Oral evidence: EU migration statistics, HC 979
Oral evidence, 26 April 2016.
Witnesses
Sir Andrew Dilnot CBE, Chair, UK Statistics Authority, John Pullinger, National Statistician, and Neil McIvor, Chief Statistician, Department for Work and Pensions, gave evidence.
Chair
Moving on to the question of migration statistics, if we can complete this within the next 35 minutes I think that will suit us very well. We are going to be joined by Neil McIvor from DWP at this point. Please could you confirm your identity for the record?
Neil McIvor
Yes. I am Neil McIvor. I am the Chief Statistician at the Department for Work and Pensions.
Chair
Thank you very much. We produced a report on migration statistics in 2013 and we said that the data based on the annual passenger surveys were, “blunt instruments for measuring, managing, and understanding migration to and from the UK. They are not accurate enough to measure the effect of migration on population, particularly in local areas, and they not detailed enough to measure the social and economic impacts of migration, or the effects of immigration policy”. Are we right to feel that there has been no material improvement in the data quality of this over the last three years?
John Pullinger
That is for me to start: there has been an improvement but there has not been a breakthrough, because we are still fundamentally using the international passenger survey as the primary source. I think in some areas we have been able to bring together different data sources to get the richer picture. A specific example I would mention is in student migration, where we have worked very actively with the Higher Education Statistics Agency, the Home Office and a range of other organisations to understand what is the journey of students as they come into the country and leave. In January of this year we published a report that illuminates that. The specific question now about how you can get a richer picture for understanding what is going on with National Insurance registrations, I think gives us the potential to go a bit further. But we are a long way from getting the breakthrough that this Committee was looking for in 2013, which was for the statisticians to be able to make use of a comprehensive administrative system that actually tracked people coming into the country and leaving the country. I don’t think we are there yet, clearly, and unless we have that kind of system the statisticians will be in this kind of jigsaw puzzle role. We can make progressive improvements and we will make progressive improvements.
Kelvin Hopkins
In response to the PASC report, you said that the only way to get better statistics was via, “Government funding for further work, including the development of new or additional statistical sources”. As one who believes very strongly that we ought to increase public spending in these areas, I wonder what your view is. Your recent business plan mentions better migration statistics only as an aspiration. What exactly have you done in the last three years to secure funding for such work? Have you bid for money?
John Pullinger
I think that is me again. It was part of our spending review submission. One thing that I said to this Committee when I came before you shortly after my appointment is that I felt that, faced with difficult financial choices, I would prioritise better analysis over more data collection. I do not think the answer is dramatically scaling up the international passenger survey. I said that at the time and we stuck with that. But I do think the answer is in having a much better analysis and access to the data that is available to us. In our business plan we have allocated extra money to do that, to support a team that is based in ONS but works with Mr McIvor and equivalent counterparts across the Government so that we get the best possible information out of the sources that already exist around Government. The short point is we said we would bid for more money. We did bid for more money. We have some more money and that is part of what we are doing in the business plan in the years ahead. We will always have an aspiration to do more but we are definitely moving on.
Kelvin Hopkins
It is admirable that you want to do better analysis but surely the Government has a responsibility to provide for better statistical collection, particularly at borders?
John Pullinger
I would be cautious about the tail wagging the dog here. I mean, what happens at our borders is a question that is not really about statistics, it is about flows of people across boundaries and that is a policy question for the Home Office. In part of that, clearly, I would make the case to ensure that we collect the information that we can then turn into statistics, but I think I have a receptive ear for that case. But as a statistician I don’t think I can say, “In order to get better statistics we need to have a new regime at UK borders”.
Kate Hoey
Can I just ask, Sir Andrew, are you full-time?
Sir Andrew Dilnot
No.
Kate Hoey
How many days a week do you spend in the office?
Sir Andrew Dilnot
I spend on average at least two and a half days a week and some weeks, like this week, it is much more than that and some weeks it is a bit less than that.
Kate Hoey
How many other non-executives do you have on your board?
Sir Andrew Dilnot
Six at the moment.
Kate Hoey
What sort of time do they spend?
Sir Andrew Dilnot
We ask them to spend two days a month.
Kate Hoey
How much do they get paid?
Sir Andrew Dilnot
About £15,000 a year.
Kate Hoey
How much do you get paid?
Sir Andrew Dilnot
Half the Prime Minister’s salary, it is about £71,000.
Kate Hoey
£71,000 for two and a half days a week, and the others get paid, how much?
Sir Andrew Dilnot
£15,000.
Kate Hoey
£15,000 for two days a month.
Sir Andrew Dilnot
The deputy chair is paid somewhere in the middle and is expected to do rather more.
Kate Hoey
Why has no data appeared from the advanced passenger information records and the scan of passports and other documents that are used to enter the country? Are we not capable of doing that?
John Pullinger
I keep feeling we are close. That is a kind of very honest response. My team in ONS are working very closely with the statistical team in the Home Office and they in turn are working very closely with the border agencies and numbers exist. I don’t want to publish numbers as official statistics until we are confident of exactly what they show, particularly in this area.
Kate Hoey
What could they show or not show? They show us who was coming in.
John Pullinger
They would show us who is coming in and who is going out. The challenge is reconciling that with the concept we are trying to get here, which is: who is coming in and staying in? Most of the flows across borders are people going for a day, for a business trip or a holiday, and it is disentangling the purpose of the trip. A scan of a passport is not going to tell you that. The international passenger survey tells you that because we ask people, but the international passenger survey just tells you what people say. To get some good numbers we need to be able to link whatever the Home Office says to what people are doing. So HMRC and DWP data is very important here, because you have some evidence that people are here earning money and paying tax or claiming benefits or whatever else is happening. In a really mature system, which I hope we will get to in the next three or four years, I would get that entry and exit data from the Home Office, I would get the data from DWP, I would get the data from HMRC, and I would be able to tell you that, of those people coming through the border, these people are just coming for the day, these people are staying for three months to 12 months, and these people are here for what we describe as long-term migration, which is that they are in the country for more than a year. So we have not produced those numbers because I am not confident of the quality of them in adding light to what is already a very heated debate.
Kate Hoey
That is a very honest answer. How far do you think that the wide disparities between the National Insurance number data and the IPS data discredit the current way that we are trying to measure the migration?
John Pullinger
I feel because of the nature of the debate I need to be able to say something that clarifies the position. I published a note on 7 March explaining what we are doing to do that. In my response to the Chairman in a letter last week, I said I will do a clarification during the month of May, which is feeling scarily close now. I want to produce something that, again, gives some extra real information rather than fuel for debate, which I do think in the absence of something the debate goes on and that is unhelpful. If I may, I will invite Mr McIvor, who is the expert in trying to get to the bottom of this, to say a few words.
Neil McIvor
I am responsible for the National Insurance numbers. I think it is very important to understand that the National Insurance registration numbers and the international passenger surveys are counting very, very different things. Very simply, my numbers are counting the number of overseas nationals that register for a National Insurance number and the period that they register. Some of those people may have been legitimately in the country for a time, a number of years beforehand, some of those people might be seasonal workers coming in that just want to work for a small period of time, going back. So the only thing you can count from those statistics are: these are the actual number of people who have registered in that time. You cannot estimate durations; you cannot estimate the number of people that are currently in the country from those statistics.
Kate Hoey
What would you need to be able to make the public feel that when they hear statistics about migration that it is foolproof?
Neil McIvor
This is the work that my team, along with John’s team in the ONS and other Departments are working on at the moment. We are looking at a range of data sources and one of the difficulties is these data sources are operational data sources not designed for statistical purposes. They are very, very complex systems that often don’t talk to each other. So we have had to bring the data in using data engineers and expert statisticians to try to make estimates of how these different systems are interacting with each other. That is a piece of work that we are currently doing with John’s team.
Kate Hoey
Do we distinguish between people coming in from the EU countries and people coming in from outside the EU?
Neil McIvor
My statistics count the EU, the EU8, the EU2, and the EU15 countries as well as those outside the EU.
Kate Hoey
Within the next few months, are you saying that things are going to get better?
John Pullinger
I have committed to the Chairman to publish something during the month of May, so very quickly.
Kate Hoey
Is that going to have an effect on the referendum?
John Pullinger
We will publish the numbers, the information that gives us the best understanding of what is going on with international migration.
Kate Hoey
Yes, and you are not put under any pressure by anyone in Downing Street, anyone at the moment, given that there is a very important vote coming up?
Sir Andrew Dilnot
If anyone were to put pressure on us they would receive the response that that is just inappropriate. If they would continue to then I would ring the Chairman.
Mr Jones
Mr McIvor, just developing this discussion, we have been told by Mr Jonathan Portes of the National Institute for Social and Economic Research that, “to date HMRC/DWP’s actions in handling the analysis of the data that they hold appears to have been directed at averting embarrassment to the Government, rather than informing the public debate or assisting ONS”. Is there any truth in that?
Neil McIvor
No, I would dispute that. These, as I said before, are very, very complex data sources. There are limitations with those data sources and it is incumbent on me to ensure that any data that I put out into the public domain are understood, are fully contextualised and are correct and tell the right story. The information that I have at my disposal is partial on some areas and has full coverage on other areas. That in itself brings its own limitations and its own difficulties, and the data sources don’t always link together as one would hope. As a professional statistician, it would be ideal if every data source and every computer system talked to each other and I could press a button and tell you a number, but that isn’t how it works in my world.
Mr Jones
But you are clearly aware of the importance of this data now, at this particular juncture. In approximately six weeks’ time postal vote forms are going to be sent out for the EU referendum. I think everybody is agreed that this is extremely important information that needs to be in the public domain before then.
Neil McIvor
Indeed, and that is why I have a team of some of my best statisticians working alongside John’s team, and working alongside statisticians in other Government Departments to really try and get underneath this. But it is a difficult thing to do. It is a very complicated thing to do. It is nothing that can be done quickly.
Mr Jones
It is going to have to be done quickly, isn’t it?
Neil McIvor
Indeed, and we have been working very closely with John’s team.
Mr Jones
The referendum has long since been flagged up. What are Government Ministers doing to assist in this process?
Neil McIvor
This process is separate from any ministerial advice I might get. I am working as a member of the Government Statistical Service, alongside John’s team in ONS, to use my expertise and to build on the expertise in other Government Departments to get to understand what these figures are telling us.
Mr Jones
When you say it is separate from what Ministers are doing, clearly Ministers must be concerned about the lack of information?
Neil McIvor
As the Chief Statistician in the Department for Work and Pensions, I work under the codes of practice and I am free from political interference in this work, which is why I am working very closely with John’s people to understand these figures.
Chair
Can I just enlarge on this very important question because, Sir Andrew, you made it very clear that you would not allow pressure to be applied on you and, of course, it is in the code. Is this a formal code of statisticians to resist any blandishments or political pressures or—
Sir Andrew Dilnot
Yes. It is formally set out in the code of practice.
Chair
Even given implied raised eyebrows and, “Please don’t be too difficult on this, Neil”, how do we know that people in Neil’s position are not being put in an impossible position? I am not going to ask him because that would invidious, but how do we know?
Sir Andrew Dilnot
I will turn to John, because the National Statistician has a crucial role in this. The answer to that is clearly set out in statute and in our code of practice that any such inference would be unacceptable. In the code of practice anyone subjected to such pressure is invited to make contact with the National Statistician, who would be expected to deal with it if that was possible, and if that was not possible that I would deal with it. Of course, we cannot be 100% confident that no such activity takes place.
Chair
When, in the past, you have written a letter to the previous Secretary of State criticising his use of administrative data—I remember that—Mr McIvor, how would you have been involved in the run up to that discussion? Would you have been raising the alarm about something like that?
Neil McIvor
We have put in place over the last three or four years some very tight protocols in the Department for Work and Pensions, wherein my statisticians and my analysts are involved in the sign-off process of any statement coming out of the Department to ensure the accuracy of those comments and those statements. Now, occasionally mistakes do happen, we are all human, but I think we have got some very, very tight processes in the Department to mitigate the risk of that happening.
Chair
So was that an occasion where mistakes were made or advice of statisticians was ignored by special advisers and Ministers?
Neil McIvor
I am very comfortable that I am heard by the advisers to our Ministers and we talk very regularly about what should and should not be said in statements. Occasionally an odd word might go in that takes the context to a slightly different place, these things happen, but very, very rarely in my Department.
Chair
If you feel that something is going out where the data is not being used properly then you would go to John Pullinger?
Neil McIvor
That is the final line of defence but I would never have to because I am very comfortable, first of all, that my professional advice is respected and well looked at in my Department, but I also have a line to my Permanent Secretary as well who fully understands my role here. Fundamentally, I could go to John but I have never had to.
Sir Andrew Dilnot
The Bean review reports a little bit on this and says that in conversations they found a number of senior civil servants thinking of the code of practices being a shield that protects them. I think that is certainly what we want it to be.
Mrs Gillan
Could I just ask you a simple question then on statistics, which is a figure you must have? In the last 12-month period, for which you have figures available, how many National Insurance numbers were issued to European citizens that are not UK citizens?
Neil McIvor
We published this figure in February 2016 and in the year to December 2015 there were 630,000 National Insurance numbers given to EU nationals.
Mrs Gillan
The year before?
Neil McIvor
The year before that—well, the year to September 2015 was 655,000. I don’t have the figures for the year to December 2014 with me.
Mrs Gillan
So over two years 1.2 million people from EU countries, not the UK, were issued with National Insurance numbers?
Neil McIvor
No, there was a big rise in December 2014 after the cessation of the—
Mrs Gillan
Sorry, I just said “over those two years” the figure was 1.2.
Neil McIvor
Yes, the two figures I gave you are overlapping periods, one was year to date December 2015; one was year to date September 2015.
Mrs Gillan
Okay, deduplicate them for me, what was the total over that period?
Neil McIvor
I don’t have that figure with me at the moment but there was a rise of some 12% of EU2 nationals after December 2014, after the restriction of the working regulations. But many of those could have already been in the country for a year or two before that legitimately. What we should not do is look at using the National Insurance numbers as a measure of flows into the country because that is not—
Mrs Gillan
No, I am not. I am not trying to measure flows into the country. I want to know what the absolute number is of non-UK EU citizens that have been issued with a National Insurance number. They may have been here legally for two years, but they are indicating that they are now starting to work and they need to get a National Insurance number. Could you produce those figures for this Committee?
Neil McIvor
Yes, I could.
Mrs Gillan
Deduplicated over, let’s say, the last 10 years.
Chair
Could we hear Mr Pullinger, please?
Kate Hoey
Just on that, about the Irish. Do you consider the Irish living here as part of that, because they are different?
Chair
Could I ask Mr Pullinger to answer—
Kate Hoey
Well, it was specifically on that.
Chair
Do you have an answer to that question, Mr McIvor?
Neil McIvor
Not on the tip of my tongue.
Chair
All right. Mr Pullinger to answer the broader question. It is a perfectly fair question, Kate.
John Pullinger
We cannot answer your question but it would be most useful, I hope, for you and others if in the report I have mentioned in the month of May, we include it in that because then you can see the whole thing in a kind of context. One of the things I have wanted to do with this publication, one of the reasons why we promised to do it, is that when we publish one set of numbers in isolation that is immediately taken to have a meaning that it just does not.
Mrs Gillan
My meaning was just I wanted to know how many EU citizens—
Chair
But we are getting this data.
John Pullinger
But I can say I will tell you, if you are happy for me to tell you as part of a wider thing that tells you a lot more things, that would be—
Mrs Gillan
I would like Mr McIvor to provide those figures because he has those already. I don’t need them set in context.
Neil McIvor
I do have that answer now for you. In the year to date December 2015 there were 630,000 EU nationals; year to date 2014 there were 590,000 UK nationals.
Kate Hoey
Does that include Irish citizens?
Neil McIvor
To be honest, I am not sure, I will have to go back and check that.
Kate Hoey
Thanks, and perhaps you could let us know.
Mr Jones
I wonder, Mr McIvor, if you could summarise what information you hope to release before the referendum, and could you let us have an indication of what sort of timescale you would be producing it?
Neil McIvor
Very simply, I am not intending to release anything. I am working with John’s people at ONS to feed into this piece of paper, which is being led by the Office for National Statistics.
Mr Jones
But Mr Pullinger requires the information from you?
Neil McIvor
Yes, and we are working very close looking at a range of data and a range of data sources, so that we can pull together the most meaningful publications.
Mr Jones
Can you summarise that, please, and let us know what sort of timescale you intend to produce it in?
Neil McIvor
We are working on a daily basis with John’s team so I am working very much to John’s timescales.
Mr Jones
But the information that you are required to produce to Mr Pullinger, can you summarise that now, please, and can you tell us when you hope to be able to pass it on to him?
Neil McIvor
We are doing an ongoing iteration at the moment. It is not a case of I am going to pull something together and then hand it, gift wrapped, to John’s people at the ONS. We have been working on a daily basis with my team, with John’s team and with other teams across the Government Statistic Service, where we have been looking at a range of data sources, some that I hold in DWP, some that is held in Her Majesty’s Revenue and Customs, and we are doing a daily iteration of our analysis using a virtual team across Government.
Mr Jones
Can you say what DWP information you will be producing or trying to produce for Mr Pullinger?
Neil McIvor
Some of the datasets that I am looking at are the National Insurance numbers. I am also looking at benefit claims. I also have a very partial extract of some of the HMRC data, but that data is old data. We get it after the event. So we are looking at some of that information. But it is important to understand with those data I don’t have continuous periods of, for example, National Insurance contributions. I know that, for example, someone may have had a National Insurance number in year X and year Y. What I don’t know is if they stayed in the country between year X and year Y. They may have spent some time doing some seasonal work, gone home and come back again. So it is a whole raft of different information we are looking at. We are really shaking the tree here to try to understand what is useful, what isn’t useful. I am not doing that isolation. I am doing it across the Government Statistic Service, with experts across the piece, who understand the different elements of these data.
Mr Jones
What is your best estimate of when you will be able to provide Mr Pullinger with that?
Neil McIvor
I am providing information on a daily basis.
Mr Jones
Yes, how much longer will you be doing it?
Chair
When do you think you will finish this task?
Neil McIvor
Again, this is being led by the Office of National Statistics. On a daily basis we have a working group, a virtual team, which are looking at the information, coming back asking for different pieces of information. We will have the task finished in time for John’s note.
Chair
I do not understand why it is only the DWP, or is it HMRC, that issues National Insurance numbers.
Neil McIvor
The DWP issues National Insurance numbers for adults. HMRC have the policy for issuing to children. We issue the number and our operatives enter that data on HMRC data systems.
Chair
There is only data system?
Neil McIvor
The National Insurance recording number is a single-based system that cuts across—
Chair
So why is it so difficult to count them?
Sir Andrew Dilnot
I think the answer is that we know how many NI numbers there are. What we don’t know is how long people are in the country with an NI number and whether they have come into the country to get an NI number or were in the country already. That is why multiple datasets are needed to try to work out, of those new NI numbers that have been issued, how many of them are here and how many of them are here for less than three months, three to 12 months or longer.
Chair
Mr Pullinger, you seem to be giving an undertaking that, whatever the state of the data, you will be publishing something before the end of May?
John Pullinger
I have given that undertaking, yes. That will include whatever we have from DWP systems, connected with whatever we have from HMRC systems, connected with whatever we have from the international passenger survey and the Home Office.
Chair
Does this data need to be published before purdah?
John Pullinger
It does not need to be. We have very careful arrangements to ensure we continue regular publication of statistics according to normal timetables, but as it happens it will be. We have a scheduled quarterly migration statistics release for 26 May and that will be happening.
Chair
Can I push on? Question 15.
Paul Flynn
Do you think, gentlemen, this could be your finest hour, in that you are buffeted by all sides, by Government and by campaigners, but your splendid answers this morning reinforce your objectivity and independence?
Sir Andrew Dilnot
Thank you for that. I think things will go on getting better and better. Seriously now—moving aside from your very generous comments—we believe that, in the Statistics Authority and in the ONS, we have made progress in recent years, but there is much more progress that we want to make and we think the sunlit uplands are still ahead of us.
Paul Flynn
Indeed. I think the Chairman and other members of this Committee are open about their partiality in this area. I have to declare my devotion to objective truth as always, not to any particular line on this. But on the question of national statistics on immigration, the Home Office has recently published a report with figures that I believe were previously available, but they have listed in a prominent way the number of asylum seekers in each local authority area. Was this something you did or was it something the Home Affairs Committee asked for?
John Pullinger
We include the asylum data in our long-term migration statistics. The recent arrangements for Syrian refugees are now getting into our figures. The responsibility for collecting those does rest with the Home Office and, again, this is another area where we work very closely with colleagues in the Home Office to ensure that we can understand those figures and integrate them with others.
Paul Flynn
Those figures reveal a remarkable situation because, while there are no asylum seekers whatsoever in the constituencies of the Prime Minister or the Chancellor, and very few in the Home Secretary’s constituency, there are hundreds, sometimes thousands, in constituencies of Labour MPs. Is this something we need more light on of why top Tories don’t do asylum seekers, although they do pontificate on policy?
John Pullinger
My colleagues in the Home Office can produce the statistics and they will continue to endeavour to do so, and that enables you and others to raise questions like that.
Paul Flynn
On the belief that it is valuable for the asylum seekers and for local authorities, whose services come under stress, particularly education and health services, if there is an excessive number and a bad distribution of asylum seekers, isn’t it essential that the true picture is known? Why should one Welsh city have more asylum seekers than all of the south-east of England outside of London? It is an extraordinary situation, is it not, and shouldn’t you be highlighting it?
John Pullinger
We are seeking to produce the best statistics that will serve the public good, understanding what is happening. The fact you are asking the question suggests that you have some information that enables you to ask it.
Paul Flynn
One of the additional problems to this, while asylum seekers are distributed in an extraordinary way that puts extra pressure on certain constituencies, there is no way of knowing how the very large number of EU immigrants who come here are distributed as well. Apparently the disparity is just as great. The constituency that has 5,000 asylum seekers has 12,000 immigrants from the EU. Shouldn’t this be published? Are these figures going to be published in some meaningful way?
John Pullinger
We publish our international migration numbers to the finest level of geographical detail we can. For asylum applications there is a very specific legal process that enables you to identify precisely where people have gone. If you are looking at longer term migrants across the piece, people coming for work or study or for all kinds of other reasons, we don’t have tags that help us understand where they are. So we are piecing together this jigsaw.
Paul Flynn
Is there any way I can find out the number of EU migrants in Witney? I know there are no asylum seekers. Are there any statistics that will explain how many the Prime Minister has taken in his constituency?
John Pullinger
The annual population survey gives us data, which is broken down by local areas, of the number of people and their nationalities. I am not sure whether it goes down to Witney. It might go to Oxfordshire. I can check the level of that. The House of Commons Library is very good at spotting these things, so I refer you perhaps to my friends. But that is a very useful source in this, because that tells you how many people are in the country at any point in time. It gives you that kind of precision about the number of people who were not born in the UK. I had a parliamentary question from your colleague, Mrs Main, recently where we set out the number of people in the country during different periods. That survey is quite a good one for that. It may well give you what you are looking for.
Paul Flynn
Could I ask you, finally—you were under pressure from the Chairman and his campaign this morning—have you equivalent pressure from Government on you to publish figures and to publish figures on a certain date, or to delay the publication of figures? Are the pressures on you about equal from both sides or are there no directives coming from Government?
John Pullinger
I don’t feel there is pressure and there are certainly not directives. What I am seeing are people involved in a very intense political debate wanting to know what is going on. I think it is our job to help everybody understand what is going on. Inevitably people want us to be better at that, but I certainly don’t feel any particular pressure from either side. This is the job that we do. This is why we are set up and this is why we are before this Committee, because it is a very important job on behalf of the country. The more data we can get out—as Andrew has said, we want to do much better even than we are doing now.
Paul Flynn
Thank you very much.
Mr Turner
I have five questions that I think will generate a “Yes” or “No” answer, but we will see. When you release the data can you commit now to this Committee to release the full data, including details such as fully disaggregating years of arrival?
John Pullinger
If you want “Yes” or “No” answers the answer is going to be no because, as Mr McIvor has said, the data is going to be hard to get, so no.
Chair
Will you provide estimates?
John Pullinger
We will provide whatever we can. You talked about transparency in the beginning. I think it is critical that we are transparent. We will give to the public whatever we can get out of the data, in a way that will stand up and I can feel confident will give some useful information.
Mr Turner
Disaggregating between different country groups, such as the EU10, 15, etc.?
John Pullinger
Yes, because I know that is very important.
Mr Turner
Thank you. Disclosing how many of these National Insurance numbers are now dormant?
Neil McIvor
I think we have answered that. National Insurance numbers don’t become dormant; they are with someone for life. If someone then comes back into the country they will be using that National Insurance number, so we give someone a National Insurance number once, they can use it as often as they need to use it if they want to work in this country.
Mr Turner
Thanks very much. Identifying separately those receiving or paying benefits, tax credits and so on?
John Pullinger
I think no to that because our focus is really looking at numbers.
Mr Turner
Presenting the information in a way that others are able to use in publishing the reconciliation to other Government data, like the labour force survey?
John Pullinger
Yes, as far as we can, because that, again, is important for understanding.
Mr Turner
Thank you very much indeed. Could I ask one further question? The 3 million jobs that are happening because we are part of the common market, the EU, do you have any information on that?
Sir Andrew Dilnot
No, that is not so. The claim that there are 3 million jobs that happened because we are part of the EU is not a claim that I think can be legitimately made. That is something I wrote about two years ago. There is reasonable work, which was originally done by the Treasury, which says that there are 3. 3 million jobs linked to trade with the EU but there is no suggestion that those jobs are linked to our membership of the EU. The fact that the trade goes on with people who are members of the EU, I imagine the argument would be that that trade might or might not go on if we were or were not members of the EU. So it is a legitimate estimate: how many jobs are linked to trade with the EU? It would not be legitimate to say, “That means that if we were not members of the EU those jobs would cease to exist”. That is a point that we have been clear about before.
Mr Turner
Thank you very much.
Chair
Can I just come back to the NINOs question? Would you be able to tell if a NINO has not been used for 10 years? Would that be identified? This question about dormant NINOs, there must be a point at which you know it is a dormant NINO?
Neil McIvor
If my memory serves me correctly, we did do a reconciliation exercise a number of years ago to look at those National Insurance numbers that have not been used for a considerable length of time. It was a very long and lengthy process because, if we could tell you that, we would be able to answer the exact question that you asked.
Chair
How many NINOs are there in issue?
Neil McIvor
I don’t have that information.
Chair
Could you provide that?
Neil McIvor
Yes.
Chair
Because from my memory it is always very much larger than the population of the country, which suggests that there are far more people out there with UK NINOs than there are people living in the country.
Neil McIvor
There will be because, if an individual comes to do some seasonal work for one year and one year only, they will have a National Insurance number registered to them. That will stay registered to them even if they go away. If they come back again in five years’ time, they will be using the same National Insurance number to pay their tax and their contributions.
Mr Turner
Is that necessarily true that they will automatically get the same number or may they also get a new number?
Neil McIvor
I would have to speak to my policy colleagues on that to see the process.
Chair
Do NINOs lapse on death?
Neil McIvor
Again, I will have to speak to my policy colleagues about that.
Chair
Presumably, you have to know that the person has died?
Neil McIvor
Indeed.
Chair
There might be quite a lot of NINOs in existence where you do not know if they have died or not?
Sir Andrew Dilnot
Particularly in the case of people who have emigrated.
Chair
When we did our report on migration statistics in 2013 we were told that the net migration figure has a 95% confidence interval of 35%, which means—
Sir Andrew Dilnot
Thirty five thousand, I think.
Chair
Thirty five thousand, yes, correct, because the example was if net migration hit the target of 100,000 it might be 65,000 or it could be 135,000. There is a 95% probability that it is that far out. How substantially different would these NINO figures need to be before we finally say, “These APS estimates are not fit for purpose. They should not be designated national statistics”?
John Pullinger
What you can be confident of from me is that when we produce the numbers you will get our best estimate of what is going on. If that is inside or outside the confidence interval you will get that. Then it will be for the regulatory part of the authority to decide whether that is fit for purpose.
Chair
My question is directed at Sir Andrew for that reason. How bad do these APS survey estimates have to be before we say, “These are not national statistics. We don’t know enough”? The net migration figure, in particular, is a difference between two very large estimates, which makes it intrinsically a broad brush estimate. Individual country estimates are meaningless and yet we still call them national statistics.
Sir Andrew Dilnot
I think my view on this is that, with our regulatory hats on, we await with eager anticipation this piece of work that the National Statistician and his colleagues will be producing. When we have that we will look at it from a regulatory perspective. There is a range of possibilities. It is certainly the case that what the NINO numbers are measuring and what the IPS is seeking to measure are different things. The ISP is seeking to measure long-term international migration, people moving for more than a year to the UK. It is conceivable that the work that is being done by our colleagues will show that these numbers, the NINO numbers and IPS numbers, are consistent with one another but measuring something different. It is conceivable that they will not show that and, if they show they are not consistent with one another, then we will make a judgment on what we think about their quality. At present, the national statistics badge is still there because while 35,000, plus or minus, is a significant figure they are still giving us meaningful information; meaningful information that shows that the target, for example, is a long way from being met, even on the most positive interpretation of their confidence interval.
Mrs Gillan
Would it be fair to say that we are not in control of this? We don’t have the information about who is coming in, who is working, who is staying, who is leaving, because all the answers from the three of you have left me feeling that we are not in control of what is going on in our own country and we don’t have the knowledge with which to work.
John Pullinger
Knowledge is always going to be incomplete. We are in the business of estimation. We try to triangulate lots of different sources. So, in terms of the numbers of people coming into the country, when we looked at the accumulation of people over the last 10 years when we had the census we were able to check how close we were. We can check now the numbers we are getting in our published figures against the number of people residing in different parts of the country. So we are continually checking but it is an uncertain science.
Mrs Gillan
So the answer is, yes, we are not in control. We don’t know exactly what is happening?
John Pullinger
We don’t know exactly.
Mrs Gillan
Thank you.
Chair
Well, on that uncertain note, as much of the debate is about uncertainty, how appropriate we should end here. Thank you. Oral evidence: EU migration statistics, HC 979