data as they have from the visa system. It is getting much stronger in the data it gets. The exit checks data particularly, which is now routinely collected, is improving the situation. There are all sorts of other political and policy considerations around having a different regime at borders that are far beyond the remit of a statistical system.

Q51   Paul Flynn: It has been suggested that a user group composed of non-government organisations, think tanks and journalists, would be useful to you. Did that suggestion come from anyone except journalists, non-government organisations and think tanks?

Sir Andrew Dilnot: I don’t know. I think it goes again to the point that John has raised: we are very clear in the “Code of Practice for Official Statistics”—our very nearly founding document—that the purpose is to meet the needs of users, and we need to find good ways of finding out what the needs of users are. As John says, while groups can sometimes be helpful, there is not always a great deal of clarity and honesty in such groups. On the other hand, at least a group can look as though it is legitimately formulated. Simply, John or me talking to our mates is not quite good enough. We have to find a mix of these ways forward. That is something that you will see in our business plan, certainly in our strategy. Another fair criticism that comes in the Bean Review, and that we ourselves feared, is that we need to get out more. That is probably true of all statisticians.

Q52   Paul Flynn: With your devotion to objective truth—which I think is a fairly genuine one for which you should be congratulated—would you be concerned that you might have a think tank that is being paid to think what they come up with, such as the IEA, which is outrageous, claiming to be a think tank and a charity but refusing to reveal the money it gets from the drink and the tobacco industries and so on? Would you be concerned to have the malign influence of the IEA advising you?

Sir Andrew Dilnot: I think my answer to that, Mr Flynn, is no. I think they are users. There are a wide range of users with which people may agree or disagree, but all, to a greater or lesser extent, are legitimate users. The point where I would be angry would be if any such user were then to abuse or misuse the statistics, and that is when our regulatory function would come into play. I think it is very important, of course, to be alert to where people are coming from, if they are coming from a particular perspective, but I think we have to recognise that people do come from a wide range of perspectives and they can all have legitimate needs. What they must do is use our statistics—this great treasure—fairly and accurately.

Q53   Paul Flynn: Do you think Mr Bean has given excessive importance to the location of jobs in the Office for National Statistics? It is as easy for me to communicate with my office in these buildings as it is with my office 140 miles away. With modern communications, it does not matter where offices are located. Would you agree with that?

Sir Andrew Dilnot: I think our own sense is that Sir Charlie—rather like Kate Barker and Art Ridgeway when they did the review for us on national accounts—has analysed the position effectively. He has noted that the move from London to Newport of economic statistics certainly caused some dislocation, as it could have done if it had been a move from Newport to London. Sir Charlie notes, as we do, that we are now building up a really exciting centre of expertise in South Wales. As you know, Mr Flynn, I am from South Wales myself, although I am from the great city Swansea rather than Newport.

Paul Flynn: Indeed. None of us are perfect.

Sir Andrew Dilnot: What Sir Charlie says is that he thinks we should carry on doing that, and indeed, in the data campus that we are building up, we will continue to do that. He also notes, with which I and the board agree, that it is very important to make sure that all of our activities, in Newport, Titchfield and in London, are connected into our core users. Because of that—and because of the concentration, particularly in London, of economic statistics expertise—there is also a case for having a London presence; not as a substitute in any way for our presence in Newport but as a complement, as a way of helping our large and dominant presence in Newport to be fully engaged with some of the debates that are going on in London. That seems to us a helpful way of thinking about things.

Q54   Paul Flynn: I will take that as an answer to say, yes, you believe the location is not as important as it might be. Are you aware of the magnet of the statistical establishment in London that has always resented—from the time of the relocation—the fact that jobs went outside of London? Does it make sense to try to get people back into polluted Pimlico, with sky-high house prices, with long commuter distances, rather than allow them to develop in the glorious, broad, green acres of Gwent?

Sir Andrew Dilnot: I think that is suggesting, Mr Flynn, that these are alternatives, and our view is that they are complements. There is no suggestion of moving people from Newport back to London. Indeed, we are recruiting more economists in Newport. We are expecting to recruit about 18 new economists in Newport this year. We are expecting to expand our economic analysis capability in Newport, but we are also expecting to build up somewhat of a presence in London to help those working in Titchfield, Newport and London do an even better job than we are already doing.

Q55   Paul Flynn: I am grateful for what you said about developing the Newport office. But do you think there is a gap in Mr Bean’s report, where he fails to appreciate the advantage of relocation in helping out areas where they have seen manufacturing jobs collapse and the really miraculous success of statistics as a brand new career in South Wales and the development of it? Do you think there should be some more fulsome acknowledgement of how successful that has been?

Sir Andrew Dilnot: I think Sir Charlie’s report is pretty fulsome about the development. I think he also recognised that there was a dislocation effect, as I say, in moving from London to Newport. The same would have been true if we had moved from Newport to London. That did have an effect on our capability in economic statistics, but Sir Charlie recognises that we are now building from strength to strength.

Q56   Paul Flynn: How many jobs will be created by Mr Bean’s report in Pimlico?

John Pullinger: We expect the London presence probably to be up to 100, but we already have quite a reasonable sized group in London. That includes both the regulatory element and the production element.

Q57   Paul Flynn: You do not see this growing and as a surrender to the London statistical establishment?

John Pullinger: No, it is not a surrender. It will enable the whole organisation, including the Newport and Titchfield elements, to be more successful, because it will connect us particularly to this place and to the Whitehall space.

Q58   Paul Flynn: As I say, I am baffled by this argument. How can it make you better connected? It makes no difference now, when we are all connected. All parts of the world are wonderfully connected by the miraculous technical instruments we have. The idea of location in this area is a very old-fashioned one, but it suits the statisticians, the journalists and the think tanks to feel that we are all cosy together in London and it does not matter what happens out in the sticks in Scotland or in Wales.

John Pullinger: I will give you one very specific example. I would love to second more people from other organisations—potentially including this place—into ONS to help me on various projects, and vice versa. That has proved to be very difficult when we have had to find people who are prepared to spend some time in Wales, if they are London-based, or Wales-based people to spend some time in London. The logistics are really difficult. I am trying to reduce the barriers to us getting the best people.

As Andrew said, we have developed Newport, and we will continue to develop it as a go-to place, particularly for smart statisticians and economists and, increasingly, digital and data people. That is starting to work, but there is a complementary opportunity for us here—it is quite a small amount in the grand scheme of things—that will enable us to get access to some of the great people. I would love to have some of my old former colleagues from the House of Commons Library working with us for a year or two, but relatively few of them would spend a year or two in Wales. They would spend a year or two in Pimlico.

Q59   Paul Flynn: What is rather dispiriting about Mr Bean’s report was his failure to reflect what the Lyons review and others said about the fact that staff from Pimlico went down, resettled in Wales—reluctantly, many of them—have stayed there, have retired there, and found it has been a great success and greatly enriched their lives by moving them out of London. Why wasn’t that positive value in this, apart from doing wonders to the local economy, reflected in Mr Bean’s report?

John Pullinger: I think it was acknowledged too, but if it wasn’t acknowledged enough let’s just acknowledge it now.

Q60   Paul Flynn: Mr Bean is part of this London establishment, your statisticians, this world of people who greedily want to have everything in London. You are not part of it, I am sure. Of course, Mr Dilnot has the great advantage of being born in Wales. Isn’t the metropolitan magnet exerting an excessive pull towards London again, and isn’t this a retrograde step?

Sir Andrew Dilnot: Let me say briefly: I think not. I think the review was clear that we should continue to develop Newport and that it would be foolishness to think of moving the Newport establishment back to London. We are not thinking of doing that. We are allocating substantial resource from a small budget to further enhance the level of economic expertise in Newport, particularly. Our business plan sees that developing. We are very excited about Newport. We also think that we can make the Newport and Titchfield activities even more successful, even more attractive, if we build up our presence in London.

Paul Flynn: Can I say one final word? Personally, I am deeply grateful for the value of the work of both of you, which is, having been watching this subject for many years in this House, a great improvement on what it was.

Q61   Kelvin Hopkins: Your letter to us said that, “Having access to administrative data is critical for our statistical work”. Sir Charles Bean said that you have lacked this in the past, particularly with reference to HMRC data. Which data sets do you have access to, which have you failed to get, and which would you like access to?

Chair: Can you give some examples?

John Pullinger: Okay. It is not so much the data sets that I want; it is the ability to inquire and to take that inquiry wherever it leads. I will use the migration example because I think it is relevant to this discussion. At the moment we have our international migration estimates based on the international passenger survey, which was never designed for this purpose, but has proved to pretty robust when you test it against external metrics. If we had had access to the administrative data sets I would be much freer to be able to go to the Home Office and get individual data, such as it is, on exits and incoming data.

At the moment that requires protracted negotiation, not for a bad reason but because my colleagues in the Home Office have a particular set of legal requirements around what data they are allowed to share with people, and there is not unambiguous clarity that I am someone they can share data with. If I am thinking about immigration, I want also to go to DWP to be able to understand what is happening with the issuing of National Insurance numbers. I want to go to HMRC to understand which foreign-born people are on the system somewhere, either because they are paying tax or claiming benefits. I may want to go to the Health Service to find out how many foreign people are using the Health Service. To understand immigration I need to put together a jigsaw of different sources to help me understand best what is going on, serve the public good and inform this House about what is happening.

At the moment the only clear legal gateway to do that is the gateway that was agreed in the Statistics and Registration Service Act, which requires me to define in advance what I want, data set by data set, in a very precise way, take it through both Houses and get agreement to that. That takes a minimum of six months. We have been trying to push it through faster and that has been working well, but again we have to do things properly. Our own legal framework and our own desire is to ensure we are open and transparent about what we are doing and that we are not asking for things that would somehow be concerning for individual citizens in terms of privacy.

Chair: That is a very long answer.

John Pullinger: I am sorry. My answer is: I would not see it data set by data set. The reason it is valuable to us, if we get it as was set out in the consultation, is that I can follow the data where the questions lead me.

Q62   Kelvin Hopkins: I have to say that is very illuminating and interesting. With specific reference to HMRC, I am one of those MPs who have banged on for years about the tax gap and the failure to collect taxes that have either been avoided or evaded. A great friend of mine, Richard Murphy, has written about this extensively. He has come up with a very large figure for the tax gap, much bigger than HMRC’s figure. Would you like more statistics to be able to research that and to demonstrate conclusively where the truth lies in relation to the tax gap?

John Pullinger: Part of my role is head of the Government Statistical Service. There is a professional head of statistics in each Government Department that I work with, and my relationship with my colleague in HMRC is extremely good. We have done a lot of research in the context of the national accounts to estimate the tax gap and ensure that it is incorporated into the national accounts, and there is an understanding of what can be known and what is yet unknown. The tax gap, by definition, is something that is going to be very difficult to pin down, because it is the absence of something. The more access to data we have the more we can bring together the analysts from the different communities to get a common understanding of what is going on, and then we can explain it to people. Yes, of course I would like more but I don’t sense there is any kind of resistance at the moment. There is just a framework that creates friction that slows us down.

Q63   Kelvin Hopkins: There are couple more straight questions here. Have UKSA non-execs been sufficiently forceful in helping to open up access to the data?

Sir Andrew Dilnot: It is a matter for legislative change. The UKSA non-execs have been involved in the discussions over both this Parliament and the last Parliament with the parts of the Government that are taking forward the legislative change. At the moment I feel pretty optimistic about that. The UKSA non-execs are absolutely seized of the importance of this. We cannot look forward to a world, in five years, 10 years, or 20 years, where we are as reliant on survey data as opposed to administrative data as we are now.

Q64   Kelvin Hopkins: Speaking about the future, how do we ensure that UKSA will have access to the data it needs?

Sir Andrew Dilnot: We change the law. We hope there is a Bill coming forward to Parliament later this calendar year that will do that. It is out for consultation at the moment. I am not sure when the consultation is due to—

John Pullinger: The consultation finished last week, so we are just starting to analyse the responses. Changing the law is one thing; another thing I think this Committee—and I hope the House—will help us with is getting the balance right. Clearly, as an analyst, I want more data but I don’t want more data than is publicly acceptable. I want to stay on the right side of individual citizens concerned about what is right to use individual data for, so help me get that balance right. To be honest, I think that is the critical thing: public acceptability about what we are doing, so there is a real understanding of the public benefit, and we are very careful in the way we set ourselves up so that we are not doing anything that citizens would feel uncomfortable with.

Q65   Chair: Sir Andrew, I will declare an interest as a director of Vote Leave, which you may or may not have been aware of when you issued a mild rebuke to Vote Leave about the use of £350 million per week as an indication of what it costs the United Kingdom to be a member of the EU. First, I will say how refreshing it is that you scatter your advice indiscriminately. I think that is commendable and is a great signal for your independence. Can I also clarify: the £19.1 billion figure in table 99 is a valid figure, nevertheless, if it is surrounded by sufficient explanation?

Sir Andrew Dilnot: Yes. Your term is “mild rebuke”; the letter I wrote was not suggesting that £19.1 billion was not the gross figure. Indeed I had written to you, two or three weeks earlier, setting out all the figures. My concern was that, for example, the poster that said something of the form, “We should stop sending £350 million a week to the EU and spend it on our NHS instead”, could be interpreted as implying that the gross figure was in fact a net figure. Mr Gough, for example, is very careful in his public speeches to say, “We send £19.1 billion, but a rebate of £4.4 billion comes to us and £4.8 billion comes back to us that goes to the private sector”. It was that that I wanted to draw attention to. Yes, the £19.1 billion figure is a legitimate figure for gross contributions. The £4.4 billion or so of rebate is another important and legitimate figure. The £4.8 billion or so of contributions that come from the EU to the private sector, so the point I was making was about making sure that people understood what the figure was being implied to be.

Q66   Chair: Given most of the population started this debate thinking that we get money from the EU, and they are unaware of how much we give to the EU, it is a very positive discussion to have. I am grateful for you also validating the £9.9 billion net figure, which we do not get back any of.

Sir Andrew Dilnot: On the £9.9 billion, it is also important to be clear about what we are talking about here. The £9.9 billion is about flows of public sector funds. It is also the case that there is another figure that is not an official statistic, which takes into account flows from the EU to the private sector and that gives a lower possible net figure. But the official statistics are the £19.1 billion, the rebate and the net official funds.

Chair: They are the Pink Book figures. They are the balance of payment figures, and they are the ones that can be measured. As you pointed out in your letter to me, that is the direct effect on the Exchequer of what we have to pay in tax to fund the EU.

Q67   Kelvin Hopkins: One other thing I would also be interested in is: there is a substantial outflow of money from Britain to expatriates who live overseas in the EU, who largely live on incomes derived from Britain rather than from overseas. There is also a substantial amount of money submitted by EU citizens working here sending money back to families abroad. Those are substantial outflows of finance from this country that are, I think, a net outflow rather than just a gross.

Sir Andrew Dilnot: I would need to think a little bit about that, because there are also EU citizens living here. There could be flows in both directions.

Q68   Kelvin Hopkins: There are not many expat EU citizens living here on income derived from abroad.

Sir Andrew Dilnot: I don’t know the answer to that. It is an interesting question.

Q69   Chair: It is an interesting question, but can I invite you to look in a similarly dispassionate and impartial way at some of the forecasts published by the Treasury on 18 April? Has that provoked many letters of inquiry to you?

Sir Andrew Dilnot: No.

Q70   Chair: I have written to you about the Government leaflet, because the claim that 44% of our exports go to the EU is an unqualified figure. While the website might annotate some of these figures and qualify them, the leaflet that went to the households, which is what people will read, has no such qualifications. When can I expect a reply to my letter?

Sir Andrew Dilnot: You can expect a written reply very soon. I am very sorry not to have signed it off yesterday, but I wanted some final quality assurance. I am very happy to give a summary now of what I think about it. In particular in your letter you drew attention to what is described as the Rotterdam effect, that, while what the balance of payments figures measure is the declared final destination of goods, in some cases in all ports it will be the case that those goods go on elsewhere, and this is thought to be a particularly large effect for Rotterdam and Antwerp, in the Netherlands.

The figures that you asked about in the brochure are something like 44% of all goods and services exports go to the EU. That is a figure in the last three years of about £500 billion per year of total UK exports. That is the exports to the world. The exports to the EU have been of the order of £220 billion to £230 billion a year, and that is the 44% figure. In fact, I think the precise figure for the last three years would be 44.3% in 2013, 44.8% in 2014, and 43.7% in 2015.

Measuring the Rotterdam effect is very tricky and, indeed, there is not just a Rotterdam effect. There will be a London effect and a Hong Kong effect, because this effect can occur in all trading ports. I think it is widely recognised it is probably larger in Rotterdam than in most places. This came up when you had the National Statistician in front of your Committee a little while ago. Subsequent to that, the ONS published a note on the Rotterdam effect that pointed out it is indeed very difficult to estimate, but said that it thought assuming that 50% of goods exports to the Netherlands would be a top-end estimate of what the scale of the Rotterdam effect would be.

If you take that as the estimate that would give you a sense of the Rotterdam effect in the last three years—2013, 2014, and 2015—being 2.49% of total exports, 2.18% of total exports, and 1.69% of total exports. Those are numbers that should be compared to the 44%. On balance, what I am likely to say in my letter to you is that I think it would be a good idea if there was a note in the Pink Book saying, “Be alert to the Rotterdam effect”. The reason for that is, while I don’t think that when talking about exports to the whole of the EU this is a large enough figure that we should be very concerned about it, if somebody was looking at British exports to the Netherlands then I think one would be seriously misled if that figure was not taken into account, so we will set all of that out. As I say, our sense of a plausible but not certain estimate of the size the Rotterdam effect would be that, compared with the total exports to the EU—which are 44% of our exports—it might account for about 2% of our exports

Q71   Chair: What do you make of the impression that the Government is intending to convey by juxtaposing jobs in the sector against the exports from the sector? How do you think the Government is intending that to be received by the uninformed reader?

Sir Andrew Dilnot: I don’t know, but it does seem to me that we—my ONS colleagues—are looking to publish some additional material in this area. We have timed it for 25 May. It has been pre-announced. One of the things that we will look at is I think it would be helpful to have information that showed jobs in the sector, the amount of a sector’s output that is exported, as well as the proportion of exports of that sector that go to the EU, because you can imagine for example that it is—

Q72   Chair: That would suggest that the Government would like us to believe that there is some absolute conditionality to the total number of jobs related to the exports and our membership of the European Union, when there might be no relationship between those two factors.

Sir Andrew Dilnot: In my letter to you I would probably draw attention to the letter I wrote a couple of years ago to Matthew Elliott, who was then running Business for Britain. It was over a concern I had received about a statement, I think made by the then Chief Secretary to the Treasury, which was thought to suggest that the 3.3 million jobs that are linked to trade with the EU—which is a reasonable estimate—might in some way be linked to membership of the EU. I think the final sentence in my letter said, “It is important to note that this is not an estimate of the impact of EU membership on employment”.

Chair: Thank you.

Kelvin Hopkins: Reinforcing one or two of the points that the Chair has made, my favourite bedtime reading is actually the monthly economic statistics produced by the House of Commons Library, which Mr Pullinger used to oversee in his previous employment, and within that it very simply shows that not just the imports but the exports are valid. It demonstrates clearly that there is a very much larger proportion of our imports coming from the European Union than there are exports going to it, and there is a trade deficit of about £60 billion plus a year. You can simply calculate the number of jobs we are effectively exporting by that. This is between the EU and Britain. The numbers are very plainly printed every month in this excellent bulletin.

Q73   Kate Hoey: Chairman, while we are on the same topic could I just ask: were you two involved with the production of the kind of fabricated Government document that has gone to every household?

Sir Andrew Dilnot: No, I wasn’t.

Q74   Kate Hoey: Were you asked anything? Were you involved in any of the statistics in it?

John Pullinger: We do fact checking for anybody who asks and we were asked with some for numbers[1].

Kate Hoey: Were you asked to fact check it?

John Pullinger: No, not the whole document, kind of specific elements. So, “Is this number described properly?”

Q75   Kate Hoey: You ticked it off as being absolutely accurate?

John Pullinger: No, and that wasn’t the question we were asked. There are many, many things in the document that involved different experts. Inevitably we have relationships with the people working in different ministries on each of the topics. What I observed was what I see in the civil service for all other documents. You are trying to write something and it has some factual material in it. We should check that factual material with who originated the factual material. We received the information on that basis and gave our comments.

Kate Hoey: Thank you.

Q76   Mr Jones: I wonder if I could raise a variation on the Rotterdam effect. It is just one example; I am sure there are others. The wings of every Airbus aircraft that flies anywhere in the world, or is sold to any country in the world, are manufactured in North Wales, in Broughton, but of course they are then sent to Toulouse in France for assembly into the final plane. As I understand it that is regarded as a French export outside the EU but nevertheless, so far as Wales is concerned, it is an export to France. Given the value of aircraft, wouldn’t that have a distorting effect as well upon the—

Sir Andrew Dilnot: I think my response would be: I don’t think it has a distorting effect but it is an important and interesting effect itself. What imports and export figures are doing—as I understand it and the National Statistician can correct me—is monitoring the change of ownership at the border for things whose final destination is specified as the place of and source of origin. Of course, increasingly, supply chains are quite complicated. To build an aeroplane is a very complicated thing; to build a piece of software is a very complicated thing; even to sell an insurance contract can be quite a complicated thing, and working out exactly where in the world all of those things happened is tricky.

Q77   Chair: Does software count as goods or services?

Sir Andrew Dilnot: I suspect the answer is: it depends on exactly what is happening. I suspect that in some cases it will be a good and in some cases where you are buying the services of a software engineer to write you a bespoke thing it might count as a service, but I don’t know and we would be happy to check up on that for you if that would be helpful.

John Pullinger: Mostly that would be services. The question Mr Jones describes is a good one. We have captured in our statistics what the conventions around national accounts and balance of payments require. When I spoke earlier about trying to have an outside-in approach, what I need to understand are the questions we collectively need to know as a nation, and supply chains around aircraft manufacture potentially are one of them. I think I will have succeeded with our business plan if within ONS we have the kind of curiosity that says, “Now that is an interesting question. Let’s find out the answer”. At the moment we have done precisely what the system requires us to do and the export of wings from North Wales to Toulouse is an export from the UK to France: period.

Mr Jones: Notwithstanding that they end up in an aircraft sold to Emirates in the Gulf.

John Pullinger: Correct.

Mr Jones: In other words, it is our contribution. It is Britain’s contribution to an export that goes outside the EU but, nevertheless, is recorded as an export to France within the EU.

John Pullinger: Yes.

Q78   Chair: How useful are trade statistics as supply chains become so much more complicated?

John Pullinger: They are useful in measuring changes of ownership as described in the national accounting framework.

Q79   Chair: In a globalised world, as regards the binary idea of trade with one country measured by one number, what you seem to be suggesting is that we no longer live in that sort of world.

John Pullinger: I believe that is correct and, indeed, I have been spending some time talking to my OECD colleagues and my counterparts in the US, Canada and various other countries are trying to understand how best we can capture in a consistent way precisely this new economic reality of being effectively borderless; at least with plane wings you can spot them moving around, but with the software example, it is really hard.

Q80   Chair: How do you feel about the Government using these very complicated and sophisticated numbers in this very binary way to try to prove a very binary point?

Sir Andrew Dilnot: I think that is probably one for me. The exports and imports figures do not tell us everything about trade. They tell us about the value of exports and imports. How accurate are the destinations? Well, they tell us what is written down on the forms for final destinations and final sources. That means that there is some uncertainty around them. I don’t wish to put words in the Government’s mouth, but I imagine that one of the points that the Government was trying to make in that brochure was that the EU is our largest trading partner. So 44% of exports go to the EU. Our next largest trading partner is the US and the figure there is a bit less than 20%. There is certainly uncertainty about the interpretation of both of these numbers but their relative scale is clear and, even if we were to take, for example, the top end of the Rotterdam effect number, we still have a big difference.

I think in the end, as so often in statistics, what matters is interpretation. As I say, it seems to me that if we are looking at the scale of trade with the EU, compared with others, then this number is fairly meaningful. If we were looking at the scale of trade with the Netherlands, then I think we would have to be much more careful because there it is a very large part of what is going on.

Q81   Kelvin Hopkins: In broad terms, where the value added is created and where people are paid to create that value, it is still significant?

Sir Andrew Dilnot: Yes.

Q82   Ronnie Cowan: We seem to have laboured the whole Brexit argument here, rather than looking at how your organisation actually works. Is it fair to say then, when it comes down to Governments and referendums, it comes down to lies, damned lies and statistics?

Sir Andrew Dilnot: No. That phrase is one of the few phrases that really irritate me. Statistics are no more or less capable of distortion than words, and statistics at their very best force us into precise thinking so—

Ronnie Cowan: People looking to seek their own end means will skew statistics, and politicians—

Sir Andrew Dilnot: They will, as they will skew language. In the end, statistics are just another form of communication and they are capable of conveying meaning that it is very difficult to convey in any other way. I think they are wonderful, marvellous things that we should treasure. That is why I so value the care that this Committee takes about them.

          Oral evidence: UK Statistics Authority, HC 978                          

22