Public Administration Select Committee (PASC)
Oral evidence: The Work of the Chief Executive of the UK Statistics Authority and National Statistician, HC 552
Tuesday 21 October 2014
Ordered by the House of Commons to be published on 21 October 2014.
Members present: Mr Bernard Jenkin (Chair), Mr Nigel Evans, Paul Flynn, Mrs Cheryl Gillan, Kelvin Hopkins, Greg Mulholland, Lindsay Roy, Mr Andrew Turner
Questions 1-94
Witness: John Pullinger CB, National Statistician and Chief Executive of the UK Statistics Authority, gave evidence.
Q1 Chair: Can I welcome our witness to his first session in front of this committee as the National Statistician, if I’ve got the title right, and congratulate him on his appointment? Please could you identify yourself for the record?
John Pullinger: I am John Pullinger, National Statistician.
Q2 Chair: Can I ask you to give short and crisp answers? We will endeavour, colleagues, to ask short and crisp questions. We will get through much more. We have plenty of questions to ask. Can I first ask you about how you are setting your strategic priorities and how your personal priorities are fitting into the priorities of UKSA as a whole?
John Pullinger: The drive around everything I am doing with setting priorities is ensuring that the statistics we produce are helpful, and particularly helpful to decision‑makers, parliamentarians, politicians of all sorts, the Government, businesses and the public. The statistical system in the UK has a very good tradition of producing good numbers, but we have not always been as connected as we need to be to the decisions we are seeking to inform. What is guiding my priorities is that, simply put.
In terms of what I think are the most crucial things for us to do to meet that remit, they are firstly the economy—to make sure we have the information that is needed by Parliament, the Government and the Governor of the Bank of England to make sound decisions about the economic situation. The second priority is much more generally to connect us with the public policy process, in domains like crime, immigration and education, where there is a desperate need for good evidence that people can rely on, but our numbers are not always quite as tightly attuned to the decisions that are being made as I think they could be. The third priority underpinning all of those is to invest in what I am describing as data capability. That includes things like having a jolly good website, so you can find our numbers, but it also includes, much more radically, improving the skills base both inside and outside the organisation, so we can actually use this evidence we are creating in a much more thorough way.
You ask how that fits with the strategic priorities of the Authority. Well, it fits very clearly. The first priority is quality, and we need to have quality statistics to do that. The second is impact, and everything I am talking about is ensuring that our statistics make an impact. Clearly at a time of financial stress, efficiency—using the resources we have very wisely—is essential. We have to have the coverage necessary to cover the whole domain of the economy and society. Finally, we need to be trustworthy because, if we produce numbers that people do not believe in, then we are wasting our time. Those priorities very firmly fulfil the strategy that the Authority has set out.
Q3 Chair: In your paper, entitled “Chief Executive’s Report”, you comment on the need to build leadership capability and you note that your staff engagement has recovered to the level of the Civil Service average but “we are way off what the best organisations achieve in terms of leading and managing change”. Would you like to enlarge on that a bit?
John Pullinger: You and I have talked about staff engagement on other occasions. We both believe that an engaged staff, who are really committed to the purposes of the organisation and are feeling that they are well led in the direction they wish to follow, is a kind of hallmark of a successful organisation. When I look at the people survey scores, the scores around leadership are the ones I am most interested in.
It is gratifying that the scores for ONS are now at the general level of the Civil Service, when they were below, but I am setting a much higher bar. First of all, I want us to be above. I guess every head of every Department wants to be above the average, but I would be the first to say that that is not likely to be possible, but I certainly want to be above the average on leadership. One of the features of the last few years in the statistical system has been that it has been very complicated and there has not been a simplifying set of accountabilities for the top team. By bolstering our leadership capacity, we will have a much better chance of improving engagement and improving the likelihood that the system as a whole is actually going to deliver the kinds of benefits that I have set out.
Q4 Chair: On your engagement figures—we have some of them in our brief—the response rate for ONS was only 55%. That is way below the Civil Service average. Why do you think that was?
John Pullinger: Last year, I understand, there was a particular campaign amongst the unions in ONS, or amongst PCS to be precise. They did not want to engage with the survey. I know, amongst some union officials and members, there are anxieties about whether the survey will be used in a positive way towards staff or in a negative way. I respect those anxieties but, for this year, the people survey is out at the moment. I am encouraging everybody to respond and hope to get a better response rate, but time will tell.
Q5 Chair: I am amazed, but obviously you have some bridge‑building to do with people who are listening to the trade union rather than you.
John Pullinger: I also want to listen to the trade unions. I have had good engagement with the trade unions in ONS so far, and they are clearly looking after the interests of their members. They are also interested in the success of the organisation as well. I am engaging, I hope constructively, with the unions and with staff at all levels.
Q6 Chair: On leading and managing change, UKSA staff seem reasonably engaged at 60%, but again ONS is down on 35%. That suggests very little faith in the ability of leading and managing change.
John Pullinger: I may be anticipating a future question, but one thing I observed quite strikingly coming into the organisation is it has been separated into quite distinct packets. There is a UKSA packet, which is quite separated in the way it thinks of itself from the ONS packet. In the UKSA box, there has been quite good clarity around the role that they fulfil and the work that they are doing supporting the Chair and the board. It has been a more straightforward task for them to be engaged and to feel engaged. They are a smaller group of people as well; that also makes it easier.
For ONS as a whole, there has been a period—the last five years—of coming to terms with the fact they are now part of a new statutory office. It is only now that the whole system is coming of age and able to come together in a much more creative way. Certainly my intention is to get the whole lot feeling much more engaged and drive up those low scores to the level of the best.
Q7 Chair: What are you going to do to increase these scores?
John Pullinger: I will focus on this clarity of purpose that everybody should be well aligned in creating an information base that is well used by the people it seeks to serve. In its most straightforward way, I have spent the last 10 years as a Librarian of this House, and we had jolly good engagement scores there. I think that is because people felt they were doing a job that was appreciated. People felt they were helping other people be successful. Exactly the same kind of thing could apply to the ONS as well.
Q8 Chair: How good do you think your leadership team is at listening and reviewing in a positive way, learning from mistakes and inviting people to tell the management chain about their problems, concerns and what they think is going on? How good do you think your people are?
John Pullinger: My observation is that the people are pretty good. The people are pretty well motivated, but there has been a legacy of quite cumbersome decision‑making practices in some areas: a legacy of some projects—I am sure we will come to the website—not quite fulfilling the expectations that they were set out to fulfil. Leadership has been quite heavily weighted by the task it has been set. I am hoping, by having this kind of unifying message around what we are seeking to do, that it will bring out the best in people, and they will be able to work much more effectively and successfully as a team.
Q9 Chair: What are the key attitudes that you need to change? You say you do not propose to launch a new change programme, but what you are talking about is certainly something that needs to change.
John Pullinger: Absolutely, and quite radically to change, not to say that what has gone on in the past is a bad thing. The expectations of us are dramatically higher. In order to fulfil those expectations at a time when finances are very tight—
Chair: What are the attitudes you need to change?
John Pullinger: I want everybody to be hungry for improvement. A word that is being used a lot in the Office at the moment is a spirit of “curiosity”. Everybody should always be looking at the numbers they are producing and saying, “Are these as helpful as they could be? What is not quite right here? Who else could help us make it better?”
Q10 Chair: How curious do you think your management is about what it feels like to work in the organisation?
John Pullinger: I cannot talk about what had gone on before. Yesterday, I had a meeting of the 50 most senior managers in the Office, and I sensed a very good spirit to try to change, improve and make the organisation more successful.
Q11 Chair: You keep talking about what things you are going to improve. I am asking about how you improve your listening and learning about your own people. What are you going to do to improve that, because that is behind the attitude that creates this sense of disengagement?
John Pullinger: I am summarising in this word “curiosity” being our own harshest critic, people not being satisfied with the way things have always been done, and thinking from the point of view of the consumer of the information about what it would take to make it better.
Q12 Chair: With an engagement level on leading and managing change of 35%, how supported and understood do you think people feel? That is only the people who bothered to return the survey. The other half of the sample did not return the survey. How understood and supported do you think they feel?
John Pullinger: Not enough, clearly. That is clearly a job that I need to do and the whole of the senior management needs to do.
Q13 Chair: How are you going to talk about that and lead that change in attitude?
John Pullinger: By systematically being consistent about the collective vision that we are all trying to fulfil to give people the confidence that they can challenge the status quo and not feel somehow threatened if they stand outside the box, to support people generally, and to push decisions down to a level that is much more sensible and consistent with good accountability, so that people feel that they have some control over their lives. A lot of the gradient in staff engagement, in many organisations, is that the people at the top feel very engaged because they are in control, and the people at the bottom feel very put upon. It is turning that around, so that management is there fundamentally supporting the people who are doing the day job, helping them be successful in their jobs, rather than issuing instructions that they feel are somehow taking away any sense of autonomy or worth in their life.
I said I am not launching a change programme. I think change programmes in ONS in particular have a bit of a bad reputation, and the same is true in some other organisations as well. This is about how people feel, as you are rightly saying. I can only do that by the way I behave and the way I encourage my managers to behave. When people start feeling it, they will start telling me they do. I do not expect it to happen overnight.
Q14 Mrs Gillan: You referred to your relationship with the unions and really implied that the reason for the low returns on this exercise was that the unions instructed their members not to co‑operate. Do you now have an understanding with the unions that, in future, they will encourage their members to respond?
John Pullinger: What I reported was what I was told. If tomorrow, when they have seen the transcript, they think that is wrong, they will correct me, and I will be happily corrected. This year, my understanding is PCS across the whole of Government has expressed concern about the people survey. That has been echoed within ONS. There is nothing specific going on in ONS this year—to my knowledge—but there was last year.
Q15 Mrs Gillan: I was going to say the union has not ventured any opinion as to what they would like to gauge the engagement, involvement and employee satisfaction. They have not put up an alternative methodology or questionnaire or anything they would be happy with.
John Pullinger: Not this year. Last year, PCS in ONS did have their own version.
Q16 Mrs Gillan: Are those results available in public?
John Pullinger: Not to my knowledge.
Q17 Mrs Gillan: Could we see whether they would be available? It would be interesting to see what they asked and how it came out.
John Pullinger: I can certainly ask the head of the PCS union in ONS if he would be happy to share that.
Q18 Mrs Gillan: For me, in coming in and guiding a new organisation, finding that there were such holes and gaps in what was happening, and possibly some unhappiness within the workplace, I would be setting some milestones. Have you set yourself some milestones, Mr Pullinger, by which we can gauge from the outside how you are progressing?
John Pullinger: The specific milestone in relation to this is to take those leadership and change scores up above the Civil Service average. That is a very specific milestone in this particular domain. Something I am seeking to do over the next six months is have a much more comprehensive range of measures against which I wish to be judged. In the staff area, they will be around this, where we have some metrics, but the most significant ones to me will be the extent to which those people who rely on our numbers feel we are being helpful to them. We already get an annual report from some of the key Government users about what we do. I want to think about how we can be much more sophisticated in the way we capture people’s views about what we are doing. Fundamentally, I want to be judged by what people who use our statistics are saying about them.
Q19 Mrs Gillan: When will that be ready for the public to look at? It sounds like work in progress.
John Pullinger: It is work in progress. My goal is that, when we begin the next financial year, we will have this properly articulated. I have been making some changes to the way the executive is run in the office that will take effect this month, so it will be much clearer that I am the chief executive of the whole organisation. As I said earlier, there was a separation between UKSA and ONS, so it is now in a more unified structure. Now I am looking at the metrics that will help us drive and improve performance across the whole thing.
Q20 Mrs Gillan: Will you be supplementing your team or changing your top leadership team in any way?
John Pullinger: Yes.
Mrs Gillan: That is fine.
Q21 Mr Turner: You have stated that your priorities for your time as National Statistician are the economy, business and finance, population and public policy, and capability. How did you come to choose these?
John Pullinger: They are from my reading of—it will be a familiar refrain—what customers are saying about us. The reports this Committee has done have been particularly helpful, particularly in areas like migration and crime, where there is a public policy need for us to do better. At the moment, I have no single person who is responsible for looking at the public policy needs and having the accountability to rectify them. That particular person will have responsibility not just for understanding what the needs are but also our data collection strategy that will help us meet them.
The economy stood out as the number one thing. There were some concerns expressed a year or two ago by the Governor of the Bank of England, the Treasury and others about what we should be doing and how successfully we were doing it. In my first month, as my report to the board said, I spent my time listening. I was talking to leaders in the Treasury, the Bank of England, the Office of Budget Responsibility and the Institute for Fiscal Studies, in saying, “Right. What do we need to do?” The economy stands out, and those are quite broad buckets. In our current structure, however, there is no single person who is actually responsible for bringing that together and making sense of it.
Having defined those two broad accountabilities that are fundamentally about the uses of the data, do we have the capacity at the moment to fulfil those? The answer is no, and the story of the website is a particular illustration of that. Much more generally, given the whole way in which the data landscape is changing—every day in the newspapers there is something new about it—I do not think we are sufficiently up the curve to do that well. A senior accountability to ensure we have the capability, both in terms of technology and people, to deliver that seemed to me to fit the bill.
How did I come to those? They came quite naturally to me from that listening exercise I conducted in the first few weeks. Three things I can cope with; more than three and my brain will start to get frazzled. In terms of engagement, something that is really simple works. In the Office, everybody knows that the economy is the number one priority of the Office now, and the people working on the economy feel pleased about that. The people working on the other side understand why it should be that way.
Q22 Mr Turner: Could you just repeat was the one word was that characterised the first thing you mentioned?
Paul Flynn: A spirit of “curiosity”.
John Pullinger: Yes, a spirit of curiosity throughout the Office. It is a very natural word in the community of scientists. You are always trying to find out more; you are always trying to do better. That will carry us a long way.
Q23 Mr Turner: What did you not prioritise and why?
John Pullinger: What you will see from this strategy is an increasing share of our resources going into the connection between what we do and the people using the information, and a much more relentless driving‑down of the costs of basic processing. Also, if it really gets tough, I would rather put money into the uses of the data than the collection of the data. Now, when we have a system that is fundamentally defined around production, the vast majority of our resources do go into collection and processing.
Now, I am being as open as I can be here, because you are asking me to give a choice and I am giving it. If we could get what we currently have used more effectively, we would add more value than just by collecting another incremental bit. The necessary consequence of that is, if there is no increase or even, as is entirely possible, a decrease in resources, I would prioritise that connection with the customer above the money that is necessary to collect the data, but that would be a very hard choice, because everybody would come screaming, including this Committee. You will say you want more people to go through the international passenger survey, more people to go into the crime survey for England and Wales, but my expectation is I will come to you with an argument that says you would be better served if I could make more sense of the information that already exists—let us stick with crime for the minute—from that survey, plus what is coming from the police, plus what is coming from somewhere else, than I can by an extra thousand people collecting information.
Q24 Mr Turner: Where does crime appear, either with the ones you did choose or the ones you did not?
John Pullinger: In public policy, that would be a changing landscape and that will require us to be much more in tune with what is happening in this place, for example.
Q25 Mr Turner: How would you decide what statistics should be cut—perhaps you have answered that—to redirect resources?
John Pullinger: There have been many efforts over the years to try to cut entire series, and they have generally been fruitless, because there is always someone, for a jolly good reason, who needs it. We also have a history of various phases in the development of official statistics in the UK where something has been cut out, only to be needed again. My expectation is that, as the policy landscape changes, some things will be prioritised and some things will be de‑prioritised. There will always be screams. One of the real challenges in this job is there is no market or quasi‑market mechanism that enables you to make the choice. All you have is people saying, “Not mine, please. Can I keep hold of that?”
I think the only mechanism I realistically have is transparency, which enables me to show that there is only so much money here; there are only so many things we can do. I would rather do the things that people want well and the things that people do not want much less well. If someone is in the less necessary box, then it will have to be reduced. In practice that happens anyway, of course. We need to be more open about it.
Q26 Mr Turner: You have said that the great problem is that most people may not want to do these things now but, in five years’ time, we will wish we had done it.
John Pullinger: I will have to have a weather eye to that and have a pretty careful look across the horizon to make sure we do not lose something completely. That is why I am saying it will be a question of de‑prioritising things. I do not think it will be losing things.
Take probably the most tricky and most significant thing for us. It is the changing shape of the economy. We still have a more significant share of our resources going into measuring the manufacturing industry, even though that is a much smaller share of the economy now than it was 20 or 30 years ago, than we do into the service economy. Yet we still care about the manufacturing industry. We still need, and certainly people operating in those sectors still need, the detailed information, but we have progressively been reprioritising between manufacturing and services, but at a pace that means we do not lose a vital part of the evidence base for manufacturing while gaining a better evidence base for services. It is that kind of approach that we need to continue.
Q27 Chair: How much of a threat to the usefulness and accuracy of statistics is the reduction in resources? One thinks of, for example, the endless wrangling and revisions of GDP data and the sheer importance of the purchasing managers’ index as a very important leading indicator, which we all know is just an estimate.
John Pullinger: ONS is not immune from the pressures that the whole of the public sector has, and we have faced our share of reductions along with others. If you have more resources, you can collect and analyse more information.
Q28 Chair: The purchasing managers’ index is a very clever indicator. Should we be concentrating on cheap and clever, rather than expensive and voluminous and not so useful?
John Pullinger: Maybe. We should be concentrating on useful, certainly. Something that is hugely consuming of resources but does not actually add much to the sum of human knowledge should be in the de‑prioritisation box. Many of these other indicators have, as their kind of bedrock, some of the official statistics that we are producing. There is a kind of ecosystem there. A principle that potentially will serve me well is that, if there is some market mechanism somewhere else that produces something like a purchasing managers’ index, I do not need to replicate it. We are spotting those parts that the system is not otherwise meeting.
Q29 Paul Flynn: Pursuing my spirit of curiosity, could you tell me if you think that the hope in the Act that set up the Statistics Authority, which was to persuade politicians, particularly Government, to be less cavalier in their use of statistics, has been achieved and been successful?
John Pullinger: Successful, yes; achieved, no. That would be my very short answer. It has been successful. In my paper, I describe as one of the successes of the last five years the role that Sir Michael Scholar and Sir Andrew Dilnot have played in putting a hand up and saying, “You shouldn’t mess with statistics. They’re far too important for that,” and doing that without fear or favour, coming in and making an intervention. It has made people more careful, and I would count that as a success. That is something we can bank and we can pursue.
Statistics are just another part of normal discourse, and they get used and abused in a natural human way. My concern is when it really undermines the currency. Before the Statistics Authority came into being, the reason the Act was debated and passed by Parliament was a concern the currency had been corroded and that was not good enough. There has been a successful turning of that tide and I expect that to continue. I observe on a daily basis that people are much more careful in the political space and also in the media space, but I do not think we can declare that the target has been achieved. It probably never will be, but progressively we can make it better.
Q30 Paul Flynn: Do you think that the Prime Minister was being very careful when he said that the Government is paying down its debts? You wrote about this and you gave the figures and very helpful charts on this, which show the debt as ever increasing. There are blue lines. This is a falsehood and, in fact, in the period involved, the Government has increased its debts by £435 billion over the original period. How would you describe this: a lie, a distortion, a falsehood, a mistake?
John Pullinger: On that occasion, the Prime Minister clearly misspoke. I heard him at another event a week later and he got it right; he said the Government is reducing the deficit. Whether or not that would have happened without the intervention, I do not know. We are in a spirit of general improvement here; the purpose of this is to make things better. I observed in that particular case an improvement in the space of a fortnight, in describing the numbers in a much more accurate way.
Q31 Paul Flynn: You can use language to distort the meaning, which I assume he did, but when he makes a clear statement, which you have rightly nailed as untrue, and says it is paying down its debts—at a Tory Party conference, I think it was—whereas in fact he is a small matter of £435 billion out in his claim, where do you get to the point when you have to embolden yourself, or the Authority has, to say, “The Prime Minister is lying; the Prime Minister is attempting to deceive the country”? The hope expressed by Andrew Dilnot before he had his present job, while the Bill was going through, was that that Bill setting up the Authority was the most significant Act of the last Government. There was a lot of truth in that—if it had worked, and if politicians of all parties were inhibited by the need to produce the objective truth, instead of going on some wild political hyperbole, which this was, surely.
John Pullinger: Each should make their own judgment, but my own judgment is that particular case was a success. There was a statement made. There was a reference made to the Authority to look at the validity of the statement. The Authority, as you have it in front of you, has made absolutely clear what claims can and cannot be justified on this, and the Prime Minister, in this case, is then accountable for whether he has misled or not. I think that is the system working. I think that is a very powerful act, and that would just not have happened had this Act not been passed.
Q32 Paul Flynn: I treasure a letter I received in answer to a worry expressed by a statistician that their work was going to be distorted by politicians. That letter was saying that this was a very unworthy suggestion that politicians would ever want to distort statistics and to pollute the work of statisticians. That letter was from Margaret Thatcher and I had it in 1988. Nothing seems to have changed, except the shock of politicians at any suggestions that they are misleading.
You wrote, on 9 October, to Mr David Hanson, about a comment by Mr James Brokenshire, who said, “We have cut net migration by a quarter since the peak under Labour.” You were asked how accurate or misleading that was. Would you like to comment on whether Mr Brokenshire’s statement gave a truthful impression?
John Pullinger: I am not going to comment on that particular case, but I am going to comment on the process, if I may. What we have with this Act, and what we have with the current Chair of the Statistics Authority, is a legal basis on which misstatements and falsehoods can be challenged in public. We have a Chairman who is very prepared to do that, without any fearfulness.
Q33 Paul Flynn: If you look at the truth of Mr Brokenshire’s statement, you have again provided very useful proof of this. The point he chose was one where there was a freakish increase in immigration and a drop in emigration. It happens to be there is suddenly a peak that goes up, and that is compared rather than any of the other periods over the 13 years. The clear trend you have here is one of increase and the graph is going skywards. The impression given by Mr Brokenshire was intended to be a false one. This is straight from the manual of how to lie with statistics. You have rightly, given the figures here, pointed out that immigration is increasing and has been increasing recently, and Mr Brokenshire’s statement is a falsehood. Can you come out? Do you think you can be less inhibited yourselves, nail them and say, “Look, the Government is lying on immigration; the Government is lying on the state of debt,” and challenge them in that way? This is the job that the Statistics Authority and you have been set up to do.
John Pullinger: From the Authority perspective, if I may say so, the kind of response you have in front of you is an even more potent critique of what was said than just a statement of an opinion. What we are doing here is putting out the evidence in a very clear way, which states our view on a particular case. That is out from our point of view. The extra thing is you now have that evidence as well, by which to challenge the Member on the Floor of the House or in any other situation.
Q34 Paul Flynn: You did another report or a comment in July about A and E waiting times and about how Government was further distorting the figures. We have a vigorous, hysterical campaign being run by one newspaper now, which is making a wholly unscientific comparison between NHS statistics in Wales and other parts of the United Kingdom. There was a scientific investigation of this carried out by the Nuffield Trust, which said that there was not that much difference between them, if you add them all up. There is now a campaign, which is propaganda dressed up as news, in the Daily Mail, which lies to their readers every day. This is then carried on by the other media, which reports what has been said. The impression is not one that is based on fact, science or any kind of objective examination.
Chair: Can you ask your question? This is a speech.
Q35 Paul Flynn: Do you as an authority feel that you should look at this and make a comparison between what is happening in the health services in the four countries, and report accordingly?
John Pullinger: If there is a case where the evidence base has been misrepresented or misreported, and that is brought to our attention, then the kinds of responses that you have in some of those other cases will be forthcoming. They will put the particular question into context.
Q36 Chair: A member of the public can report a newspaper to you, can they? Do newspapers fall under your regulatory remit?
John Pullinger: As you know, in the Act, I am on the production side of the business rather than the regulatory side of the business. We are very careful to respect that distinction. The fundamental job of the Authority is about the production and use of statistics in the public good. I can look up the Act if you like. That probably is not the best use of our time.
Chair: Okay. We will move on.
Q37 Paul Flynn: Are you satisfied that, when you do a report, it gets enough public attention? There are screaming headlines for the lie but, when the truth is published, it is either not reported or reported in very small type. Has the Statistics Authority built up a reputation where they are respected and they can beat down the distortions and hyperbole of politicians?
John Pullinger: I believe so. I think that has been a success, going back to the original question of the first five years, and I think we can build on that success over the next five years. We all have a joint interest in making sure that this sort of distortion is continuously reduced. The approach the Authority has taken to doing that has been a really good start, in its first five years.
Q38 Chair: In our report, we did criticise the lack of separation between the assessment function and the rest. Can you explain to us what has been done in response to that?
John Pullinger: Yes. It was before I started, but the Authority had a very careful look at its governance after that report, and has streamlined and made it much more straightforwardly separated now. Until this summer, four separate committees were the core Authority governance. There was the Authority board; there was the monitoring and assessment committee; there was the ONS board; and the committee on official statistics.
There are now just two. There is now a regulation committee, which makes it much clearer that that is the function that assessment is doing. That has no executive involvement on it at all and is completely separated from my world. On the other side, there is now a national statistics executive group, which I chair. That is the counterpart. That is an executive body, which is responsible to me for the production and delivery of the statistics. The whole thing is overseen by the UK Statistics Authority board, which is chaired by Andrew Dilnot.
Q39 Chair: Just to be clear, there are no executive members of the assessment committee.
John Pullinger: Correct. Not all of the non‑executives are on it, but the non‑executives also meet in private without me, when the main board meets, and I am a member of the main board. That separation is about as clear as it could be now and that is a change since the Committee considered this before.
Q40 Chair: On the question of designating statistical data, in our report, we were unhappy with the label “national statistics” as the kitemark of quality because, to the layman, it does not mean anything about quality. It just means it is something to do with the Government or official statistics. Have you thought about what, for example, Michael Blastland told us? He said that: “an understanding of the distinction between national and official statistics does not really exist in the public at large and it does not exist at ministerial level either”.
John Pullinger: That is beginning to change. One of the things with creating an identity—a brand, if you like—is that it takes time to get some traction in the minds of the people you are seeking to get that identity with.
Q41 Chair: Your own Chairman said, “National Statistics does not mean anything to me. It sounds as though it relates to the whole country or the ONS.”
John Pullinger: He did say that. My observation, certainly since I have come into the job, is that national statistics as a kitemark is now being held up as a gold standard. “Gold standard” was a phrase used in relation to police‑reported crime statistics in this Committee. People are wanting it as the sign that this has that standard.
Chair: You think it is something we just carry on working on?
John Pullinger: That would be my view, yes. Frankly, messing about with this now would be another distraction from getting on with business, in my view. If there was a feeling that this thing was somehow getting in the way—
Q42 Chair: You have decided not to look for an alternative term?
John Pullinger: The decision at the moment is to make it work and reduce confusion as much as possible by primarily talking about this.
Q43 Chair: We could put the word “quality” in front, Quality National Statistics. Is it a QNS? Yes.
John Pullinger: You could do.
Q44 Chair: I still think it is something that you need to perhaps adjust.
John Pullinger: We need to look out for it. We are measurement people. We currently have our survey looking at public attitudes towards us, and we are asking what people think about the label.
Q45 Chair: You have not completely closed your mind?
John Pullinger: No. I have never closed my mind to this or anything, but I am just saying, at the moment, this is not top of my list.
Q46 Greg Mulholland: Mr Pullinger, the ONS website is dreadful, isn’t it? We recommended previously that it should be improved, but it has not. It is not being visited. Numbers are falling. What is going on?
John Pullinger: At its simplest, we are starting again. You have described it as not being improved. One of the lessons from the past has been that there have been several attempts to improve the website. It was actually conceived before the latest iterations of technology really came into being. Cutting that knot and starting again will serve us much better than trying to build on what we have had before. It is not good enough. I said so before I came into the job and I say so again. In my spirit of being the harshest critic, you can be confident I will be the harshest critic of the website until it actually works for the people who use it.
At the beginning of last month, we started again and we are now working much more actively with the Government Digital Service and with a private sector partner that is much more in the space of the modern generation of websites. We are very actively working on what we are describing as an alpha site—the new system that, I am hoping, within a matter of weeks will be able to be tested. It may not surprise those of you who know me from my past life that I have given the House of Commons Library the task of trying to test this thing and make sure that it is useful to people who need to use it in anger. Others are welcome to be part of that testing community. The test of this new website is whether it hits the spot for the people who use it, and that was the fundamental problem with the existing one.
Q47 Greg Mulholland: That sounds positive. Can I be clear? Is it going to be an entirely new website with a different way of presenting the data? Are you starting from scratch?
John Pullinger: Yes.
Q48 Greg Mulholland: Could you give us a sense of the time scale on that?
John Pullinger: This alpha phase is running between now and Christmas, and that will test some prototypes. It will have some real data in it. It will have a real search engine in it and it will show whether the thing hits the mark for the people who rely on it. We are testing that with a variety of real people, but also around some personas—a word the web community uses—of the archetypal users: an expert user who just wants to get the data, someone who really wants to understand the analysis or just a casual member of the public wanting to know what the latest number is. They are the three broad measures. We will have that by Christmas.
If that is successful, we will then move into a beta phase, which will progressively be operational during the course of next year. What I am being told is that, by early 2016, we should have something completely operational, but my experience of websites is that, by then, we should be in another iteration mode. There is the phrase “continuous beta”. Once we get into the beta phase, progressively, people will start using the new one, rather than the old one, because it will be better for them. It will ramp up during next year and be ready certainly early in 2016.
Q49 Greg Mulholland: I am not entirely sure that talking in terms of alpha and beta phases makes something that has been described as “like venturing on the magic roundabout” particularly clearer, and I would advise perhaps finding a more user‑friendly way of describing the progress. You are saying it is going to take all the way until early 2016.
John Pullinger: The challenge with the ONS website—and it is true for most websites, but it is particularly true for us—is you are trying to jump from one moving train to another moving train. You look at the list of statistics publications. There are almost 100 publications, and some of them are really blockbusters, so crime, inflation and the labour market last week. These are very large things that we need to get right first time, at half past nine, every morning they come out. Before having something that is fully operational, we will want to test it to absolute destruction, so that we do not lose something that is absolutely critical to the market and to the policy community.
Before we switch the old one off, we need to be confident of all the heavy‑duty aspects of that, but also the security aspects and other things. We can build a prototype, which is the better word for an alpha version, in a few weeks, which we will do. We will then progressively build strength to it during the year, but moving some of these big things across we will really want to do carefully.
Q50 Greg Mulholland: Can I just ask you about the data.gov.uk website? It seems that, to some extent, there is confusion, because there is the ONS website, but then there is the data.gov.uk website, which was launched in 2010 and has all sorts of data sets from Government Departments. How do you see those two linking and do you not share the concern that there may be confusion as to where to go to find out credible, reliable, national statistics?
John Pullinger: There has been confusion and we need to get rid of it. The two have been independently conceived, and a central tenet of the new approach that we are adopting is that we will work hand in glove with the Government Digital Service. One of their senior staff is seconded into my team to help make sure that happens. From the user’s point of view, they do not know because, as you say, they do not care. You do not want to be faced with the choice, “Do I go here or do I go there?” You just want to know what unemployment is or whatever the issue is. We are conceiving what ONS is creating in partnership with the Government Digital Service, who do that data.gov platform.
Q51 Greg Mulholland: Are you saying that data.gov.uk would go and it would all be on the ONS website?
John Pullinger: Again, the drive is how people are using it. Data.gov and gov.uk cover a lot more than just official statistics. There is an intersection there that needs to be thought through carefully. Data.gov and gov.uk are fundamentally an exercise in transparency by the Government. They are the whole range of things that every Government Department is doing, led by the Cabinet Office, to open things up. Even in data.gov, there is a lot more than what we would call statistical data in there. It is information that can be re‑used by other people. The confusion comes at the intersection. There is no suggestion here that data.gov would be replaced by what ONS is doing, but there is a suggestion that what ONS is doing would actually fit within that bundle of things that officialdom is creating, and not be seen as something separated from it.
Q52 Chair: How much does the website cost?
John Pullinger: The running cost is about £5 million a year.
Q53 Chair: Quite a lot of money is being wasted here, isn’t it?
John Pullinger: No, because the website is publishing huge quantities of information.
Q54 Chair: If it is not a quality product at the moment, it has not been a good use of resources. How are you going to reduce the cost of the website?
John Pullinger: At the moment, a lot of the cost is the fact we are nursing a legacy system, so a lot of staff time and technology time goes on keeping the thing going. A new one, conceived in a more up‑to‑the‑minute way, should be less of an overhead. At the moment, we have a website that is publishing numbers. We have published “Public Sector Finances” since we have been speaking this morning. The website functions for that; it is just jolly hard. I am taking the hit that that is a problem, but it does work.
Q55 Chair: Coming back to the engagement, supporting and understanding bit, there must have been people in your organisation who knew that this website was not working, was not right, who had views and ideas about how it should be improved, and who probably were not being listened to. How is your organisation going to learn from this experience and be seen to learn from this experience via your own people?
John Pullinger: The story on engagement is getting the best out of the people who we have. My priority around data capability reflects the fact that that is an area where we do need to bolster our leadership. We need some good leaders there. We have got some very good leaders there. We need some more. We have, in the last year or so, so before I started, hired in a significant number of professional web people, who are part of creating this new prototype site.
Q56 Chair: That has been your initiative, has it?
John Pullinger: No, that started before I came. Clearly I am embracing it and taking it forward.
Q57 Chair: They had already learned that they needed to fill some skills gaps?
John Pullinger: Absolutely, yes.
Q58 Mrs Gillan: Whilst you are renewing, refreshing and redoing this website, I have a particular bugbear, because there are so many statistics that are held out on departmental websites or on gov.uk. I also then do not really know whether to believe them or not, for the simple reason that the very Department putting up, for example, rail passenger number statistics, is the Department for Transport. Has some thought been given to your linkage to all these departmental so‑called statistics, and bringing them in, if not signposting them from your website, so that you actually have one stopping‑off place as a member of the public or an interested party; but also at looking at those statistics as they are produced and seeing whether they should be endorsed, queried or questioned? Is there a relationship between your organisation and those statistics that are produced out in all the other Departments?
John Pullinger: That is something we will develop, I hope quite rapidly, over the next few years. It is beginning, but I certainly recognise it as a problem. It links a bit to the Chairman’s comment about the kitemark. I am seeing more people saying, “We’re not sure about these numbers that are produced in this agency,” and then a conversation goes on: “If we want to be sure about it, we need to get the kitemark, because then we can be confident it is going to be properly regulated and, if someone is doing something bad with it, there is this mechanism to shout about it.” I am seeing that dynamic beginning, so I am hoping that will continue at some pace from now on.
Now, the mechanism that I have is I am the head of profession for statisticians working in Government Departments, so there is a head of profession for statistics in the Department for Transport, for example. I work very closely with that network thinking about how we can improve the quality of numbers that are produced, not just by ONS but by Departments as well. One of the plus sides of the British system is we have this richness. We do have a very well-articulated group of people there and those heads of profession are responsible for adhering to the code of practice for official statistics in their Departments, and most of them, including the Department for Transport, have quite a lot of statistics that are already badged as national statistics, but some which are not.
The debate between the people using the numbers and the Department that starts saying, “We think these numbers matter so much to public debate that they should carry this kitemark,” will give strength to the statisticians in the Departments to make the case in their Departments, but will also increasingly be recognised by Ministers as being something that is valuable for them. Instead of arguing about the numbers, they are taking the numbers as read. I have already seen that in some places. The discussion around crime statistics that resulted in crime statistics moving to the ONS was an example of this, where my predecessor and the Home Secretary both came to the conclusion that, “We need to have confidence in this; in this particular case, we will move them across.” I do not think that will be the case in all examples, but there is a mechanism there. As I say, I think it is growing and that people like you who care about particular sets of numbers will give the pace to it, so we make the changes where they are going to add most value.
Q59 Mrs Gillan: Just to go back to the website itself, would it be an idea to make sure that you do signpost all those departmental statistical blocks within the Departments and also say whether they have that kitemark or not, springing off from you? You are the natural place that you would go to. I went to the ONS website to have a look at travel data and rail data to find that it was mostly about international travel, tourism and so on. The section called “rail” bore very little relation to where I wanted to get to, but I could not even get to the Department for Transport statistics. I think there should be some sort of signposting and then some sort of classification as to whether these had got your approval, the kitemark, or not.
John Pullinger: On the basis of my own logic, the answer can only be yes. If I am starting with people using the numbers, you are interested in transport; you put “transport” into a search engine. If you cannot find it straight away, you want to be able to find it. That would be a test. If my test is the user test, until we meet that one we have not done it for you.
Q60 Chair: Just on administrative data, everybody raves about the potential of administrative data and, indeed, the census is going to rely much more on administrative data than the actual census data in the future, particularly if the census is to be phased out after 2020‑21. We recommended that ONS now scopes out and sets out a more ambitious vision for the creative use of administrative data. How are you going to do this?
John Pullinger: The centrepiece inevitably is the census. Since the Government accepted the recommendation of my predecessor that we should go ahead with a largely online census for 2021, but with a really quite aggressive look at the potential of administrative data to replace it thereafter, we have been working very precisely on that case. That case has now gone to the Treasury and they are looking at the numbers to see whether it stacks up in financial terms. We clearly think it does, but that is going to be something that will have to go through the mill of the public finance system. We know that is going to be quite a tough test to pass.
If we can get access to the data, the driver then becomes what questions we can most usefully answer, in short order, from administrative data that is likely to become available to us. Basic characteristics of the population seem to me to be something we ought to be able to do reasonably straightforwardly. We use NHS data on GP registrations at the moment. There is tax and national insurance data—relatively simple and non‑contentious datasets. With the right legal gateways, we can start looking at whether we can establish the number and character of the population there. That may help with some of these questions such as migration. Progressively, the strategy will be to build, between now and 2021, a complete collection system, at each stage testing the public acceptability as well as the technical viability of all of that.
Q61 Chair: There needs to be a plan for this.
John Pullinger: The plan is part of this business case that has gone to the Treasury. The viability of the plan will depend, to some extent at least, on the amount of money available.
Chair: There is a business plan, an implementation plan, with costs and budgets, in preparation and discussion with the Treasury.
John Pullinger: Yes.
Q62 Chair: What about the legislative changes that we need in order to make this happen efficiently?
John Pullinger: In conjunction with the Cabinet Office, we have been working on what the Cabinet Office describes as an open policy process that has been testing possible legislative options on data sharing. That, from our point of view, is reasonably positive. There seems to be quite good public acceptance around the idea that data brought together for specifically statistical purposes is something that could be supported in the right environment. There are relatively simple changes to the legislation in this space that fundamentally would be about simplifying the processes that currently exist in the Statistics and Registration Service Act.
Q63 Chair: What about the reassurance of the public that administrative data in Departments that has been collected for certain purposes, and might be regarded as personal or confidential, is not going to be misused for a different purpose? Presumably the legislation is going to be required in order to get over these kinds of barriers, but public consent will be needed in order to get those legislative changes. Have I understood that correctly?
John Pullinger: You have mostly. Public acceptability is absolutely essential to all of this and we need to test that very carefully, but the scope of the current Data Protection Act is quite careful in its description of how to treat statistics and research. For statistics and research, you are not using the data in order to find out anything about the individuals. You are using it simply to combine it with other information and produce aggregate statistics. The Act already recognises that as a purpose not inconsistent with the original one so, theoretically, there is not a new public acceptability test to pass, but we would always want to be confident that we are not going to do anything that is going to undermine public confidence in us, because we rely entirely on the confidence of the public for the collection of everything that we do, not just the administrative data. We will set ourselves a higher bar than a legislative hurdle. The legislative hurdle is much more precisely about the kind of technical ability to get access to the data, which is currently prohibited in some cases.
Q64 Kelvin Hopkins: Congratulations on your appointment. I am very pleased about that. I am sorry I missed your earlier answers. I have a particular interest in economics. I used to teach economics, but I also taught economic statistics at a modest level, and I have serious concerns that the ONS might be under pressure from Government to cut what it does. I would resist that personally very strongly indeed. I hope that you will do your best to preserve those valuable time series that might be under threat if the Government started cutting and burning and whatever. Those are my thoughts, but an open‑ended question, first of all: what further improvements to economic statistics are needed?
John Pullinger: The director general of ONS commissioned Kate Barker to do a review of our national account statistics last year, and she reported in the summer. That has given me quite a good starter for 10, in terms of the improvements that need to be made. The first thing to say is that her overall conclusion is that our national account statistics, at the moment, are of a good quality. She had in her team Art Ridgeway as her helper in all of this, who was the senior person in Statistics Canada, which is widely held up as a beacon in this space. That was encouraging for me.
There are a number of areas, however, where she really did prioritise improvement. A specific one, which we are working on very actively now, is looking at financial data. While in the macro‑economy you have the possibility at the moment of looking at what is produced, alongside what is spent, alongside what is earned, and you can calibrate your numbers with a reasonable amount of success, the financial side is less well articulated. Concerns around the numbers used during the financial crisis were a particular trigger here. What is described as the “flow of funds” is a priority for us, so looking at how money flows around the system. It is of particular interest to the Bank of England, looking at the new regulatory regime around the banks, so we are trying to look between us at what the project would look like to improve our understanding of the flow of funds through the system. Other countries are in exactly the same boat; they are all looking at what lessons we have learned from understanding the financial system post the crash. That is a particular highlight to me.
Other ones are slightly more technical. The Chairman mentioned revisions earlier. There is a process called supply use, so you are looking at what sorts of commodities or activities are used and where they are coming from. We have a gap in our current armoury in that space at the moment, which holds us back, namely information about what kinds of products businesses are buying, which enables us to track flows through the real economy. We used to have a survey that did that, but that was one that was a consequence of a reduction in a previous time and we have not had one since 2006. Reinstituting that is another priority and we are beginning that now.
There is a plethora of things. I raised those two as the top two, because one will help us with the financial side of the economy; the other will help us understand flows around the system. Those are two things we need to do where we need to spend money and we are prioritising that. I said to the Chairman when he asked about priorities that economic statistics are our number one; they do matter most to the country and we will make sure we put as much money towards that as possible. More generally—I have lost my train of thought.
Q65 Kelvin Hopkins: Governments are notoriously mischievous when it comes to economic statistics. They always want to paint themselves in a good light, but there is one area where there is an enormous difference between what is publicly admitted and what has been privately calculated. That is the tax gap. It is not directly your responsibility, but HMRC estimates the tax gap—that is between what should be paid and what is actually paid in terms of tax—at under £40 billion, and yet Richard Murphy, tax expert, now actually working with the Treasury in these matters, has estimated it at £119.3 billion to be precise, three times the size. The numbers are so vast that, when the Government is talking about saving £100 million here and £100 million there, we are talking about drops in the ocean compared with the tax gap. This is very important stuff, but the Government constantly plays down the tax gap and even leaders of my party, in opposition, do not make much of the tax gap. They talk constantly about other things. With financial statistics in that field, you have a danger of being too political, but could you not shed some light in those areas possibly and talk about financial statistics?
John Pullinger: Thank you for the question, because it helped me with where my train of thought had gone. What I was going to say is, more generally, we should be looking to move into analysing puzzles, where people cannot quite make sense of what is going on from what they see. One of the recommendations that Kate Barker made in her report was that we should establish a small team, probably based in London, that would bring in expertise that would be able to make sense of some of these things that people continue to ask about, in a way that is entirely political neutral, but just trying to get to the bottom of what is happening. Productivity is top of the list, but potentially the tax gap could be in that sort of domain, so that we are not left wondering what is really going on here, and you have someone like us that is capable of stepping in there.
I would want to be quite careful. It is very easy to get some of these things wrong, and so what I want to start with is just a new mechanism for bringing in the right experts, so potentially from the IFS, the Treasury, the Bank or from a consultancy, as well as people from ONS, just to sit them down in a room for three months. What is going on with this question? What new light can be shed on it? That would be a new approach. It is in keeping with the spirit of curiosity that I have described, and will help us systematically to get the economic statistics that we need to understand the problems we are worried about.
Q66 Kelvin Hopkins: It is more a role for the Chair of UKSA to be political and challenge Government when they misuse statistics, perhaps. Do you work well with Andrew Dilnot?
John Pullinger: Very much so. He is a numbers man; he is enthusiastic about numbers. He loves everything about numbers, so we have had very energised conversations on almost any subject to do with statistics. He is very good and I am very happy with the whole board of the Authority. They are very well aligned and very committed to making our numbers the best they could be.
What I am finding very interesting is this relationship between the regulatory function and the production function, which I sense we are beginning to see in a properly distinct but mutually supportive state. Andrew, in his capacity as regulator, is holding the hand up and saying, “Look, don’t mess with statistics. They’re too important for that. This is not good enough,” but that is then a spur to me to say, “What statistics do we need to answer these questions that people are arguing about?” What I need is a much slicker mechanism to be able to occupy those gaps and restore credibility in those areas where there is currently anxiety.
Q67 Kelvin Hopkins: I had the pleasure of interviewing Andrew for the Chair and also knew him 20 years ago, when he was director of IFS, when he did a superb job again. We have been very lucky in both respects. Some of my favourite bedtime reading is actually the Library’s monthly economic statistics. I do use it a lot, particularly trade figures. I have always worried about trade figures, because again it is an area where Governments are mischievous. They constantly talk about exports, but never talk about the balance of trade. You can have substantial exports but, if your imports are £1 billion a week more than your exports to the European Union, which is the case now, that is significant. I keep raising these things in the House of Commons, but again eyes glaze over and I never get an answer. The reality is we have a massive trade deficit with the European Union, which is equivalent to at least a million jobs being exported from Britain to continental Europe.
Martin Weale, now of the Monetary Policy Committee but who used to be Director of the National Institute of Economic and Social Research, said that the Government was almost Goebbels-like in its distortion of the argument that we would lose jobs while leaving the EU. Now, I am not arguing for or against leaving the EU, but they say, “If we left the EU, it would cause a loss of hundreds of thousands, millions, of jobs,” when if there was a complete trade stop, we would stop those million jobs going overseas immediately. It is crude stuff but, nevertheless, is it not the job of the ONS sometimes to point out to Government, “You can’t just do this. You can’t talk simply about exports without mentioning imports, because the balance of trade is more significant than exports on its own”?
John Pullinger: What we can do is explain what is going on in the most clear, rounded way. Our trade statistics release does a pretty good job of that, so that anybody who is wanting to get a rounded picture will be able to do so. If someone is putting out one side of an argument, then it is possible for someone else to put out another side. It feels to me it is the inevitable currency of politics that people are going to talk in a way that bolsters their argument.
It was always a great joy to me when a House of Commons Library research paper was referred to on the floor of a House by a member of one party, and then the same research paper referred to by a member of the other party, quoting a different paragraph. The great thing for me was that anyone who was actually interested in what was going on would go and read the research paper. That is exactly the challenge that ONS has. It does, I am afraid, link back to the website. The more we can make our presentation of the data much clearer and easier, the more readily people like you and others will be able to challenge if someone appears to be only using one side of the story.
Q68 Kelvin Hopkins: Years ago in my previous employment, I was an avid reader of trade statistics. The balance of trade is very significant. Have we lost some of the detail over time in trade statistics, in terms of exports and imports from particular countries? I know that everything is much more fluid with trade these days, but have we lost some of the detail? I do not seem to recall seeing quite so much detail as we saw years ago. We used to have a very detailed breakdown of trade back in the 1970s and 1980s.
John Pullinger: The detail is still there, but it is slightly different. Before the single market came in, there was clearly a requirement within the customs system to be monitoring the flows for policy purposes. Since the single market has been in existence, the statistical system has sought to, under European regulation, measure the same things. We are still getting the data but it is harder to really be as confident in it. What has changed is that, in the old days, in the 1970s, we would be carrying around our publications with all the numbers in. You do not need to do that now; you can get the data from a database. The publications are shorter, but the richness of the data is there. In fact, what you can do with it is dramatically greater, because you have the power of technology behind it. There is still very detailed trade data collected through HMRC feeding into the trade figures that ONS publishes.
Q69 Kelvin Hopkins: There was one more question. We have a serious national numeracy problem, I think. Lord Moser has produced a report saying that 50% of the population do not understand what 50% means. They cannot do percentages, elementary stuff, and yet Governments constantly bamboozle people by exaggerating—by suggesting that £150 million is a lot of money and glossing over losses of billions elsewhere. Is there anything that ONS can do to try to help the general population understand statistics easily? For example, I write articles for newspapers, and I do it quite a lot and I used to use percentages all the time. Now instead of 10%, I put one in 10, because I think people can understand one in 10 more than they do 10%. Is there anything that ONS can do to help with that problem?
John Pullinger: The short answer has to be yes, but there are also others who are well placed to do it. I had three years chairing the Royal Statistical Society’s “Getstats” campaign, which was looking at various events. You and I were involved in several of them in this place for Members, looking at how you understand the data coming out from school league tables, hospital waiting lists or whatever. There was quite a lot of take‑up, particularly amongst research assistants. The Royal Statistical Society is really well placed, as a charity that cares about statistics, to make some of those things happen. They can be confident of my enthusiastic support for those kinds of campaigns.
There are specific things that ONS can do. One of the roles that I have is as a member of the policy profession group in Whitehall. They are looking at the skills that policy officials need to have. Clearly statistics is one of them, so ONS is working with them to think about a curriculum for policy officials on understanding numbers. There are lots of ways in which ONS, because of its role inside the system, can help. In terms of leading campaigns, campaigning bodies are quite often better placed to do that, and the RSS is certainly a very good one in that respect.
Q70 Kelvin Hopkins: Just as a humorous finish, the first question I used to ask my students was always: what is the difference between a million and a billion? Most of them did not know that one is a thousand times bigger. It is that elementary—the ONS saying, “Millions and billions are very different, and it is important to know the difference.”
John Pullinger: You have to have people listening, haven’t you? One of the things the RSS has been doing is working with the curriculum authorities to define, for example, the maths curriculum post‑16, where the UK is an outlier in the number of children who have done anything mathematical after 16, so they are working with the curriculum authorities to fix that. Again, there is a lot more freedom for a campaigning organisation outside to have those sorts of conversations with Ministers. That is a place that is changing quite quickly. There is a recognition that there is a gap, but there the gap is in our education system. ONS does not particularly have a voice into schools. You and I can keep making this case. We are winning.
Kelvin Hopkins: I could go on, but I have had my fair share, I think. Thank you.
Q71 Chair: Can I ask particularly about trade data? We export a total, including goods and services, and invisible earnings, of about 43% of our GDP, but in terms of manufactures, it is only about 19%. Of that 19%, only about 10% our GDP goes to the EU, but that is overstated, isn’t it? Can you explain to us the Rotterdam-Antwerp effect?
John Pullinger: Rotterdam and Antwerp are entrepôts, where goods come in and then go out somewhere else. In our trade figures, we seek to identify the final destination, or the people filling in the forms are asked to describe the final destination. Oil is probably the most straightforward thing to think about. Someone who is bringing oil in from the Middle East to Europe will likely take it into Rotterdam. As far as they are concerned, the final destination is Rotterdam. It then gets put into a big tank and someone else takes it from Rotterdam to the UK, so the transaction of that bit of oil is Rotterdam to the UK. It is very difficult to disentangle that, because the oil is all mixed up, isn’t it?
Q72 Chair: This is very important because, in the debate about our place in the European Union, people regularly talk about the proportion of our manufactured trade that is exported to the European Union. In your quarterly release, you have just shown, for example, that Germany takes 21% of our manufactured exports; the Netherlands takes 16%, but quite a lot of the Netherlands’ 16% is in fact exports to the rest of the world, not to the Netherlands. Is that correct?
John Pullinger: The 16% will cover all of those exports where the final destination has been stated as the Netherlands. It does not, however, only go one way.
Q73 Chair: I appreciate that, but this is the hot figure that people are always using and then linking to the so‑called 3 million jobs, which UKSA has rightly rubbished. This 16% is an overstatement, because those goods are not really going to the Netherlands. That is just an interim destination. Quite a lot of those goods are being exported to the rest of the world. How are you going to disentangle this?
John Pullinger: I do not know that we can. We are measuring a transaction, and the person undertaking that transaction is telling us that the final destination is the Netherlands. They may well not know.
Q74 Chair: You cannot count it, but do you not have an obligation to produce, on a regular basis, alongside your quarterly release, an estimate based perhaps on survey data of how much is exported to Rotterdam and Antwerp that is actually being exported to the rest of the world, so that this whole debate about how dependent we are on exports to the EU is put into a proper context? Otherwise people will go on misusing these figures and pretending that these figures are exports to the EU, when they are not really exports to the EU.
John Pullinger: I can see where you are coming from. I just cannot think of a way to unpack it. It does work the other way as well, you see.
Q75 Chair: I appreciate it works the other way, but when people talk about exports, they are talking about exports. I appreciate that imports from the EU are overstated too; that is what you are going to tell me, but exports are what we are talking about. I just think you need to address this, because otherwise everyone is going to be using a figure that is basically overstated by, perhaps, I am told from the research that has been done unofficially, as much as 1% of GDP or even more, which is quite significant. There must be a way of estimating or publishing an unverified survey estimate. There must be some basis on which it can be estimated. Given that it is such an important and political number, are you not under an obligation to do so?
John Pullinger: We should be seeking to inform that debate.
Q76 Chair: There is nothing about it in your quarterly release.
John Pullinger: No, but what there is in the quarterly release is clarity about what we are collecting.
Q77 Chair: Yes, I know, but there is nothing about this lacuna in your data. Quite a lot of that 16% of manufactures that goes to the Netherlands does not actually finish up in the Netherlands or anywhere in the European Union. It is actually businesses exporting to the rest of the world.
John Pullinger: What we are getting is a statement that this particular crate has gone from here to the Netherlands, and the Netherlands has been stated as its final destination. If the final destination is somewhere else and has been stated as such, we will do it, even if it is going through Rotterdam. You are not getting the entire Rotterdam effect. The only thing in those numbers, if I have understood it properly, is we have those where the final destination of the Netherlands has been stated.
Q78 Chair: The fact is you and probably I have to say “if we’ve understood it properly”; do we not have to understand it properly? If there were a referendum in 2017, which there may be, this is going to be quite a hot topic. I think Parliament, through this Committee, would like you to give us an undertaking that you will address this.
John Pullinger: I will certainly look at it but, as I say, I am anxious about giving an undertaking that we can produce some numbers for something that nobody has been able to produce numbers for before.
Q79 Chair: As I say, numbers have been produced unofficially. Even if you referred to those unofficial numbers or found a way of referring, every time you do make a release, to what proportion of our exports are going to the EU, it would have a qualification attached to it, with some kind of estimate of how wide the qualification might need to be. Is that not desirable? Does that not put your figures in a realistic and more truthful context than pretending that they are what they are? At the moment, I think they are misleading.
John Pullinger: I would say they are realistic and truthful in terms of what they are saying.
Chair: I am sure they are, but they give a misleading impression.
John Pullinger: They do not give an answer to the question that you are asking.
Chair: They do not actually say what is actually happening. All they do is record what is recorded on bits of paper. They do not actually record what is happening to those goods that are actually being exported to the rest of the world.
John Pullinger: The question is whether it is possible to know what is actually happening.
Chair: I agree that is a challenge, but is it not a challenge that should be addressed? Otherwise, you are just falling into Mr Flynn’s category of inviting politicians to use statistics that do not actually mean what they want them to mean.
Paul Flynn: It is questionable that you should be using your time and that of your staff to support one particular contention for a referendum that might not take place. Surely if there is a dispute, and we know that the Chairman is not exactly independent on his view on Europe and the referendum, would it be premature to spend your time arguing if A or B is right on the referendum campaign when there is no certainty that it will happen anyway?
Chair: Sorry, that was not a question; it was a statement.
Paul Flynn: It was a question: is it a sensible use of his time?
Kelvin Hopkins: In a way, it is a question of truth versus distortion. The truth is we have a serious balance of trade deficit, a massive balance of trade deficit, with the European Union. The distortion is only to quote export statistics. Notwithstanding your referendum, that is significant for our economy and says something about the failure of our manufacturing and is something that we ought to do something about economically. If UKSA or ONS were able to say that the Government should actually say there is a balance of trade deficit—exports is one thing, but we also have a gigantic level of imports from the European Union, much larger than our exports—this is in a sense telling the truth as opposed to a distortion.
Q80 Chair: I think you are on a slightly different point than Rotterdam-Antwerp one, but could I invite you to write to us on the subject of the Rotterdam-Antwerp effect and give it some thought?
John Pullinger: Of course.
Chair: I think there is huge public interest in the matter. Even those who want to sustain the European Union should be interested in actual statistics.
Kelvin Hopkins: On that point, Chairman, your point makes the deficit even bigger.
Chair: No, it does not affect the overall deficit. I do not think that is correct.
Kelvin Hopkins: It means that our exports are not nearly as big to the European Union.
Chair: I see; it makes the trade deficit with the EU even bigger.
Kelvin Hopkins: Our exports to the EU are actually less than are being portrayed because of this, and that makes the trade deficit bigger.
Chair: I agree, so we both agree that this is an important matter.
John Pullinger: I will certainly write to you on that.
Paul Flynn: I must put in that our two Members speaking are leading lights on the “get Britain out of Europe” campaign or whatever it is called, back to the dark ages. They are shamefully misusing their position on this Committee to put forward a point of view.
Chair: That is something you never do yourself, Paul.
Kelvin Hopkins: The statistics are showing that the European Union is going back into the dark ages itself.
Chair: We will not get into that. That is not the purpose. We are only concerned about accurate statistics, and I am sure Paul would agree that we want accurate statistics. Indeed, I am surprised that you are not in favour of more accurate statistics.
Paul Flynn: No, no—I am all in favour of statistics, but not using the Committee as a platform for your own views.
Q81 Lindsay Roy: Good morning. How do you engage with the devolved Administrations in gathering relevant statistics and how do you ensure consistency and offer comparability?
John Pullinger: Within the system that we have, there is a Chief Statistician for Scotland, Wales and Northern Ireland, and the UK Statistics Authority is a UK body. Northern Ireland is slightly different, because the Civil Service is run differently, but the four of us operate as an inter‑Administration committee to try to ensure comparability where we can, but also to look at how we can ensure that the needs of the devolved Administrations are met. Particularly over the last year, we have had intensive discussions with the Chief Statistician for the Scottish Government, and increasingly also for the Welsh Government, looking at the implications of the future settlement. That committee system is in place, and we are engaging with looking at the data that is going to be needed for whatever future settlement comes for each of the nations.
Q82 Lindsay Roy: Can you explain why on inflation and price levels statistics are only made available on a UK‑wide basis and not on a more localised basis?
John Pullinger: Primarily because economic policy has been a UK‑reserved matter and the pricing systems are worked out to meet that need. If there is a need in Scotland, Wales or Northern Ireland for inflation numbers, then we would have to think about how to meet that. At the moment, the whole system is built around creating the UK‑wide estimates of consumer prices and the various other prices.
Q83 Lindsay Roy: Is it not just the home nations, the devolved nations, but also the regions within England that might want that information?
John Pullinger: Yes, potentially. Again, it is a choice and the choice needs to be determined by how it is going to be used. I am just saying that, to date, the primary uses have been at the UK level, rather than at a lower level of geography. Potentially of course you are right.
Q84 Lindsay Roy: Has the experience of the Scottish referendum influenced your decision as to how to proceed in the future in providing that information on inflation price indices?
John Pullinger: It will do, yes. That is what we are working through at the moment as part of the broader discussions around devolution, and how we create the statistics that will support whatever powers are devolved and how we ensure that there is some consistency across the UK.
Q85 Lindsay Roy: Can you give us a notion then of how you plan to make progress with your colleagues elsewhere and whether that would apply to English regions as well?
John Pullinger: We will be trying to work in tandem with the political process. As the Bill comes before the House relatively soon, we will be looking at the statistical requirements that will flow from that. On a practical level, I am meeting with my counterparts to look at what the emerging requirements are likely to be. As they become clearer, we will be thinking about how we fulfil them, what the costs of that are particularly, but also what the technical aspects are, and how we design the measures that will be required and how we actually make that happen.
Q86 Lindsay Roy: You have two indices: the Land Registry and the ONS. Why are there two and are they compatible?
John Pullinger: There are more than two. We have developed a suite of price indexes over the years to meet different needs. One thing I might have said in response to Mr Hopkins’ question is that looking at the requirements for inflation measures is something that we are currently doing. Paul Johnson, the Director of the Institute for Fiscal Studies, is leading a review for the UK Statistics Authority at the moment thinking about what inflation measures we need for the future. Inevitably, all the current political demands will be fed into that, and that will help guide us as to whether we need one, two or more. Fundamentally, you need as many as you need to inform the debate.
Q87 Lindsay Roy: Have you come to any initial conclusion?
John Pullinger: We are awaiting his report, which is due in the next few weeks, certainly this side of Christmas.
Q88 Kelvin Hopkins: Can I just touch on that point? We are finding at the moment, when prices apparently are very stable or you have low‑level inflation, that some people are still saying they are seeing price rises. Is it not important that we do have these different indexes for different sections of the population? For pensioners who do not drive, oil prices do not affect them so much as for people who do drive, and so on. It is important to make this point, especially to Government, which is constantly saying, “We are having great success,” when some people, because of the nature of their purchases, are suffering.
John Pullinger: We are certainly looking at that question as part of Paul Johnson’s review. There is one nice thing on our website, which is the personal inflation calculator, which enables you to put in the kinds of purchases that apply to you and to get out what inflation is for you, which is very personalised. As to the kind of level of the headline numbers we publish, that question is clearly something that is before Paul Johnson and something he is going to have to consider quite carefully. There are quite tricky trade‑offs in all directions but, as I said to Mr Roy, we are awaiting his report with interest and we will share it with the Committee.
Kelvin Hopkins: I have to say that the strengthening of the pound is helping my wine bill no end.
Q89 Chair: Can I just ask about English data? The Scots and the Irish and the Welsh seem to have available to them data about their GDP and their productivity, but we do not have the equivalent figures for England.
John Pullinger: We have some equivalent figures. Andrew Dilnot wrote to you recently just explaining what we have, but the drive has again been their use. After 1999, the Scottish Government decided that it wanted to get a richer suite of gross value‑added indicators and it put the money in and got them. They now complement what we are producing for the UK, England and the English regions, so there is data that covers the whole thing broken down.
Q90 Chair: We are now moving into a period where there is going to be greater fiscal devolution to the Scots. That may well be replicated across the other parts of the United Kingdom. Therefore, the Barnett formula as a means of distributing resources is going to become less important as they gather more of their own resources, which are sensitive to their own tax receipts rather than to decisions made about spending in England. It always strikes me as rather odd that the Scots should actually want a system that determines the level of health spending in Scotland by a vote on what happens to health spending in England but, there we are; that is what they want. There is going to need to be more data to assess the absolute fairness of the distribution of the block grants to each of the component parts of the United Kingdom based upon actual need. Do you think ONS can provide the Government with all the statistics it needs to make a needs‑based assessment for such a decision?
John Pullinger: Some of the variables are easier than others. For something that inherently has a geography around it, it is more straightforward. Things related to people are relatively straightforward. Some of the economic measures are just inherently difficult. When you have a system that has a UK‑registered company, it is much harder to track whether that is in Scotland, Wales, Northern Ireland or England, when you have never had the need to do that before. It is definitely something that we need to think about. Certainly the committee I mentioned to Mr Roy is thinking about how we get a system that creates good data for England, for Scotland, for Northern Ireland and for Wales to serve the needs of the various territories, but also, as you say, to look at the equity of the system across the whole. The more things are devolved, the harder it gets. In areas at the moment like education, crime and health, it is quite a challenge to get a consistent picture across the UK, because you now have quite different systems that are operating.
Q91 Chair: We also have very variable systems, increasingly variable within England, but we manage it. I do not think that is a unique problem. What assessment are you making of the data that will be needed to better measure the fairness of the distribution of funds across the United Kingdom?
John Pullinger: I have asked a group of my colleagues to look at the data for devolution needs, in the light of the emerging political debate at the moment. That can be fed into the discussions as they take place in the House and elsewhere over the coming months. I am currently gathering, for each of the policy domains, what evidence and what data we need to collect in order to inform that debate.
Q92 Chair: Is that not quite urgent, given the argument is going on now in the Smith Commission in Scotland, which is likely to set the framework perhaps for some considerable time into the future?
John Pullinger: As of now, we can only make the most of what we have. All the data that we have is being used by the Smith Commission and others to inform that. We can break down the data we have in various ways, and that gives the information. What I am talking about is how we have a system that is sustainable once the new powers are in place. For now, there is a new collection we can suddenly make that will change things.
Q93 Chair: Clearly the Scottish Executive champions Scottish data. The Welsh executive champions Welsh data. The Northern Ireland Executive is very interested in Northern Ireland data. Who is the champion of English data?
John Pullinger: I think that goes to a deeper political question that I might need to defer to you on. I am the champion of UK data, and I am trying to ensure we have a coherent system across the country, whatever level of devolution is voted for.
Q94 Mrs Gillan: When you look around the world at other equivalents to your organisation, how do you think we rate? Is there a particular country whose statistical operations you would like to emulate or do you think we really are at the top of the tree?
John Pullinger: We rate quite well, but Canada is generally held to be a particularly strong system, as is Australia and the Netherlands. Certainly I look to them to help understand where we can improve. I mentioned our review of economic statistics. Art Ridgeway, who was the director in Canada with responsibility for that area, was a core member of that team, so I am absolutely keen to learn from whichever country is good in different spaces, but those would be the top three that I would list.
Chair: Interesting. Thank you very much indeed. We have enjoyed our first session with you. We hope we have left you with plenty to think about and we look forward to having your Chairman with us at some stage in the future possibly, but maybe he is going to get away with it until after the general election. Thank you very much indeed.
Oral evidence: The Work of the Chief Executive of the UK Statistics Authority and National Statistician, HC 552 29