28
Examination of witnesses
Mike Keoghan and David Freeman.
Q1 The Chair: Good afternoon, and welcome to our first oral evidence session in this new inquiry into where all the workers have gone in the UK labour market and workforce. Please introduce yourselves, Mike and then David.
Mike Keoghan: I am the director-general for economic statistics at the ONS.
David Freeman: I am head of the labour market and households division at ONS, and I work in Mike’s area.
Q2 The Chair: Thank you both very much for coming. You are absolutely in the best vantage point in many ways to answer a lot of questions that we have for you today. We are going to try to peel the onion in certain ways, looking at labour supply trends, long-term sickness, international comparisons, trends by sector, self-employment, and demographic trends. The reason I say that is I will start off with a very broad first question, so you can try to set up answers that you might come on to give.
If there are points that you would like to explore in more detail and give us further written evidence on, feel free to say so because we will not necessarily have time to go into lots of detail.
We will start with a very simple question: who has left the labour market? Can you give us a broad view and then we can begin to explore each section?
Mike Keoghan: That is a helpful steer at the beginning and, in addition to extra material that we may provide afterwards, we also have several surveys in the field relating to these issues, which are not ready yet but are likely to come to fruition over the next four to six weeks, so we will provide the committee with those once they are available.
Just before getting into the question, we thought it would be useful to place who is left in the labour market in the current labour market context. As the previous Prime Minister said this morning, in terms of unemployment we are close to 30-year record lows, at 3.8%. Employment is below the pandemic percentages at around 75.5% and inactivity has risen over the recent period and is above pre-pandemic levels.
When we look at employment, there are two trends or compositional changes that are worth discussing or raising. One is around a switch between self-employment and inter-employment. There has been a big change in terms of the real-time information that the HMRC provides in terms of jobs. That is at record levels. Self-employment has fallen quite a long way over the pandemic period, so there is a switch there. Some of that switch we think is classification and we can explore that in further evidence, and some of that we think is real.
Secondly, there is a switch in migration, which again, I see as part of the question set. The evidence suggests that net EU migration is close to zero. There has been a substantial increase in non-EU migration. That is the broad situation in terms of the macro position on the labour market.
In terms of supply, there is the inactivity position. There are about 281,000 fewer people than before the pandemic. There has been a reduction in activity of 521,000. That has been concentrated in the 50 to 64 age group, which has reduced by 334,000. The reasons that are given on this are largely to do with retirement but, in particular, long-term sickness.
David Freeman: Picking up on the point of reclassification of self-employees, that partly comes from the way that we collect the data when we ask them if they are self-employed. We think some of that reclassification was down just to people changing their minds, possibly due to the support schemes that were available at the time. There were also changes to the tax system, the IR35 changes around contractors, but we do think some of the reduction in self-employed at this time is genuinely people moving out of that sector.
The Chair: I would like to come back to IR35 when we come on to self-employment. Before I pass on to Viscount Chandos, may I ask you a question about data and its reliability? Are there any areas in what we are going to come on to talk about where you think that we do not have good visibility on data?
Mike Keoghan: The main data source here is the Labour Force Survey, which is a national statistic, and we are confident in what it is currently telling us. It is fair to say that during the pandemic period the LFS had some challenges. It was difficult to do in-person surveying, given the social distancing restrictions. However, we have been able to use alternative sources of data—for example, real-time data from HMRC—to supplement that position and improve our understanding. We have also been lucky to have a census, which has enabled us to triangulate where we think the LFS is.
We think that the LFS is in a reasonably good state. It can get better, which is why we are moving to a labour market survey. That allows us to have bigger sample sizes and reduce some of the biases, but we are pretty confident where we currently are in the LFS, given what other sources of data tell us. There have been changes in relation to the migration statistics and quite a lot of work there. As I am sure Dave will talk about, the current migration statistics are experimental and so we are in a less confident position there.
The Chair: We will come back to migration in a moment.
Q3 Lord Livingston of Parkhead: How do changes in the average number of hours worked affect these numbers? Do you set a standard given 35-hour assumption for a full-time person, or if two people are working 30-hour weeks does that make a difference, or take down or increase their number of hours? Reading through the data, I could not see references to changes in hours within employment patterns.
David Freeman: We collect that data directly from the Labour Force Survey, and we ask people two questions: what hours they usually work, and what hours they worked in the reference week. It is a measure of the amount of work they are doing, and those all then feed through, so we do not have a standard full-time or part-time number. You will see those come through into the data.
Lord Livingston of Parkhead: Can you give any indication as to what they show? If people are still employed, are they working more hours or fewer hours?
David Freeman: Part-timers are working slightly more hours. Full-timers are working slightly fewer, but if you average the two out it is virtually where we were pre-pandemic, but, as Mike says, because we have fewer people in employment the total hours worked in any quarter are below where we were before.
Lord King of Lothbury: Mike, would it be fair to summarise what you said as that we should focus on comparisons between the data just before the pandemic and the latest numbers you can get now, rather than look at numbers during the pandemic, when the LFS perhaps found it difficult to get as accurate numbers as normal?
Mike Keoghan: I will defer on the specifics to David, but I think the way that we have chosen to present the numbers in terms of the public communication on labour force data has been to talk about the pre-pandemic period and the current situation.
Q4 Viscount Chandos: The Bank of England’s most recent report highlighted a number of reasons that might lie behind the reduction in labour supply: long-term illness, early retirement and lower immigration. Which do you think of these are most relevant?
Mike Keoghan: They are all relevant. When we look at the data what we observe is long-term illness as the predominant reason behind the increase in inactivity. That said, there has been a large change in the over-50 group. We are investigating it further. We did an over-50s survey at the beginning of the summer, which investigated a bit more about the attitudes towards work. We have another survey out in the field that completes this week, which goes into much more detail about health issues in particular, and we expect that to be released at the end of this month.
Q5 Viscount Chandos: Attitudes towards work leads me to the question that has been at the back of my mind. A huge proportion of the workforce experienced a completely different lifestyle through the pandemic. How much do you think behavioural issues have changed attitudes to work and the trade-offs that people may be willing to make, even though furlough and the financial support that existed then no longer applies?
Mike Keoghan: This was the motivation behind the over-50s survey. We found that over 50% decided that they had retired, and they did not plan to go back. A number, 25%, said that they had lost their job and then retired, 25% said that they lost their previous job because of the pandemic and were not looking for work, 18% said they were furloughed and lost their job and were not looking for work, and 11% talked about illness and disability. There is some evidence from the previous survey that suggests that the moment of the pandemic forced a variety of behavioural changes on people. What we do not know is how durable it will be, and that again is the motivation for the current work.
Q6 Baroness Kramer: Have you looked at the impact of childcare? I can hardly open a newspaper without reading about another early years childcare programme closing. Anecdotally, many of my friends who fall into that 50s age group who might have gone back to work are now essentially stepping in and becoming the childcare and providing grandparent care. Is anyone paying any attention to this and whether it is very fringe or playing a significant role?
Mike Keoghan: I am not aware of that. David?
David Freeman: Nothing specifically. If you look at long-term trends in particularly females looking after family and home, that has been going down very consistently even through the pandemic. We do not have anything to say where the childcare for these people is coming from—if it is older people or not.
Baroness Kramer: Is that worth exploring or at least trying to be aware of in the survey work that you are doing?
Mike Keoghan: Yes, I think it is. What I cannot talk to is to what extent we picked that up in the current survey, or whether there is opportunity in the free text response to see whether it is there. Anecdotally, it is clearly a barrier to rejoining the labour market, so it is something we should be interested in.
Baroness Kramer: I was thinking about it from both ends, because sometimes people only look at whether the parents do not go back, but they do not also look at the impact on the older generation.
The Chair: I do not want to get too far ahead in our list of questions, but if it is the 50 year-old-plus bracket that you are looking at the most, how does that square with where the vacancies mainly are? In terms of those vacancies in those sectors, I am looking here at where the sectors seem to be most affected by vacancies. Is there an issue? One associates these sectors—especially food service activities—with mainly the 20 to 30 year-old plus age bracket, so what I cannot quite understand is how that is reflected in the spike of 50 year-olds plus who have left the market.
Mike Keoghan: There are two things going on. Some of the sectors are areas such as construction, which does have an older workforce, so it may well be part of that. The other element is that people move around, so what is happening—and this is entirely speculative—may be that in a high-pressure labour market, if there are jobs appearing that were for older people, others are moving into them and, therefore, leaving jobs available in other parts of the distribution.
The Chair: Are you going to do any work to try to work it out?
Mike Keoghan: We have a lot of work ongoing on vacancies. The centre of expertise in government on vacancies and skills shortages is the Department for Education. It has established a Future Skills Unit, whose role has been to try to pull together the data to identify which sectors have gaps, where the labour force may come from, what the skills needs are, what the education needs are, to enable planning for education and training provision.
Lord Rooker: What you are doing in answer to the question is pick up employer age discrimination against the 50 to 64s, simply because they do not normally recruit those ages. Would the figures pick that up, or would it just be people anecdotally saying, “Nobody wants to interview me because I am over 50”, for example?
David Freeman: We probably would not pick that up in the data. For the number of vacancies, we have just asked a sample of companies how many posts they are actively recruiting for. That is something that we could potentially do around age distribution by industry, but again it would not necessarily directly answer that question.
Mike Keoghan: The other issue that we find from the survey is a lot of this seems to be choice. It is not that people are looking for jobs; they are choosing to be inactive, so they are not out there looking for jobs and being discriminated against. They are choosing not to work.
Lord Rooker: That is the employee. What about the employer? Who is asking the employer, “You are short and you cannot recruit but do you have a policy about going for the 20s to 50s, rather than the 50s to 64s?” Does anybody ask that question?
Mike Keoghan: It is not something we have asked.
Lord Fox: This is not so much the point that Lord Rooker was making, but I think the point that Lord Bridges was making is: how easy is it to measure displacement? I think what you were talking about was displacement, so in a sense one group displaces another in the absence of the other. Is that measurable or is it in a sense statistically impossible to pick out, because of the nature of what you are looking at?
David Freeman: From the data that we have it would be hard to see. We have evidence that shows that there are more people moving from job to job from one quarter to the next. It was around 700,000 pre-pandemic and it is now about 900,000, so more people moving.
Lord Fox: Turbulence?
David Freeman: There is more turbulence, more movement in the labour market, but we do not have the level, the size of sample, that we need to be able to do anything in the detail you are asking about in particular industries.
Lord Fox: With the durability of the people coming out of the labour market, which Mike alluded to just now, there is some anecdotage that I have picked up that people who have left the labour market are now considering or trying to get back into it because of finances, if nothing else, and the fear of future financial issues. Are you picking up any sense that this is a permanent or temporary absence from the market, and will your future field work be where you are going to pick up that particular movement?
Mike Keoghan: When it was originally done there was a signal from the survey that suggested that people responding thought this was permanent. Obviously, things have changed in the interim so that is one of the purposes of the current survey: to try to get a handle on whether that is still the case, given the change in the economy.
Lord Fox: Looking at the nature of the people who have left the labour market, how many of them is it possible to reverse versus those who have properly retired or are too ill to work?
David Freeman: The 50 to 64 age group has gone into inactivity slightly more because of ill health—but again, it might depend on the job. Slightly under half have retired early, so again, that is potentially reversible. With the 65-plus it is more to do with retirement than ill health. If you look at the data we had back from the survey that might refer to the over-50s survey, around 40% said they would consider returning to work but would look at work that would suit them and their lifestyle, and be more flexible.
The Chair: Can you go into a bit more detail? For example, how many of them have savings over a certain amount? How many of them have mortgages? How many of them are university educated? I am interested in those things because this would reflect their resilience in the face of economic hardship and their ability to retrain or find a new job. Do you have that data?
Mike Keoghan: The question there would be: from the active cohort of over-50s in the LFS, would we have information about degree status, height of qualification and previous job?
David Freeman: I would have to go back to look at the questionnaire. I know that we have highest level of qualification as part of the over-50s survey and I believe the sample is a subsample from the LFS, so we might be able to do some linking back.
The Chair: Also, levels of savings. I do not know if you have that kind of data or if you can somehow find that out.
David Freeman: I do not think we collect that, because the survey is very much focused on the labour market status. We would not be picking up the level of savings.
The Chair: I will ask you a layman’s question. Obviously, the data that credit bureaux such as Experian collect has enormous amounts of data that can show what is going on across the market. Could you tap into that and use that to help inform what your insights are, or is that not allowed?
Mike Keoghan: We have huge access to alternative sources of data other than national statistics. I think you have raised a good challenge. We do not have the information to hand, but there are probably a couple of ways into this. One is, as you say, looking at alternative sources of data. The second would be to what extent we are able to match across that group to some of the household surveys that we run, to see if we can identify them there. Another one would be to see if there is anything we can do with respect to the census and the census microdata. I think you have raised a good challenge.
The Chair: If you could look into that, it would be extremely helpful. I know Lord Skidelsky and Lord Stern want to come in.
Lord Skidelsky: I will postpone my question, because it is about different ages.
Q7 Lord Stern of Brentford: This is about how to understand permanent and temporary. Do you have any panel data that you can use? Are you planning to go back to the people you spoke to in your over-50s survey? If this is a moment where people are changing and may go on changing, it is important to try to understand that from individuals over time. I know that panel data is expensive and there are problems with dropout and so on, but if you are long-term sick, presumably, if you go back, you could still be long-term sick, you could be better or you might have perished. It is important to understand, if long-term sickness is so important, how that is changing over time, and similarly with people who have decided not to work and then think about going back. It would seem to me that panel data is pretty important in this context.
David Freeman: I am not sure with the over-50s survey whether there is any overlap from one sample to the next. On the Labour Force Survey, we do have an overlap between one quarter and the next and the annual version also has a big overlap, so there are certain elements that we can look at longitudinally.
Mike Keoghan: I do not think it is big enough to offer anything. I think that was the hope, but the numbers have not allowed it to happen. We should provide you an answer on whether there is a parallel, but the point is a good one around looking at long-term sickness and the permanency of the issue. What we do see, though, is that in terms of the increase in the LFS figures on inactivity, long-term sickness is the biggest single driver.
Lord Stern of Brentford: Which is why it is so important to try to understand it more deeply.
The Chair: I owe you an apology, Lord Griffiths, because I began to ask your question.
Lord Griffiths of Fforestfach: I noticed that.
The Chair: My apologies. Overexuberance on my first day back.
Q8 Lord Griffiths of Fforestfach: How reliable is data on labour supply? Are there any gaps or areas where you are less confident about the reliability of data? There were three things that struck me. How much, in collecting the data generally, do you talk to people? One thing that my colleague Lord Lothbury has impressed on me is that we construct large econometric models of how the aggregate economy works, but how many people go out into the field, talk to businessmen, talk to people who are unemployed or changing jobs and so on? What I am not clear about, for example, is, if somebody declares “I am suffering long Covid”, is there any verification of that? Do you go back to check things? In a way this comes out from a point Lord Stern was making.
Secondly, from my point of view, is there a bag of gold of these surveys that you are doing? Do you want to say anything about the topics that you are looking at, which may be relevant to the way we think about things in the future?
Thirdly, I was very interested in an article in the Sunday Times—I think it was last Sunday or the Sunday before—by Irwin Stelzer, in which he quoted a very well-known professor of demography who was at Harvard, Nick Eberstadt, now of the American Enterprise Institute. This goes back to the question raised by Baroness Kramer about the relationship between the statistics and family structure. Eberstadt says: “The strong relationship between family structure and employment status for men is undeniable”. What they say about the unemployed is that, if you compare men who are part of a family structure and men who were not, there is a noticeable difference in the way they have registered in relation to inactivity. Do you have any information in this area for the UK?
Mike Keoghan: On your second question, could you say a bit more about what you are looking for?
Lord Griffiths of Fforestfach: Just the titles of what you are doing.
Mike Keoghan: In terms of the reliability and if we talk to people, I think David could talk about how the LFS runs.
David Freeman: The LFS is one of our biggest household surveys and the way that it currently runs is we go out to talk to people, so we have the random sample of households across the country. Prior to Covid we would have turned up on their doorstep with the questionnaire and gone through that with them. Post-Covid we are doing most of it on the phone, but it is still talking to the householder. If we can, we try to talk to every individual aged over 16 in the household. If we cannot get hold of them, we will talk to one member of the household who will then answer on behalf of other people, and so they get a standard questionnaire.
As you say, a lot of it is self-classified, so we do not ask people if they are employed or unemployed, but we will ask a set of questions that will let us determine that. Things like employee/self-employed will be self-classified. If they have a limiting disability linked to a long-term illness, it will be their classification of whether they feel that is suitable for them.
On checking things, we do not as part of publishing the survey triangulate against other data sources. We do some checks to highlight implausible answers such as people earning £30 million a day, or 85 year-olds with very young children and so on, but most of the data will go straight through. As part of publishing, we do then put the LFS against other measures such as the tax data and business surveys to make sure that it is making sense and we can tell the labour market story.
Mike Keoghan: To pick up on that in terms of the verification, it is about the quality of the data. There is a huge amount of effort within the ONS to improve quality and to make sure the quality of the data is good. We obviously had Charlie Bean’s review in 2016, which raised some questions. Since then, there has been quite a substantial investment programme to improve the quality and relevance of the data that the office produces.
I joined the ONS in January and I have been impressed by the processes; what they call the curiosity processes. It is not like any individual data release comes out without anyone looking at it. There is a huge amount of testing, kicking tyres, testing against other forms of data, testing against what market expectations are looking like, testing against admin data, so that we can make sure that wherever we come out we have real confidence.
We are in a good place, but we can always get better. The process that Charlie Bean established is ongoing and is working through the full suite of economic data that the office produces.
In terms of topics we are looking at, we publish forward look of research and survey work. It is called the “ONS release calendar”. It is probably not on the top of everyone’s Google bookmarks, but it is on the top of mine. It does give you an indication of what is coming out, what we are looking at. We produced something on marginal workers in the labour market and measures of underemployment in the labour market. That picks up some of the LFS data that David was talking about, the micro level. It uses different measures drawn from, in particular, different central banks and how they look at labour market slack to try to get a sense of how the UK is looking. That was done yesterday, and it is a good piece.
We have our surveys that are coming out, but we are interested in what is happening in the labour market. We have quite a lot of surveying and research work in train. The quickest way into that is to look at what we have said we will publish.
Baroness Noakes: Can I make sure that I understand the LFS? Is it the same set of questions that are repeated and then you get different answers in terms of the data and then you wish you would have known more about that, or are you constantly changing the questions? My concern is that we are seeing some quite significant issues emerging from some of the data. Have you asked enough questions during that survey to get to the bottom of it, or are you hamstrung by a set of questions that have been used for a long time, and therefore you are working in arrears? Is it real time or in arrears?
David Freeman: The current LFS is in arrears. We have a set of questions, and we try to maintain consistency so that we can get consistency over time. There are certain sets of questions that we will always ask, because they help us get to the fundamental core things such as employment and unemployment. We do revise questions, so we do have a regular process of adding questions in, rephrasing questions for when certain things come to light. We had a set of questions around Army veterans a few years ago when that was a big issue for the Government. We do have add-in questions, but it can take quite a while given how the current Labour Force Survey works.
It is a good point to say that we are in the process of transforming the Labour Force Survey, moving from the current interview towards something that is online first, with a bigger sample, and with the option of subsampling from that to do other surveys. The ambition is to be a bit more reactive to things that are happening and set a survey up quickly, a bit like the over-50s survey that we put in the field this year, in order to tackle specific issues in a much more reactive way.
Mike Keoghan: On whether we talk to people, the LMS will retain a face-to-face element, so we will continue to do that. One of the things that the ONS has been doing over the pandemic period has been to try to be much more agile in relation to emerging problems and not be constrained by the existing set of surveys. One of the things I would highlight here which probably is useful for your inquiry is the reactive surveys such as the over-50s. We also did a similar one on Ukraine and the experience of people who have come from Ukraine. But we can do essentially reactive, agile surveys on the most important issue of the day.
We also run two opinion surveys, one for individuals and households and another looking at business—a business condition survey, which I think is similar to the Bank’s decision-makers panel survey, which asks businesses about how they are experiencing the economy, whether they can recruit, what the big cost pressures are and whether they can push prices on. The structure is very flexible, so new and emerging problems, and issues or opportunities, can be placed on to those surveys within a four to six-week lead time. So we are in a position to act faster than perhaps traditionally we have been.
Q9 Lord Stern of Brentford: I have a slightly technical question about sample selection bias. If you have switched from doorstep to telephone, then the issues of sample selection bias would change. On the probability of hanging up the phone, if double glazing people or computer people in your area come for a survey, they are all triggers for people putting the phone down. If that probability of hanging up is correlated with employment status in a different way from a reaction on the doorstep, your whole sample selection problems will have changed.
David Freeman: You are absolutely right, and we did see that at the beginning of the pandemic: first, a shift in response rate, so the response rate was lower moving to just telephone, even though we put mitigation in place—an online portal for people to give their telephone number.
We probably saw two things happen to the sample response. First, that we received a smaller proportion of rented properties responding. We also saw a step change in the number of non-UK nationals responding to the survey. To mitigate those, we reworked the weighting of the survey. So when we took sample responses and we worked up to the UK level figures, we brought housing tenure into that weighting process, and we brought in nationality linked to the new data source we received from the real-time tax data, split by nationality. While we got that shift in sample response and a bit more bias, we have managed to remove as much bias as we can through recalibrating our weights.
Mike Keoghan: It is also worth saying that, as we move from it, we are also doing a dual run of the LFS and the new survey to try to ensure that we can identify variances that may stem from the change in method.
David Freeman: We talked earlier about moving to an online first version, as Mike says. It is online first, but if we cannot get an online response we then approach via telephone and then finally go for face to face. We are running that new survey currently in parallel with the headline Labour Force Survey and using that to say, “Is the new survey giving us better quality data? Is it biased in a different way? What do we have to do to the weighting to ensure that it is at least as good as the current LFS, if not better?” We are looking potentially to make the switch next year between the current and the new survey, but that is dependent on what comes out of our quality assurance processes.
Q10 Lord Layard: The thing that is stimulating this inquiry is the shortages. The extraordinary thing is that the level of vacancies is so high relative to the level of unemployment. Unemployment is the same as it was, essentially, pre-pandemic, when nobody was talking about shortages and now the vacancies are so much higher. You have hinted at one possible explanation, which is a higher level of turnover, but we are talking about the reliability of data. Could you say something about the vacancy data and how credible that is?
David Freeman: The vacancy data that we use as our headline is based on the survey of businesses, so we take a sample of businesses each month and we write to them and ask how many vacancies they are actively recruiting for. It does not necessarily mean how many vacancies they have and how many they are advertising or interviewing for. There will be some sample variability around that, but during the pandemic we have had access to some online vacancy data, through a company called Adzuna, and we have been doing some work analysing that data. That is showing similar trends to what our vacancy survey is showing.
Where they do diverge, we have had to do a bit more work, but the underlying trend is they are both showing very big increases from about the beginning of 2021 onwards. We are confident that the business survey is giving us reliable data but there is a bit of a conundrum here about why we have such high levels of vacancies with unemployment being the same. We have talked about people moving out of the workforce into inactivity, and that is one of the drivers I mentioned before regarding changes in migration, as well and the increase in long-term sick. But also, as time goes on and the population increases, employment will generally increase, so you have a couple of years of population increases and therefore demand for jobs will go up, and we have not had that increase in the available labour force to go alongside it.
Q11 Lord King of Lothbury: I will start with a technical footnote to this discussion: do you produce estimated standard errors for any of the numbers in this area?
David Freeman: We produce them for everything that comes from a sample, so in the monthly publication we put out there are sample errors for all the headline labour market statistics: employment, unemployment, inactivity. There are also standard errors for the ones that come from the business surveys, such as earnings growth.
Lord King of Lothbury: Did these errors go up a bit during the pandemic, when you had to rely on telephone?
David Freeman: They did go up slightly, again because the response was down, so that does naturally increase the sampling variability. One of the things we did to mitigate for the Labour Force Survey was to increase the sample size. So we had a similar number of responses, even though the response rate was lower, but that does have an impact on your sampling variability figures.
Q12 Lord King of Lothbury: Thank you. That was a footnote. I want to ask three questions about long-term sickness. Could you summarise again your judgment as to the contribution of long-term sickness to the scale of the rise in inactivity? Secondly, could you remind us of your definition of long-term sickness and say whether you feel that is truly objective, or whether people may tell you they are long-term sick when they are too embarrassed, for example, to say that actually they decided not to work again and do not want to?
Thirdly, do you have any reliable information on the contributions to the number of long-term sick from things such as long Covid—we do not have a good definition of long Covid—NHS waiting lists and so on? Those are the three questions I would like to put to you, and perhaps you could divide them between you.
David Freeman: The definition comes back to the fact that the Labour Force Survey is self-classified. When we get to the relevant part of the questionnaire and we have established that someone is economically inactive, we will say, “Which of these reasons is it?” and if they say it is long-term sick, we will not necessarily ask them to demonstrate. We do ask a follow-up question, “What is the nature of the long-term illness?”, which is not part of our main publication but is there for additional analysis—so if it is a mental illness, physically based, musculoskeletal and so on. While we do not publish that regularly, it is there to provide some additional analysis if anyone wants it.
Mike Keoghan: Regarding the other questions, what I do not have to hand is the proportion of the increment in terms of long-term sick. All we have to hand is the proportion of people who are inactive who claim that they are long-term sick, and that is 27%.
In terms of the other questions, we have two pieces of ongoing work. The issue around if they are on the NHS waiting list is a question in the second wave of the over-50s survey, so we should have some insight from that. There is also some work going on that tries to link the work that we do in the LFS with the Covid infection survey, in order to try to get a sense from those two surveys, if you put them together, of whether they can back out any insight around prevalence. That is ongoing work, and we expect to publish it.
Lord King of Lothbury: There must be some difficulty in using phrases such as “long Covid” or even “long-term sick”, phrases that we use to describe general concepts, and matching that to responses. Of course, the changes from pre-pandemic to now may also reflect changes in the way people have thought about phrases like long-term sick or have started to think about issues, and they come up with words to describe it, whereas they would not have done it three years ago. Do you have a feel for how reliable these numbers are likely to be? I can see that if you ask someone if they are on an NHS waiting list and it is a yes or no answer, it is reasonable to take that as a given. For the other things, what do you and those in the field judge, having carried out these surveys? How reliable do you think they are?
Mike Keoghan: Is it worth talking about the guidance that we give to the field force to enable them to answer these questions?
David Freeman: We can probably send the guidance that we give to each of the blocks of questions on the labour force. There is a preamble where we talk about what the next question is going to be. I will have to come back to the committee, as I would be speaking off the top of my head and I cannot remember. I think we do give an idea about what long-term sick is. I think it is how long they expect the illness to last. I think it might be six months or 12 months as a cut-off point, but we will have to come back on that.
Mike Keoghan: We can provide the guidance that we provide to the interviewers.
Lord King of Lothbury: Will your surveys over time and over the next two or three years be able to map the persistence of any particular cause of withdrawal from the labour market?
David Freeman: At the aggregate level we will. We will not necessarily be tracking the same individuals over time. The individuals will be on the Labour Force Survey for 15 months, but it is something that we regularly publish and, as Mike said, one of the things we look at are the reasons for inactivity and how they change over time.
Q13 Lord Fox: John Burn-Murdoch in the Financial Times directly equated the UK’s position in terms of people being long-term sick with the NHS waiting lists and a higher level of chronic disease. Do you ask people if they are on an NHS waiting list? Is there a way of correlating those that are with the statistic? Have you seen movement of that statistic before and after Covid and is it a statistically significant thing, or is it happening in parallel but not necessarily causing what is going on?
David Freeman: On the Labour Force Survey we do not ask directly if someone is on an NHS waiting list. However, in the current incarnation of the over-50s survey that has just closed, which Mike talked about, we did include that as one of the questions, so we will get a bit more information later this month on the proportion of over-50s who are on the NHS waiting list and how that correlates with their labour market status.
Lord Fox: I do not know if you saw the Burn-Murdoch article but perhaps we should let you have it. We would be interested to know if you think that he is on to something, or whether he is disappearing down a rabbit hole. It was quite strident information, so perhaps we could make that available to you.
The Chair: We should certainly try to make that available. I am going to bring in Lord Rooker and then Lord Livingston.
Q14 Lord Rooker: My question is very similar and on the same point. According to our background notes—leaving aside the pandemic; I am talking about before Covid—between 2009 and 2020 NHS waiting lists virtually doubled: 2.3 million to 4.5 million. The question Lord Fox and the journalist asked concerned the effect of people being on the waiting list and whether getting more ill affects their willingness to work or their safety to work. How do we measure that, or do some of them just cop out and say, “I will take retirement; I will opt out completely”? How can we measure the correlation in terms of such a big increase? This is pre-Covid. It has shot up enormously since then, but the statistics are there for the decade before Covid of a doubling of the waiting list. That must have had an effect on people’s capacity and willingness to work, and your statistics are not going to pick that up, are they?
David Freeman: Again, we will not be linking and asking about whether someone is on the NHS waiting list. The number of people who are inactive because they are long-term sick goes back to the mid-1990s, so we have our labour market data for that period, but we have not done any analysis on trying to correlate against NHS waiting list data. We can let the committee have a longer time period of the inactivity numbers, but I do not think there was any significant change. Long-term sick is one of the more stable elements of that, but we can let you have a longer-term time series.
The Chair: Can I take a step back? We have been going for 50 minutes now and, as Lord Fox and Lord Rooker have said, there is obviously a dispute and you have raised the issue of long-term sickness. There is the issue of whether it is the over-50s who have retired. Here we are and we do not know the answers to these questions, and it strikes me as odd. The Governor of the Bank of England said back in May, “I have to be honest. We do not know very much really about what is behind all this”, “this” being the fall in the workforce.
It strikes me as very odd that in this age, when data is pouring out of our ears, we somehow do not know the answer to this question. It is not as if we have had several months to think about it. Is this the circumstance of Covid? Mike, you mentioned the agility of the ONS to change surveys or to collect data. Why are we in this situation? We have all the ONS might, we have all the Bank of England, and we cannot seem to answer this question in a straight way.
Mike Keoghan: This was an issue that was spotted in the LFS towards the end of last year. In the early part of this year, we put out the over-50s survey, and surveys can be agile and fast but they do take time to collect. We received that one back and that provided some useful insight into what was happening, but that then created a second round of questions, which are the issues you are raising today.
I fully expect that if we were having this conversation a month hence, we would be able to provide you with a lot more insight as to, for example, how many people are on waiting lists and how many people have now decided they want to go back to the labour force. I think this is one of those ones where the policy-making and policy interest cycle has moved faster than the ability of the survey machine to do it, but we are moving quickly. We are pretty confident that in the next four weeks we will have answers to the questions that the Government have been asking us on this, and which also reflect where the committee is.
Q15 Lord Livingston of Parkhead: The UK, regrettably, is not unique in experiencing Covid, so what is of interest is the work you have done. How does the UK’s experience compare and contrast with other rich industrialised countries?
Mike Keoghan: We have seen most countries increase their employment rates for the three months to March 2022 compared to the same quarter the previous year. The UK had a relatively low increase, with USA, Canada and Australia all seeing larger increases in their employment rates. Ireland saw the biggest gain in employment rate with an increase of 7%. Most countries have seen a drop in inactivity rates for the three months to March compared to the same quarter the previous year; there was no change to the rate in the UK for the same period. Therefore, the UK was one of the few countries that did not see a decrease in its inactivity rates and Ireland saw the biggest fall, with a reduction in its inactivity rates of 5.9 percentage points.
Lord Livingston of Parkhead: You produce the international comparison of the labour market, which I assume you are referring to. It does highlight this point that the UK’s inactivity levels are getting worse, whereas in countries such as Germany you see a huge bounce-back. In fact, the inactivity rate is much lower than pre-Covid. These numbers for the UK only take you to December 2021, and arguably we are still in a Covid world. We are now in September 2022. Do you have any update as to what has happened? Is life changing, and is what you are seeing affecting countries differently? Certain US states exited Covid restrictions, for example. Is there any update on those inactivity numbers? The crucial point seems to be that the UK’s experience is vastly different from almost everywhere else.
Mike Keoghan: I think the numbers I gave you were up to March 2022. Are they the latest we have?
David Freeman: I think they are the latest, comparable across all countries. We have a few UK figures that go up to Q2 at the moment, but for the comparable base I think we only have up to Q1.
Lord Livingston of Parkhead: Any update on the survey you did would be helpful, because understanding whether this is a temporary or more fundamental change in the UK employment market is critical to our inquiry.
Baroness Kramer: Which sectors are the most and least affected? I have some follow-up questions on that, but can you give us your view?
David Freeman: The hospitality industry stands out as the most generally affected, particularly by Covid. That is where you saw the biggest drop in employment at the beginning of the Covid pandemic and currently, although it is not the highest level of vacancies, the number of vacancies in that industry has more than doubled compared to the pre-pandemic level. There would be about 180,000 vacancies.
Other areas that have been particularly hit are in the health and social care sector. Again, that has the highest number of vacancies. That is up over 68% on where it was pre-Covid, wholesale and retail. Again, there is lots of demand there. In terms of least affected sectors, I know there was a question—
Q16 Baroness Kramer: Can I ask about farming and construction? I think they were mentioned in some of the data that we looked at.
Mike Keoghan: In construction, 75% of businesses reported shortages of workers.
David Freeman: Farming is one area on which we do not have as complete information as others, because in a lot of other areas we write to businesses and it is hard to write to individual farms. But I believe we do have some vacancies data.
Baroness Kramer: Do not worry too much. I am not looking for a specific answer on farming for the sake of farming. You were going to mention some sectors that seem not to be impacted.
David Freeman: The public sector generally has maintained its employment levels or increased over the pandemic period. That is partly an increase in NHS employment but also partly from the Civil Service as a response to the pandemic. Those sectors were relatively unaffected, but most sectors have seen some sort of impact and mostly it has been negative. Again, construction, accommodation, food, creative arts and recreation were also affected by the pandemic. We have also seen more demand for professional, scientific and technical jobs—things like barristers, solicitors, accountants, veterinary surgeons, all those sorts of very specialist jobs.
Baroness Kramer: For the sectors that you have highlighted as the worst hit, do you analyse the quality of the jobs on offer? It strikes me that social care is very badly paid, and it is a lot of hard work. Hospitality, unreliable; shift work, low paid; farming, seasonal, low paid; construction, some of it well paid but otherwise low paid and often project by project. Do you run any test of job quality?
David Freeman: We collect some information on job quality. There was a project run by the Carnegie Trust a couple of years ago, just before Covid, on measuring quality of employment. That recommended 18 variables and we currently have data for just over half of those. Some of those are quantitative, like level of pay and number of hours. We have started asking a few more qualitative questions around the amount of control you have over your job. In moving to the new online survey, this is something we will look to do a bit more comprehensively and try to cover all those 18 variables, measuring quality of employment.
Mike picked up some of the additional analysis we have done. We have not done any since Covid but, just prior to Covid, we did some analysis looking at where we had data on quality of employment, looking at the percentage of jobs that are good quality based on a set of criteria, and compared that across the country and across sectors. We could send you a copy of that if you are interested.
Baroness Kramer: Have you seen a correlation between inactivity of people who have traditionally worked in those sectors versus people who have worked in other sectors, or is that not something you would be aware of?
David Freeman: That is not something we have done any analysis on. We do collect information on where someone’s previous employment was, if they are unemployed or inactive, but I do not have any specific analysis on that now.
Mike Keoghan: That kind of analysis would be undertaken typically by the sponsoring department of those industries, so BEIS or the HSE or the Department for Education.
Baroness Kramer: Have you looked at outsourcing? I do not mean conventional outsourcing. For example, I deal with a lot of tech and fintech companies and, being unable to find the skills that they wanted domestically, they have basically taken chunks of their activity and outsourced them to a firm—Ukraine used to be a favourite place; Romania is a favourite; Portugal is a favourite—or indeed have recruited individuals who remain and do remote working from those countries, something that did not exist prior to Covid. I wonder if that is having an impact on where you are seeing vacancy rates, because there are obviously implications for the UK economy if there is a changing pattern to sourcing from locations outside the UK.
Mike Keoghan: That is not something that we would pick up in the labour market surveys. I am not familiar with whether we do pick it up but if we did, it would be in one of the business surveys, in particular the business conditions survey, because it would be linked to questions around how they are dealing with an inability to hire people and deal with costs. At this point I am not familiar with whether we are asking that question, but if we were to ask it, that is where we would find it.
Q17 Lord Layard: Could you tell us about the impact on the public sector as compared with the private sector in the main dimensions?
David Freeman: Employment in the public sector increased over the pandemic period because the nature of the work that was happening for the NHS increased in size, and more civil servants were employed to respond to the pandemic; it was a testing phase that led to higher employment levels.
Pre-pandemic, the public sector accounted for 16.5% of employment. It went up to 17.5%. The level in the public sector increased by an estimated 18,003 over that two-year period. In the private sector, employment fell by 581,000 over the same period.
Pay levels in the private sector fell during the Covid period, mainly because of the impact of the furlough scheme, when a number of people were getting 80% of their usual wage. In the public sector, pay was almost unaffected over the same period because those people were still working their usual hours, albeit in a different format.
Lord Layard: I think you said that the vacancy situation is more serious in the private sector. Is that right?
David Freeman: We do not split the data by public and private sector. Health is the area with the most vacancies at the moment, and there was also strong growth pre-pandemic. The figures will include some public sector vacancies in the NHS and also in social care in local authorities.
The only specifically public sector data we have is in the public administration industry and will include mostly civil servants, local government administration and so on. I can send you some figures on that if you are interested.
Mike Keoghan: It is probably worth thinking about the current pay situation because that is one of the ways that markets adjust. The most recent data shows that total pay in the economy is up by 5.1%, regular pay is up by 4.7%, real total pay is down by 2.5%—that is the inflation effect—and for regular pay without bonuses the figure is minus 3%. The sectoral split, in nominal terms, is 5.9% for private and 1.8% for public. There are some interesting splits in some of the sectors we have been talking about, such as wholesale, finance, fintech. Wholesale was up by 7.7% and finance, in nominal terms, by 6.3%, according to the most recent Labour Force Survey.
Lord Monks: You have given the key statistics. I am interested in the fact that now, post the pandemic—if I may be so bold as to forecast that, which may be extremely foolish—is there not going to be a strange conjunction between the shortage of labour, the high demand for labour and pay growth, which according to the figures you have just given is well below the level of inflation?
From a trade union point of view, which is where I originally hailed from, these figures are quite—dynamite might be putting it too strongly—capable of stirring up a lot of groups to think that they might need an increase to keep up with inflation, and they are in a market position to do so because there are shortages and employers are looking for labour. Do you have any insights into that situation?
Mike Keoghan: The ONS does not forecast. We do not project. We can tell you as best we can what the current situation is. What we most recently see is relatively high levels of unemployment, low levels of employment, negative pay growth for most workers and some compositional shifts, and at the same time, large numbers of vacancies. There are people in the room better qualified than us who have a wider view of how this plays out. All we are in a position to say is that this is what the current labour market position looks like.
Q18 Lord Monks: Turning to particular sectors, which we did get into with the conversation on public/private, from your data are there not some fairly clear differences between, say, financial services, where wage and bonus earnings have been pretty strong recently, and other groups that are looking round for comparators and seeing that there are some pretty big increases in some sectors? Are you picking that up?
David Freeman: We normally see very high growth in the financial sector, particularly around bonuses, and we have seen that in recent months. We have seen bigger bonuses this year than we have seen in the 12 months previously. We are also seeing it in other sectors, particularly in construction but even in wholesale, retailing and hospitality. We have seen growth, including bonuses, of 7.7% in the most recent period.
Where pay growth is slightly lower among private sector industries is in manufacturing, which is currently at 4.1%, but that is still above where we are in the public sector, which is at about 1.8%. While there is a great disparity, it is not just finance and business services that have seen that big growth.
Lord Layard: Could we say that vacancy rates across sectors are related to wage growth?
David Freeman: It does appear to be. The big demand, the big increase, has been in wholesale, retailing, and accommodation and hospitality—they are the top sectors for numbers of vacancies. The public sector and administration industries have a relatively low level of vacancies; although the numbers are increasing, they are fairly low. It does appear, therefore, that there is some pay growth where there is demand. However, a couple of months ago we saw a big gap between including bonuses and excluding bonuses. Businesses may have been using bonuses to retain or attract staff, but in a way that does not increase baseline pay bills; so, people have pay increases in the baseline pay, and bonuses are being used as a more flexible way of rewarding workers or enticing workers to come into particular businesses.
Lord Skidelsky: I am glad that wages have entered this discussion because they seem to be pretty crucial. We have heard about the increase in inactivity rates among older workers, the 50 to 64s. Could that not be partly explained by the fact that they cost more than younger workers and, therefore, if you are thinking of hiring, you will hire someone who costs less than an older worker? Are wages not very important, if older workers will get towards the top of their pay grade and businesses can get younger people more cheaply? Have you considered that? This is a conceptual point; I do not know whether the statistics pick up any of this, and it is hard to know quite how they could.
David Freeman: We do some analysis of earnings by age, and you are right. If people stay in the same industry occupation, their earnings do tend to increase over time.
Our conundrum is, if someone has withdrawn from the labour market to inactivity and then decided to re-enter, would they necessarily be able to go back to where they were previously? It would perhaps depend on their level of experience, but it may be that they would have to come back in at an entry-level pay grade, which might prevent someone going back into the labour market unless it is worth their while going back into the same industry at the bottom again.
Lord Skidelsky: They have either been made redundant or decided not to work in the first place.
David Freeman: That is what came out of our over-50s survey. The majority of people in that age cohort who have withdrawn from the labour market do not intend to go back. Only 40% of them would consider going back, but it would depend on what the job would be giving them. It would have to be a job that they would enjoy and could be flexible—either working from home or flexi-time—so perhaps when we ran the survey, they were being a bit more choosy about the jobs they would do. It will be interesting to see what people are saying this time around, because of the changes in circumstances, the increase in energy costs and other things happening in the economy, compared with six months ago.
Q19 Baroness Noakes: I will shift to self-employed people. In the 20 years before the pandemic, rates of self-employment grew pretty healthily, steadily. They took a big dive during the pandemic and have now flat-lined and have not gone up again. I think I heard you say that there were classification issues—people calling themselves different things—possibly related to furlough, to changes in IR35 legislation and in some cases to people literally moving out of working.
Can you say any more about what you know to date? If not, can you describe what additional information you might be able to get on this in future? This committee has had a strong interest in the impact on the economy generally of IR35 legislation and what that is doing to employment patterns.
Mike Keoghan: Like you, we saw a big change and our interest was in how much of it was real and how much of it is due to reclassification. In any one month, you will see people doing the same role in the same job move their employment status—a contractor who is taken on, or the worker who gets taken on and is deemed to be a temp worker. That is just a feature of our labour market, but during the pandemic we saw some big swings in one direction. Clearly there were people either actively moving into employment or realising that they were employees and, therefore, should be on payroll. That effect has wound down. That behaviour looks normal, so we think there has definitely been a reclassification element to that reduction.
There is also a question about the consequences of certain policy decisions. We are not doing any work on that, but we understand that parts of HMRC are looking at that one. If that is something that the committee is interested in, it would probably be for HMRC to tell you where it is on that.
Baroness Noakes: We have been in dialogue with HMRC already and are looking for any additional information to help us in our long quest to understand what is happening. Are we likely to get much more analysis of what is happening? Is this an area you are still actively looking at, or is it a question of, we have dropped down and it is not of interest any more?
Mike Keoghan: Our interest was around the extent to which it was real or whether it was an artefact of the data collection. I think we have satisfied ourselves that we understand that part of it. The extent to which it is a consequence of policy is where the interest of the ONS naturally stops.
The Chair: Therefore, we need to go to HRMC. Fine.
Lord Fox: It was not clear to me from your answer what percentage is reclassification and what percentage is people changing from being self-employed to being employed. Did I miss that?
David Freeman: Reclassification applies to people who, according to their answers, do not change their job; they change the classification, and everything else is the same apart from whether they say they are self-employed or employed.
There are always people who change, and pre-pandemic that rate was fairly flat. We saw probably an extra 200,000 people make that change over the pandemic period. We are back to roughly where we were pre-pandemic.
Lord Fox: The change they made was that they signed up to have a full-time job rather than being self-employed. At the beginning, you said there was this change and you have explained that there are two sorts of change. What I am not aware of is the quantum of the number of people who have actually moved across the floor from being self-employed to actually having a new, different job that is employment. That is what I am seeking to know.
Mike Keoghan: We will send you that information. We do have it, but not here.
Q20 Lord Rooker: I have a question on the geography. Are there any significant differences in labour market trends across the regions and nations of the UK? I will ask that question minus one of the words: are there any significant differences in labour markets across the regions and nations of the UK?
David Freeman: We do this as part of our regular publication. Certainly, for the latest quarter, most regions are showing similar trends. We are seeing increases in employment across every area of the UK apart from Wales. Wales is the only part of the UK to show a decrease in employment and an increase in inactivity.
We are seeing different patterns across the UK. We are also seeing differences in unemployment. The north-east had the highest unemployment rate at 5.1%. The lowest was in the East Midlands, 2.4%, a record low rate for the region. Scotland has an unemployment rate of 3.2%, which is a joint record low.
Lord Rooker: I have looked at your latest figures, in table 3, the latest headline estimate. The inactivity rate in the north-east of England for 16 to 64s is 24.9% and in the east, it is only 18.9%. That is one hell of a difference. I am not a statistician and I know it is only 6 percentage points, but it is actually a 32% difference. Looking at the employment rates of the two regions you are highlighting, in the east of England it is 78.4% and in the north-east, it is 71.4%. That is a 10% difference, whichever way you look at it.
The question arises, therefore: what is the reason for this astronomical inactivity rate—I think it is nearly 25%—in the north-east? How much of that is related to redundancy—the old big jobs going in the heavy industries—and the theme of a lot of our questions, long-term sickness? There is a big gap between the highest inactivity rate in the nation and the lowest, in the east of England, of only 18.9%. I say “only” but 18.9% seems high, and there is a big gap between the two.
Mike Keoghan: A lot of it is historic. Historically, the north-east has generally had relatively low employment rates compared with the rest of the UK, and high rates of inactivity.
It could go back to the days of heavy industry such as coal mining having an impact, particularly on people’s health. We generally also see high employment and low inactivity in London, the south-east, the east of England. Some of this is built in to the structure of the UK and the local economies.
However, the north-east has seen improvements in its employment rate, and a decrease in its unemployment rate. If you go back three, four or five years, unemployment rates were in the double digits in the north-east. The rate is now about half of where it was.
Therefore, we are seeing shifts but the relative rates of unemployment and employment in parts of the UK do seem to be almost fixed over time. While they might improve and rates of inactivity might go down, they tend to go down together so you still end up with that discrepancy between the regions.
Mike Keoghan: We know that there is more to do on regional statistics generally. We have a quite substantial investment programme going on to improve the quality, quantity and representativeness of our regional statistics. As an office, we are also planning to put more statisticians into the English regions, not just to provide data to the minor authorities but also to understand what the problems are on the ground, so that our data collection reflects economic reality and what people are finding. This is a work in progress but there is some very rapid work going on to try to understand what is happening out there.
Q21 Lord Griffiths of Fforestfach: I want to ask you about Wales. Not only is there a decrease in the employment rate there, whereas there is an increase in the employment rate everywhere else, but the smallest decrease in unemployment was in Wales, the largest increase in inactivity was in Wales and the smallest increase in the number of payrolled employees was in the East Midlands and Wales.
I must declare an interest—I am Welsh—but that is a pretty poor record for Wales. Is the explanation perhaps what you hinted at earlier in respect of the north-east—the influence of de-industrialisation—or attempts to increase use of the language and so on, which has held migration back? I wonder if any work has been done on that.
The Chair: Can I bring in Lord Fox as well? We can hold Lord Griffiths’ question meanwhile.
Lord Fox: I am not Welsh. I am interested in the incidence of chronic illness because there are regional variations. Is there any mapping of the sort of variation that Lord Rooker was talking about, on the instance of chronic illness?
The Chair: Two questions: Wales and chronic illness.
Mike Keoghan: For Wales, presumably the same sorts of historic patterns are broadly the same sort of story. That is as far as the ONS could probably say, given our current knowledge. Further information is probably a matter for the Wales Office or the Welsh Assembly Government. If we wanted to map chronic illness, what question on the LFS would that track to, David?
David Freeman: Potentially, the inactivity due to long-term sickness. There are some health questions but I do not know if we have them on the annual version.
Lord Fox: Some towns are famously worse for chronic illness and would, therefore, have a higher level of inactivity, so it would seem to me to be perfectly possible that there is a correlation between the areas that have the lowest levels of activity with those that have the highest levels of chronic illness. Is it possible to get that kind of correlation?
Mike Keoghan: As part of the ONS, we do have quite a substantial health practice so we can take that question back to the office. We will add that to the list.
The Chair: That would be very helpful. Thank you.
Q22 Lord Skidelsky: We have heard evidence—although it is inconclusive—about why older people are leaving the workforce. What factors might encourage them to rejoin or remain active for longer? One can think of a number, and they have been mentioned: improved job quality, levels of pay, more capacity in the NHS, less remote working, retraining and so on. First, do your numbers enable you to rank these causes at all? Secondly, what should we as an inquiry, as a committee, be concentrating on? What factors might encourage people to rejoin or not retire so early?
Mike Keoghan: Our survey does ask what would encourage a person back. A small proportion said they would go back. There were some issues around flexibility, job enjoyment and ability to fit other things in—for example, caring responsibilities—around their work. We do have some quantification but the survey to come will put a lot of emphasis on that. As for where to focus, what do you think, David?
David Freeman: Is your question what you should be focusing on specifically around the older workforce?
Lord Skidelsky: I wonder, first, whether you have any way of ranking any of this. There are lots, dozens, of factors. From the surveys, do you have any sense of the relative importance of the factors?
David Freeman: Nothing from the current surveys. I think we recently asked the over-50s what would encourage them back to work, and people just ticked some factors. We have not done any ranking and I am not sure we could do that with the current survey.
Lord Skidelsky: Because you do not ask those questions?
David Freeman: No, we do not ask people to rank them in importance; we ask: what would encourage you to come back?
Lord Skidelsky: We are left with a lot of factors.
David Freeman: Yes. You could potentially rank them according to how many people say such and such is a factor. If over half said it was a job they enjoyed, slightly fewer said they would be encouraged by the money and under half said it would be a job that suited their experience, would that change six months down the line from the previous survey? Are there changes over time in what would encourage people to go back, depending on the economic circumstances?
Lord Skidelsky: What you have just said is useful. I wonder whether we could get some figures that reflect what you have just said.
Mike Keoghan: I would want to provide them following the next survey. We could provide the current ones, but the next ones would probably be better.
The Chair: When will the next survey be done?
Mike Keoghan: It has been completed. It is now going through quality assurance. We expect it by the end of the month.
The Chair: This month?
Mike Keoghan: Yes.
The Chair: That would be very good. It would be great to see those findings. That would be very helpful.
Q23 Lord Stern of Brentford: You have been very helpful on long-term sickness and early retirement, and you are promising us some goodies on those two subjects—a justified promise, it sounds—in a month or so.
It is less easy to be specific about migration, and the Labour Force Survey must be weaker on migration-related areas. It does depend on people being at home or near their phones and so on. If you are long-term sick or have taken early retirement, the probability of being at home is rather higher than for others. You are giving us that but, clearly, there are problems with understanding the role of migration. From ad hoc observation, you hear lots of people in hospitality, farming and social care—as you have indicated directly—referring to the lack of immigrants relative to the position before Brexit came in strongly. Can you tell us how you see the role of the immigration story, including sectorally, and also how you intend to go deeper into the role of immigration?
Mike Keoghan: Is it worth saying what we have done on immigration and the immigration statistics, David?
David Freeman: We are going through a major transformation process on our migration statistics. Prior to the pandemic, the biggest driver was the international passenger survey, where we interviewed people at ports and airports on their intentions—people moving in and out of the country. We are now looking to use a much wider range of data, including things such as tax data and visa applications. We are in transition, moving to that new data source. There is a lot of work being done to improve our understanding of people coming to and leaving the country. That is out in the public domain and we are working through it now, supplemented by the work that my team does using the Labour Force Survey on the number of people in the country who are of particular nationalities, what jobs they are doing and whether they are active or inactive in the labour market.
Mike Keoghan: We also have data from payroll. That provides us with some information. The current situation is that we are seeing falls in EU employment in agriculture, forestry, fishing and the arts. The area in which we do have some information on this is from our business impact survey, which does say that about 23% of businesses are citing the inability to access migrant workers as a barrier to their growth.
The Chair: How certain are you? I note that you said, “ONS has produced various estimates of EU migration, but they are subject to large uncertainty” and this, I dare say, is because of the pandemic. How certain are we going to be able to become? I hear what you are saying about the replacement of the APS, but we are thinking about what has happened over the last three years. How much certainty can we get from that data to answer the basic question of how many EU workers have gone home? In our excellent brief it says, “between June 2019 and June 2021, payroll employments held by EU nationals fell by 6%”. How certain are you about that figure?
David Freeman: The RTI data is probably one of the more robust elements of the data field. It is a count of people on payroll so there are no sampling issues there. It is matched to the national insurance numbers—so, somebody’s nationality when they got their national insurance number, which may have changed in the interim, but that will be a fairly robust measure. It is now feeding into our Labour Force Survey weighting but based on survey responses, it appears fairly consistent. We have seen a drop in the number of people from the EU who are employed in the UK. We have also seen an increase in the number of people outside the EU employed in the UK but that does match what we see in RTI data. It also matches visa-application figures that we have received from the Home Office. We are getting a better understanding of the role of immigration in the UK labour market, but we do have a way to go.
Baroness Kramer: How granular is your information on immigration? For example, do you track the age groups? Do you in any way take note of skill levels, or is it just whether somebody is a non-UK national, whether or not they are from the EU or from beyond?
David Freeman: On the Labour Force Survey we collect the same data irrespective of whether the person is from the UK or the EU, so we do have some information on levels of qualification, and on age, which we also get from the tax data. The limiting factor is that because it is a sample survey, if we try to break it down too far, we do not have enough people in there to give accurate estimates. While it would be great if we could get that level of granularity, we simply do not have the data to be able to do it.
Baroness Kramer: You would not be able to say that a significant proportion of people who had returned to the EU were low-skilled or high-skilled, or whether it was older folk or younger? You would not be able to give us a sense of the profile of those who had left or the profile of those who have arrived.
David Freeman: We know certain things about the profile of people from the EU when they are in the UK. People from the EU tend to be concentrated in areas such as the construction and hospitality industries. Non-EU citizens tend to be more in the professional, scientific and technical areas. So, you can make the assumption that the ones who have left fit that profile, but it would be hard for us to do the analysis based on the data we have.
Q24 Lord Rooker: I was not going to raise this issue until I listened to the last few questions, but just over 20 years ago when I was at the Home Office, it was pointed out to me quite forcefully that the UK was the easiest country in Europe in which to work illegally, mainly because we have no ID system. Nothing has changed since then. Do you pick up anything relating to illegal working?
David Freeman: Not knowingly. In the Labour Force Survey, we do not ask someone if they are working illegally or not; they may be. We do make some adjustments for it in national accounts—one area is household income, where we do make an adjustment for illegal working, the black economy—but we do not have a direct measure of that.
Lord Rooker: You mentioned national insurance numbers. You cannot use national insurance numbers. I think the figure must have reached 100 million now because they have to be maintained on the system because of dependencies, so everyone who is now deceased is still on the system because pension benefits are paid. National insurance numbers are not a factor that is helpful in measuring any of the things that might go wrong. You need one—the employer will want a national insurance number—but how you get one, of course, is another matter. It is not a very fair question to ask, by the way.
The Chair: Can I ask a fairer question? Is it the case, just in simple terms, that the declining number of EU workers has been replaced by non-EU workers in the workforce? Is that basically what we have seen over the last two to three years, from the figures?
Baroness Kramer: Could I follow up on that? I think what you said was they do not necessarily replace the same job.
The Chair: Exactly, so can you address that point and if not now, maybe in writing—people who might have left to go back to the EU and, if they have joined the workforce and are non-EU workers, in what sectors? I know that there are some good graphs here, but I am interested in some granularity.
Mike Keoghan: We can provide some granularity, but at a high level. The number is lower, but it is broadly—ballpark—net zero on EU and 251,000 on non-EU, but they are in different places.
The Chair: Very good. Unless anyone else has any other questions, thank you very much. There is a lot, I think, potentially to follow up on, particularly the over-50s and maybe overlapping with that, those who are long-term sick. I will ask our team to talk to your team about how we can follow up on some specifics. I would also be very grateful if you could send us as soon as possible a short summary of what you are intending to publish over the next four to six weeks, so that we have some clarity on that. That would be very helpful indeed. We may well come back to you with some further written questions as we take further evidence.
Thank you very much indeed.
