13
Examination of witness
Darren Morgan.
Q244 The Chair: Welcome to this evidence session of the Youth Unemployment Committee. The meeting is being broadcast live via the parliamentary website. A transcript of the meeting will be taken and published on the committee website and you will have the opportunity to make corrections to that transcript where necessary.
We have one witness today, Darren Morgan from the Office for National Statistics. Darren, thank you very much. Before we started on the committee’s inquiry, you gave us a briefing on the background to ONS numbers. That was not a public session and we have asked you along today for the final evidence session of this committee. Please say a word or two about yourself by way of introduction.
Darren Morgan: Thank you, Lord Shipley, and thank you for the invitation to attend the committee. I am the director of economic statistics development and analysis at the Office for National Statistics. I have the privilege and responsibility to produce and analyse a lot of the UK’s official economic statistics, including on the labour market, which I am very much looking forward to talking about today.
Q245 The Chair: Thank you. You are preparing lots of statistics on youth unemployment. Could you tell us what you think the shortcomings are in the figures that you have produced and what the key factors are that you consider when developing and preparing the facts and figures on youth unemployment?
Darren Morgan: When looking at the key factors in compiling the estimates, starting at the beginning, definitions are very important: what is unemployed, what is employed, and what is economic inactivity? There are a couple of things we need to think about there. One is making sure that our data is as internationally comparable as possible so that we can compare with other countries; the second part of that is confidence that we are using international best practice. The definitions that we use are in line with the International Labour Organization, which is a common standard used to produce labour market statistics across the world.
Another factor we consider is the population we are measuring, and we use households for the population. We must think about how we collect that data from people, so we use a range of surveys and administrative data to measure the labour market. Our main survey is the Labour Force Survey, a household survey based on interviews with members of households from across the UK. Then we need to think about how we estimate people we cannot collect information from. We weight our results up from that large survey. Our survey collects information from 30,000 households and 80,000 people, but we need to weight that up to the population, which we do, to get an estimate for the UK as well.
On all the points that I have just made, we are aligned to international best practice. Nevertheless, of course, as you suggest, Lord Shipley, there are challenges when we do that, and one challenge is with sample size. We have a large survey and the 16 to 24 year-olds are a small part of that population. This year, between April and Juney, we interviewed around 7,000 16 to 24 year-olds, so that sample size is a bit smaller than for the rest of the population. You can imagine that getting hold of younger people can sometimes be a bit more difficult than for other age groups within a household; if we cannot get hold of the younger people within a household, we ask the parents or guardians or other people in it to respond for them. It is a proxy response, and we tend to get more proxy responses for younger people.
The student population is a challenging part of the community to measure because they are transient. They live in halls of residence, they are in private accommodation and they can be quite challenging. I think it is fair to say that younger age groups can have a lot of casual interaction with the labour market. They can have a lot more casual jobs and part-time work. We compile our estimate in line with international standards but of course, as you suggest, there are some challenges as well.
The Chair: Thank you very much for that. In the course of the questioning we will be expanding on some of the answers that you have given, but now we go to Baroness Newlove.
Q246 Baroness Newlove: Good morning, Darren. I have written down a few notes. Thank you very much. That was very clear for me as a lay person with the ONS. This question is really important. What is the headline rate of youth unemployment today, and how does that compare to the past few decades? How do you define the headline rate and what factors are considered when determining it?
Darren Morgan: It is hot off the press; we published the very latest estimates earlier this morning. The unemployment rate for those aged 16 to 24 for the three months between May and July was 12.9%. That is still higher than it was at the start of the pandemic when it was just under 12%, but it has been falling quite quickly from the highest rates that we saw during the pandemic where it touched close to 15%. We are seeing the picture for the younger people improving relatively quickly as we come out of the pandemic.
What has been often reviewed during the evolution of the labour market during the pandemic is how it has compared with the financial crisis of 2008-2009, which was perhaps the last significant event. Very interestingly, in the pandemic youth unemployment shot up far more quickly than during the economic crisis, but it never reached the highs that it did then. The unemployment rate hit 22% in the economic crisis. We are seeing it falling and recovering much more quickly. To put that into context, I think it took until 2016 for the unemployment rate to get back close to 13%, so a much longer period than it has taken to recover from the pandemic. As I said, we are already below 13%. The reason why I have mentioned that 13% a couple of times is that over the period since 1992 a stable low rate for youth unemployment has been around 11% to 13%.
Q247 Lord Davies of Oldham: Thank you very much for the statistics you have already given us, which are, of course, of singular importance. I am keen to discover how you define the issues of employment or unemployment. So many people we meet seem to straddle the categories, and I wonder if you are prepared to comment on that aspect.
I noted that you not only talked about the national figures here but emphasised that the methodology you adopt is in keeping with what other countries do. That is of considerable importance to all of us for understanding levels of unemployment in this country. I think that point alone is an important one that we all need to bear in mind when we look at this.
You very kindly addressed yourself very quickly to the issue of youth unemployment, which is the main issue as far as this committee is concerned. Are you detecting a greater rapidity of change, a greater incidence of change among young people than among older workers, or does it fit into a pattern?
Darren Morgan: I will go first to the first part of your question, on definitions. You are absolutely right that definitions are the building block for everything else. Employment is defined, as I think I said in response to Lord Shipley’s question, by international standards and best practice. It is basically defined as having done any paid work or having a job you are temporarily away from. That paid work could be one hour, but you are still classed as employed if you have done any paid work. The last part of the definition, having a job you are temporarily away from, has been important over the last year for obvious reasons, given the furlough scheme. It has meant that people on the furlough scheme have been counted as employed. They are still in employment.
Unemployment is defined in line with the International Labour Organization definition as not being in employment but you are actively seeking work and you are available for work. That is the definition of unemployment.
On the behaviours of the younger people in the labour market compared to the older group, I think it is fair to say—and it will be interesting to see how it comes up after the pandemic—that the younger age group tend to move jobs much more regularly as they are starting their careers. Their jobs tend to be much more transient and casual in nature than the older age groups and they tend to be split into a couple of categories. They are very concentrated in hospitality, for example. They are very concentrated in retail and jobs of that nature and, perhaps not surprisingly, those industries tend to have an online presence. They tend to be very concentrated in those areas. They can be quite a different group compared to the rest of the labour market.
Lord Davies of Oldham: I very much appreciate those answers and I am glad you introduced the concept of furlough. That is a feature that we have not had to wrestle with before, but it is clearly going to present itself as a significant issue, as you alluded to.
Q248 The Lord Bishop of Derby: Good morning, Darren. My question follows up from the previous one and asks you to expand a bit on definitions. How do you define economic inactivity? Out of that definition, how does that work when we are looking at data for economically inactive young people? Do you feel that the overall definition reflects that well for the demographic that we are looking at, young people?
Darren Morgan: It is a definition that is not overly widely recognised for the rest of the labour market, but it is incredibly important for younger people. Economic inactivity means being neither employed nor unemployed. The way I look at it in its most simplistic terms is that we are all in this committee room and, if we are not offering our labour to the labour market, we are just not doing that, and that is economically inactive—you are not looking for work, you are not in work. The reason why it is important to look at economic inactivity as well as unemployment for young people is that students and people in education may choose not to look for part-time work or any work, and they will be classified as economically inactive. You have a picture there. When you are looking at youth in the labour market, it is important to look at unemployment, but to get the full picture it is also important to look at the economically inactive, because a large proportion of young people are in that category.
The Lord Bishop of Derby: Thank you. The definition of economic inactivity seems to be about whether somebody is economically rewarded for their labour. How does that intersect with labour that generates economic benefit but the person who does it does not receive economic reward for it? In our category, that might be young carers, for example, or people engaged in voluntary work, which generates enormous economic activity for society but people are not paid for it.
Darren Morgan: That is a great question. If I can clarify the other part of economic inactivity, one of the definitions is that you are not being economically rewarded. That is part of it, but equally if you are looking for work and you are volunteering and you are employed by a charity or an institution, as you touched on, you would be classified as employed because you are offering your labour to the labour market. They might not be getting paid, but because they are being altruistic they will be counted in our employment figures.
Q249 Lord Baker of Dorking: I find the measure of youth unemployment rather confusing in the UK because there are NEETs as well and then you also have another number, those who draw benefit, and they all come from different sources.
In university technical colleges every July, we record exactly what happens to each student in our colleges. We can do that because there are about 500 or 600 students. We know exactly who is going to university, who is going to an apprenticeship and who is going to a job, and who is not and wants to become economically inactive. I have tried to persuade the Department for Education to adopt this system, and they are totally resistant to it. If you ask for a school’s unemployment record, it is almost two years out of date, so they measure two years later whether the student who left the school is in employment, which is not a judgment on the school at all, because they might have had extra training. Do you think there is any merit in the UTC system?
Darren Morgan: Lord Baker, what you were discussing sounds very robust, and this could be a potential data source. What you rightly say is that at the moment we do not use that information for youth unemployment. We collect a large household survey from the Labour Force Survey, as I said previously, sampling and talking to households to collect that information. You are quite right that we do not use that data source at the moment. I was not aware that it existed, so it is very helpful that you shared that with us, but it would be good to know a little bit more about it.
Q250 Lord Baker of Dorking: Yes, of course. I have a few questions about disadvantaged students and unemployment, The 12.9% is clearly an average figure, and I appreciate that. NEETs compare at about 9.8%. Evidence has been sent to us for 2020, when 16 to 17 year-olds were measured at about 14% plus, the figure that I think you mentioned. Resolute did a study of 6,000 students and found the figure was 20% and not 14%. We have also been sent information by a think tank called Gatsby, which analysed all the results of the boroughs in Birmingham, because in Birmingham each borough has to report. It came up with an average of about 10% or 11%, but then it had a chart that showed that if you spread the Birmingham from Sandwell near Wolverhampton to Warwickshire and Solihull, Solihull youth unemployment was 7% whereas at the Wolverhampton end it was 20%. Your figures do not reflect the level of youth unemployment for disadvantaged students, although you do manage to reflect the level of students who are BAME or an ethnic minority. I think that is one of the flaws in your system. I believe that the level of unemployment among disadvantaged students is very much higher than the figures you publish.
Darren Morgan: It is a very fair point, Lord Baker, and a very fair challenge. We cut our data into lots of different categories, into NEETs, zero contracts, disability, ethnic minority, but you are right, we do not capture disadvantaged children. I am trying to think why we do not do that, and I wonder whether it is definitional. How would you define disadvantaged? That does not mean we should not, but how would you define disadvantaged children? What qualifies someone to be disadvantaged?
Lord Baker of Dorking: Generally in the education world, it is those who get below level 4 in the GCSEs, 4, 3, 2, 1; that is a very considerable number, about 2 million students. In fact, if you go below 5 it goes to nearly 3 million. The number of disadvantaged students who are unlikely to get anywhere near a university is very high in our system, so you can define them and measure them.
Darren Morgan: That is helpful to know. At the end of this month we are going to publish recommendations and launch a new inclusive data task force at the ONS. That would be good to follow up on. This is a key part of inclusivity, so I would be very happy to follow up on this point with you and the committee. It is a good point.
Lord Baker of Dorking: The other point, and this is my last point, is that economic inactivity is a mystery, quite frankly. We find in UTCs in more prosperous areas that there is a little bit more economic inactivity because families can afford to sustain their children at home. In the disadvantaged areas that is well-nigh impossible to do. A few students may decide to go on, but they are on their uppers, quite frankly, if they are economically inactive in a disadvantaged area. The family does not have the money to support them and they try to get very part-time, very low-paid jobs. That is a consideration I shall leave with you.
Darren Morgan: Yes, it is a very good point. Thank you, Lord Baker.
Q251 Baroness McIntosh of Hudnall: I think this is linked to the question that you have just been answering from Lord Baker. You said earlier that definitions are the bedrock, the building blocks of the way that you analyse the data. The definition of unemployment or claiming benefits is also quite mutable. Could you talk a bit about how you measure claimants, who they are? In particular, as far as I can see, you do not distinguish between those who are unemployed in the narrow sense that they have no paid work, which is partly what the Lord Bishop was talking about earlier—and those who are employed but need benefit support because their pay is low. Can you talk to us about how your statistics are affected by those nuances?
Darren Morgan: Yes, I can pick up on the easier part of this first, where you mentioned about people not being paid for work. Let us talk about the claimant count first for benefits. We have claimant count figures from our colleagues in the Department for Work and Pensions, which seeks to measure the number of people claiming benefit principally for the reason of being unemployed—so we have that. It is based on administrative data supplied by the Department for Work and Pensions for claimants of jobseeker’s allowance and of universal credit who meet the conditions of searching for work. You touched on this a little bit. It does not seek to measure the number of people who are out of work as such, or the number of people who are unemployed, but it gives us an estimate of people claiming benefits. Historically, that has often had similar properties to unemployment, so it was a good indicator.
The challenge is that the claimant count has always been less representative of young people who are unemployed compared with other age groups. Many young people—and this touches on your point again—who may be searching for work would not be eligible for those benefits, such as students or 16 to 17 year-olds. Prior to the pandemic, the claimant count had also been made an experimental statistic because the rollout of universal credit had gradually brought a broader span of claimants into the claimant count as well. In addition, during the pandemic the benefit system has played a further role in supporting those who had income that was disrupted, which has meant that the claimant count has become even less of a proxy for movement in unemployment at this time. We are strongly steering people away from the claimant count as an indication of unemployment. The claimant count gives us a very broad picture of those claiming benefits.
Could you ask your second question again, Baroness?
Baroness McIntosh of Hudnall: I suppose this is quite a difficult political issue, so I would not expect you to opine on the politics of it, but the rhetoric around benefit claimants is often to do with an assumption that people who claim benefits are not working for whatever reason. In fact, we know that many of them are working and are receiving benefits because their income is insufficient to meet their needs. Is there any mechanism within your data for trying to make that point?
Darren Morgan: That is a very helpful clarification. It is challenging to draw that from the data that we have. The one thing that people could do to help would be very broad. The claimant count, as I say, gives you the totality of people on the claimant count. We split that up by age, industry and so on. You could look at some of the other measures that we have for those different categories for employment and unemployment, to see if there are any links there. I do not think we could make that distinction within our data as it stands. You are absolutely right.
Q252 Lord Clarke of Nottingham: You mentioned a minute ago the pandemic having the effect you described because of the rollout of universal credit. Has the pandemic had any other effect on your collection of data? It must have been difficult for you to do your household survey in the deepest depths of the lockdown. Should we treat the figures from the ONS for the last couple of years with a little bit more scepticism than we normally do?
Darren Morgan: Thank you, Lord Clarke, and you are absolutely right that it did have quite a big impact on our Labour Force Survey. The reason for that is it is face-to-face interviewing. The first contact with people in the households who complete our survey is face to face. During the onset of the pandemic we had to stop that. We basically reverted to a telephone-based survey. We do contact respondents in households in other interviews by telephone after that first contact, but we moved all our collection to via telephone. As you can imagine, the impact was we found very quickly that the people we were able to contact more easily tended to be people who owned their own homes because they had a telephone number and they were easier to contact compared to those who were renting. That caused us a bit of a headache, because people who own their own house can sometimes have different characteristics to those who rent. The people who rent can be younger, a bit more transient in their movement, and we were seeing that affecting our figures. We had to take some measures to address that. We changed our methods to help us adjust for that.
Another impact of the pandemic was to reduce the overall response rates. With the pandemic, people’s priorities lay elsewhere than helping us compile statistics, and that is completely understandable. There was a risk that our sample of households would become very low, but in the pandemic we doubled our sample size to maintain the level of people we contacted to be as close to pre-pandemic levels as we could.
The third thing that we thought was going to cause a problem—and we touched on it a little bit earlier in the session—was furlough. We thought that furlough could cause us a headache, but in the end it did not because people were still classified as employed if they were furloughed. That was so clear that people tended to be able to respond to our surveys in the right way.
There is one final thing I could say to reassure people of the data for the quality of our survey during the pandemic. There have been lots of challenges—but can you have confidence in our estimates? I think you can. Part of the reason for that is our colleagues at HMRC. We secured access to data through HMRC’s real-time information data, which is basically for workers on payroll. Every worker on the payroll is on that HMRC dataset. We have led on that indicator during the pandemic. It has 29 million people on it; what gives us some reassurance is that the patterns seen in our survey and HMRC’s data have been very similar, including for young workers. That gives us a level of confidence in our survey results as well. Nevertheless, we are stressing that there is increased uncertainty in our estimates compared to what there would be otherwise, outside of a pandemic.
Lord Clarke of Nottingham: Thank you for a full and helpful reply, and fairly reassuring.
Q253 Lord Storey: First, my apologies for being late but I had all sorts of technical issues trying to join you. I do not know if this has come up before. My question is about how we get effective regional variations and whether there are any ways that we do not think are capturing the regional scene? Are there ways that we can look at how we can present that regional data?
Darren Morgan: We do it in a couple of ways. I have talked a little bit about the Labour Force Survey; we use that survey based on a sample of addresses throughout the UK. We know interviews are conducted within a geographical area, then we weight that based on the population estimate of the geographical area. If I can bring that to life a little bit, effectively an estimate for the north-west is based on interviews that take place in the north-west. We know that. Then we weight them to the population size of the north-west. The biggest limitation we have on regional data is the sample size again. The more you keep the data with more granularity, the smaller you are estimating for, so that is a challenge.
What we are doing—and I think that is something we have been leading on to get better coverage this morning—is that alongside that survey data, as I just spoke about to Lord Clarke, we are also able to present the information from HMRC’s real-time information data, which is far more robust and allows us to cut the data into 370 different regions of the UK. Although it does not follow international standards—it does not do that in the definitions—it allows us to provide very reliable estimates for workers on the payroll at the regional dimension. The coverage this morning was that it is only London and the south-east and Scotland now that are below pre-pandemic levels with workers on the payroll, so our ability to develop regional estimates is far more improved now than it once was.
If I can say one final thing on the survey side, we are also currently developing a new labour market survey to become a successor to the Labour Force Survey. This will be online first but in mixed mode, so it is telephone and interviewing as well, and it will have a much larger sample size than our current survey. There will be around 120,000 households in that survey compared to the 30,000 now. We know how important regional data is as we move forward, and we hope to be able to respond to that very effectively, combined with the HMRC data and our new survey.
Q254 Lord Empey: Good morning, Darren. To some extent my question has been touched on already, but with your data can you effectively capture ethnic minority communities, including very hard to reach groups such as the Traveller community, Roma, and people like that? What steps might be needed to improve your ability to capture this data? It has huge public sector and public policy implications, which is why there is an emphasis on this type of question.
Darren Morgan: Thank you, Lord Empey. I completely agree with you on how important this is. This time I am going to touch on the challenges first, because there are challenges here, to be completely candid.
The Labour Force Survey covers households within the UK, based on a register of addresses. It does not cover communal establishments or those who do not have addresses, so straightaway you can see that—as you said when you listed the Gypsy, Roma and Traveller communities—some of those groups may be underrepresented in our survey, because they are less likely to have a fixed and registered address, and we would not cover them if that was the case.
In addition, while the survey collects a lot of detailed information about ethnic minorities, the sample sizes are very small and do not support robust estimates of youth employment by ethnic group. We produced estimates and we shared them with you back in March; we released them, but we put a hundred caveats on them when we did so. Nevertheless, it got a lot of coverage, understandably, because it is such an important topic. It is fair to say that this is an area where there are some challenges in producing reliable estimates, just because the sample sizes are so small in the number of households.
To help support estimates for ethnic minorities, if they are sampled, if English is not the first language when we interview them we offer them a few translation options, such as a family member or a friend who can help, or translation services. If they are selected, we are very supportive in making sure we collect that information. There is a lot of challenges at the moment and we do not regularly publish data on ethnic minorities. We do not do that at the moment.
The game-changer here is the new labour market survey that I just talked about. We are increasing our sample size to 120,000 rather than 30,000. We can design it in a way that allows us to think a little bit more about other groups of society that we need to better measure, and this is definitely one of them.
Lord Empey: I get what you are saying, but is it not the case that it could very well be that ethnic minorities may have geographical concentration and, therefore, your sample definition will not do that? I think it is important for our report to get our heads around this. I understand the absence of addresses and the definition, and that some of those groups move around. The ethnic minority one is quite an important one, but does it tend to be a geographical concentration of people? How do you reflect that in your surveys and statistics?
Darren Morgan: You are spot on, Lord Empey, if I may say so. That is why I think our new survey is a really good opportunity to build in a design of our sample to take account of what you are talking about to reflect that. At the moment, for the geographical breakdown in our survey, we do not think of more detailed characteristics in our geographical data as we design the sample. With the new survey, which will be next year, we will be able to design it in a way that addresses some of the points you are making. That is why I think it is a game-changer.
Q255 Lord Hall of Birkenhead: Good morning, Darren. The question I am about to ask may also be answered by your point about the new survey. This is a question about one sector: young people with disabilities. Does the data you are currently gathering effectively capture the landscape that they are facing or is this something that in your new survey you intend to improve? I know we all feel it is really important that we gather good, reliable data on this area.
Darren Morgan: Thank you. You are quite right, Lord Hall, that it is linked but our challenges for including the disabled are not as significant as they are for the ethnic minorities. We publish regular estimates for people with disability, and we do that as part of our release. If you are a young disabled person, you are as equally likely to be sampled and interviewed as anyone else. We have that, and that is reassuring. The one challenge we have is that we exclude those living in communal establishments, as I said previously. People with disabilities may be more likely than the wider population to live in communal establishments. If that is the case, obviously we are underrepresenting them.
Your point about the new survey is equally valid here. I think that our estimates are reliable for disability but we can improve, and the new survey is an opportunity to do that as well.
Lord Hall of Birkenhead: Darren, will you be looking in your new survey at those disabled young people who are in communal establishments and who are not captured right now?
Darren Morgan: The thing is how we collect it. You are exactly right. We are looking at ethnic minorities and at a more geographical breakdown generally, at disability and at age groups in a little more detail. Across the board, we are saying that this is a good opportunity to bring the survey to reflect where we are in society. Yes, I agree with you.
Q256 Baroness Blower: Darren, good morning. You have already covered some of the elements that are in this question, but I am going to put it to you anyway. The question is about whether the ONS effectively captures the situation for young people who are not in traditional full-time employment. In the nature of what you have been talking about, there will be a lot of people in part-time, zero-hours and temporary work. You have already covered retail, hospitality and so on, but other than those, are there other types of work that are hard to measure and, if so, what might be done to improve the way they are measured? For example, do we have statistics about how many young people there are working in care, who regrettably are often on zero-hours contracts or temporary arrangements?
Darren Morgan: We touched on it a little bit, but I think there is more detail in your very good question, so I will try to cover it as well as I can. All young people are equally likely to be sampled and brought into our survey, regardless of the nature of their employment or otherwise. It does not matter whether they are a doctor or whether they are working as behind a bar, they are equally sampled in our survey.
The types of work that you touched on represent a challenge to us. That is because someone may get income from work but they do not consider it to be work in the normal sense. They could be doing a little bit of casual or occasional gig work, a bit of retail, and when we ask them the question they might not think of that as employment—but they should be classed as employed, because that classes as employment. It comes down to the response that the person gives us in the survey, so it is a challenge. As I said, some people may not think of it as employment because of the nature of the work that they do.
We get information and publish breakdowns of zero-hours contracts though, and that is quite a robust estimate. That includes an age breakdown, given how important that form of employment is to young people compared to other age groups. For example, 9.1% of young people in employment have zero-hours contracts. That compares with just under 3% of all other age groups. If you are a young person in employment you are three times more likely to have a zero-hours contract, basically. We believe that we do that well.
I think that the way we collect information at the moment is reasonably good in this space, but is there more we can do? It comes down to the interview responses that we have to rely on and whether there is any more we can do to make sure we get the right response from the people who do the type of work that they might not think of as employment. We could think about that but I do not think it is not a step change. I think we do this reasonably well.
Baroness Blower: I think there is a question here about whether a zero-hours contract in anybody’s mind is actually employment, given that there might be a lot of periods of time when they are not working. There may be people who are not responding in the most helpful way to the survey simply because they do not think of themselves as employed because in law, strictly speaking, they are probably not.
Darren Morgan: That touches on the point I was making, Baroness Blower. You are right in the sense that they may not see themselves as employed, but for our purposes, even if they earned one hour in the previous two weeks during that period they are employed. The point you make is valid, yes.
Q257 Lord Layard: Darren, thank you. You have talked so far about two main problem groups: people who are unemployed or people who are essentially not doing anything. From the point of view of a young person’s progression in life, it is also very important if they are employed that they are getting training. A third problem group is a large number of young people who are in jobs that are not leading anywhere because they have no training. If we are thinking about the youth problem as a problem of people developing their skills, we would want to know how many people are not in education, apprenticeships or some other form of training.
I have been able to find out the number for the end of 2019 for 18 year-olds from your lovely publication called Participation in Education, Training and Employment, where there is this extraordinary figure. I must say it shocked me to the core that 29% of 18 year-olds are not in education or apprenticeship or a job, or any other form of training. This is nearly a third of our young people aged 18, on the threshold of life and not in any form of education, apprenticeship or training. I think that that has to be one of the starting points for our committee.
Do you have more up-to-date figures for that? Could you send us a note on where that publication is? The latest I could find was for end 2019. I guess it is not going to be much better now, but I think we would need an up-to-date figure if you have one. I wonder why ONS gives so much prominence to the data on NEETs. I have not seen a lot of public exposure of this other figure. Is there some reason you do not give equal prominence to this estimate?
Darren Morgan: To answer your first question, of course it would be a pleasure to double check and write to the committee with the most recent estimate on the figures that you just referred to. I do not have that at hand.
On the prominence and the regularity of it that you referred to, and the NEETs that you referred to as well, it comes down to the data collection. Rightly or wrongly, our data collection at the moment has been designed for NEETs, and we are collecting that information more regularly. It is easily defined, we are publishing it more regularly, and that is why it is given more prominence. I think the reason we perhaps do not give the analysis that you refer to as much prominence is that it was analysis rather than part of our regular data collection. We use the data we collect and then do a special analysis of it, which is perhaps why that is not as prominent.
I completely accept your point. It is a very fair challenge why it should not be more prominent and, again, I am very happy to look at whether we should do more of that technical analysis going forward. I think it is a really fair challenge.
Q258 The Chair: Thank you. If there is anything that you can add in writing, Darren, that would be very helpful to the committee, and specifically in answer to Lord Layard’s point.
We are coming towards the end of this session, so thank you, Darren, very much indeed for giving us the up-to-date information and the benefit of your professional advice. You talked about that changes are coming in the way you collect that data. From your personal standpoint in the collection of data, if you could change one thing about it what would it be, and is it going to be implemented as part of the review you are now undertaking?
Darren Morgan: I think you are absolutely right about what I said in my response. If I were to ask for one thing it would be around the Labour Force Survey, which has done its job but is very old. It has been around for ever. I think putting a new survey into the field, which has a much larger sample size, allows us to design it in a way that addresses some of the challenges that we have talked about a little bit today, while building on the strengths of the current survey. I think that is the one thing that I would wish for. We hope that we will be able to produce a monthly estimate of the labour market, which we do not have at the moment from our survey, so we can put the labour market with those detailed characteristics on the same footing as our other high-profile economic statistics such as GDP and inflation—a brand new survey, newly designed, with a larger sample size.
The Chair: Thank you. Just remind me, did you say when this is going to happen?
Darren Morgan: It is in the field and already in testing. We are testing it as we speak. We are going to start a formal parallel run next spring, so we will parallel the Labour Force Survey and the new survey next spring and we will produce analysis off that. In reality, switching over to the new survey is more likely to be 2023, but it starts next year.
The Chair: Good. Darren, thank you very much indeed. On behalf of the committee, I thank you for joining us. On the very first occasion you gave the committee some advice and have now updated it formally with the latest facts and figures, so on behalf of the committee thank you very much indeed for joining us this morning.
