20
Examination of witnesses
John Burn-Murdoch and David Finch.
Q87 The Chair: Thank you for joining us at this meeting of the Economic Affairs Committee and our inquiry into where all the workers have gone. I am delighted to be joined by you both this afternoon. David Finch, would you like to introduce yourself first, and then John Burn-Murdoch?
David Finch: I am the Assistant Director for the Healthy Lives team at The Health Foundation.
John Burn-Murdoch: I am chief data reporter at the Financial Times.
Q88 The Chair: May I start with a general opening question? Please do not go into too much detail on long-term sickness because we are very keen to get into that in a moment. Will you both broadly summarise your views on why the UK is seeing a rise in inactivity among our workforce? We are hearing, at a high level, very different views from very different organisations and institutions.
David Finch: In the big picture, we are seeing a longer-term trend become more apparent—the increase in long-term sickness and inactive people. That has been particularly noticeable because of how well things such as unemployment have fared since the pandemic.
There are different views, which I think we will cover a bit later, on the extent to which the pandemic is driving particular reasons for people moving into inactivity. It goes back to the situation before the pandemic, if we are thinking about the labour supply, where we had high employment. We had decades or so of being very successful at getting parents into work but, if we wanted to increase labour supply pre-pandemic, it would have required a focus on getting people with long-term sickness back into the labour market. At the same time, you had some reductions in inactivity because of things such as the state pension age rising through the last decade.
Ignoring precisely what has happened in the last year or two, this is a bit of a re-emergence of a longer-term trend where we were seeing a stubborn, inactive population of people with poor health.
The Chair: We keep coming back to poor health.
John Burn-Murdoch: The overarching point that I will probably make until we are all tired of it over the next hour is that this is multifaceted and there is a danger of saying that it is one thing to the exclusion of others.
My high-level view is that ill health is playing a role here, as David mentioned. It is partly playing a role among people who were already inactive, but there is at least some evidence that it is accounting for some of the potential flow of people from activity into inactivity. Again, there is a danger of saying, “Well, it is just people taking early retirement or it is people going out of the workforce because they are ill”. There is often a lot of overlap between those two, and we will come back to that.
My view is that a general population-wide issue of increasing chronic ill health is playing out among people who are economically inactive as well as the active and, to the extent that we have seen people drop out of the workforce in the last two or three years, there has been a clear inflection point. Health is playing a role in that, but the question is whether it is a small role or half and half with retirement.
Lord King of Lothbury: You produced some very interesting charts about the movement in different countries. The UK seems to stand out as being rather different from others. How does that fit into your general description of why inactivity has gone up in the UK?
John Burn-Murdoch: It is tricky. It comes down—and we will come back to this point a lot, I imagine—to how we define health as the reason that someone drops out of the labour force. In the headline Labour Force Survey statistics, people are asked for the main reason that they are not active. You then have some people who may not list ill health as the main reason they are inactive. However, elsewhere in the survey they say that they have a health condition that is limiting their ability to do paid work. Then, your loosest definition would be to go a step further and include anyone who simply says they have a health condition regardless of its role in their decision to be looking for work or not.
My preferred definition is the middle definition: people who do not necessarily say that health is the primary reason for their inactivity, but say that they have a health condition that limits their ability to do paid work. If you use that, there is evidence that in the region of 20% of the flows from activity to inactivity over the last couple of years are coming about through health.
If you combine that with evidence that we have on how access to healthcare has been disrupted in different parts of the world over the last couple of years, it feels plausible that that 20%, which is absolutely not a majority, could be related to the fact that British healthcare has been more heavily disrupted than in other countries. We will come on to what I mean by that.
The Chair: We will come on to that, if that is okay.
Lord Layard: That was my question. You have made a huge impact by doing the country comparisons and showing that we are the only country that is right off-trend. If you did the same for the health state or the health states of old-age groups or whatever, country by country, would you also find that we are the one that is off-trend?
John Burn-Murdoch: That is much trickier. I wish that data was available. One of the things that I have found myself coming back to time and again over the last three years is the fact that the amount and quality of publicly available data in the UK is far beyond what most other countries publish. I am able to go into the Labour Force Survey and see, right down to specific health conditions that people list, what might be playing a role in their participation, but that data does not exist for France, Germany, Italy and Spain—at least, if it does, it lags by several years.
At the moment, the international comparison that I make is from data on unmet healthcare needs. That is the percentage of people who say that they needed either an examination or a treatment in a given year and were not able to access it. On that figure, before the pandemic, the UK was already one of the highest in Europe. According to the latest figures around one in six adults in the UK in the last year has needed healthcare access and not been able to get it, whereas in peer countries in Europe the figures are much smaller.
The Chair: Lord Layard, I am very happy to come on to international comparisons later; I just want to follow through.
Q89 Lord Stern of Brentford: My questions are on definitions and data quality, before we get into causal stories. The first is the formal definitions of long-term sickness, mental and physical, used by those who collect the data. What is “long-term sickness”? What is “long-term” and what is “sick”?
The second is about information on data quality in one or two dimensions. How does it seem that those questions are interpreted by the respondents, if there is any feel for that? Are there any problems with differential response? Conditions may determine response: if you drink very heavily, you may be comatose when people ring you or knock on the door, or you cannot open your email.
My final question is on something that you began to touch on and is still a quality issue. How do these answers intersect and overlap with other answers? Could you help us first with formal definitions of long-term sickness, physical and mental, and then help us with those data-quality things—interpretation, differential response and overlaps?
John Burn-Murdoch: I will give an answer, but then I might hand over to David in case he has thoughts as well. There are a couple of interesting things on the definitions. The Labour Force Survey allows people to list one of 17 or 18 specific groups of health conditions that they suffer from. In the last two years, autism was added to that list, and it looks as though the addition of that as a factor may have influenced the number of people who tick any health condition. It is small in the grand scheme of things, but it is just an example of how the wording or the options available can change the response rates.
Similarly, it is worth acknowledging—we will come on specifically to mental health—that, as we as a society are talking more about mental health and there is more encouragement and better diagnosis of that, it is possible that some of what may look like very steep increases in the prevalence of conditions may partly be better detection of those conditions.
In the general definitions, the question is asked: “Have you had a condition lasting a year or more that affects your daily activity?” There are gradations in terms of whether it limits your ability to do your job and whether it severely limits or only partially limits your day-to-day activities.
On general data quality and interpretation, one thing that we might come on to is that there are two important datasets: the general Labour Force Survey measures stocks and the longitudinal Labour Force Survey measures flows. The danger when looking at net change in stocks only is that it may look as though people are going from employment to health-related inactivity, when they may simply be retiring, and at the same time a batch of already retired people are developing poor health. That is where the longitudinal flow analysis comes in, but the tricky thing is that it is a much smaller sample size. A lot of what we will discuss on data is about weighing up those two slightly imperfect sources and seeing whether we can come to the best overarching interpretation of what they show.
Lord Stern of Brentford: Are there problems with differential response across different categories?
John Burn-Murdoch: I do not know the answer to that. I suspect that there may be. Certainly, the trends in most of those health conditions are fairly stable over time, with the exception of categories such as mental health, or particularly autism, which was added as a new category. Apart from the autism instance, there is not a case where one would look at the data and say, “Something strange seems to be happening there”, but I do not know whether there are longer-term reasons why people with some conditions may be less likely to respond than others.
Lord Stern of Brentford: What guidance is there on answering questions on the difference between physical and mental health?
John Burn-Murdoch: I can bring up the questions. They are relatively specific. People are asked whether they have problems affecting their arms or hands, which is one category, problems affecting their legs and feet, which is another category, and problems affecting their back or neck, which is another category. Then you go into difficulties with sight or with hearing, and heart or blood pressure issues. The categories that people are given are relatively fine-grained. However—I spoke to a few other people who work with this dataset—people will perhaps move between the categories, not necessarily the same individual. Two people who have the same condition may tick a different box.
The main headline indicator that I use in my analysis is whether someone declares that they have any long-term health problem. People can tick multiple boxes as well. You can see the impact of people having multiple conditions on their participation rates. You can also see where there seems to be overlap.
Something that has come up in a lot of conversations that I have had with policy experts is that chronic ill health is almost always multifactorial: physical and mental health can bleed into one another and exacerbate one another. To the extent that we get into the details of individual conditions, that is probably less black and white than the headline numbers.
Lord Stern of Brentford: How fine-grained are the answers, and what guidance is given on answering mental health questions?
John Burn-Murdoch: I am not aware of specific guidance beyond simply the overall tick list. There is, especially in the mental health space, probably some scope for different people to interpret the same options differently.
The Chair: David, do you have anything to add?
David Finch: John has covered it really well and comprehensively. It is definitely self-reported within the Labour Force Survey. We looked at health administrative data and compared where you have both those options. It looks as though there is some overlap. You will not always have everyone reporting their health condition if they are diagnosed with one, for example, but the trends and overall prevalence are broadly in line with what we would expect. I suspect that there probably are some reporting issues, but it seems to at least be broadly comparable to what we see in some of the administrative data.
Q90 Lord Rooker: Thanks for coming in and for the material you have been producing. Anybody who has looked at OECD health statistics over the last few years has to start from the presumption that the UK’s NHS is not the best health system in the world, notwithstanding our individual benefits from it, so there is a scepticism there.
To what extent are the increases in sickness-related inactivity occurring among those who are already inactive for reasons other than their exit from employment? The IFS paper of 26 October used five years. There may be other factors. I do not know why it used five years; it was probably a convenient figure that was in the stats.
Has there been a similar increase in those working while still reducing their hours or capacity to work or taking extended time off? Has it affected those who are out already but are struggling on while they are still working?
John Burn-Murdoch: It is a great question. The IFS piece was fantastic. I spoke to the authors in my preparation for today. It is clear from what the IFS did that, certainly if one looks only at overall ill health among the inactive, it gives the impression that the overwhelming majority of the rise in inactivity has come from ill health. This is where I have to confess that that was the analysis I was doing originally, but I have revisited that and essentially replicated what the IFS did and looked at the sensitivity of its findings and conclusions to those different definitions of health-related inactivity.
In essence, if one uses the strictest definition of someone being inactive and flowing from active to inactive due to ill health, in the region of only 5% of the increase in inactivity can be explained by ill health. The bulk of it is people simply taking early retirement, and the bulk of the observed increase in health-related inactivity is from people who are already inactive.
If, however, one uses that intermediate definition that I described earlier where you ask, “Do you have a health condition that would limit your ability to work?”, and then, “Are you inactive?”, that figure goes up from about 5% to the 20% that I mentioned. In neither case are we talking about it being absolutely dominant, but it could be one of two or three relatively similarly significant trends here. It could be that you have people whose health is deteriorating and that has taken them out of work; people who have simply taken early retirement for other reasons; and people who say that they are now inactive for other reasons. That is excluding things such as looking after their children or the home and students. It is just an amorphous “other” group.
In short, it is clear that, to the extent that we are seeing an increase in health issues among the inactive, the bulk of it is within people who were already inactive, and that has fewer direct implications for the labour market but, as we may come on to later, big implications for population health in general, especially because the increase in ill health among the people who were already inactive matches extremely tightly with the start of the pandemic. As we will discuss later, there are some other health conditions that were already increasing before March 2020 but, when we look at people who were already inactive and whose health has deteriorated over the last two or three years, it is a very sharp Q1 2020 onset.
On your other question about whether this is impacting the ability of people who are in the workforce to continue working, I have not explored that specific question, but David and others and I will look into that kind of stuff in the coming weeks.
David Finch: I have a general point on the difficulty in interpreting the data on what is happening over a relatively short period with what is quite a small group once you start to break it down into the different components. There is definitely a difference in what the stock are reporting as their main reason for inactivity in the flow. It is drawing a difference between some people leaving employment and why, against the change in the total stock since the start of the pandemic, which goes back to that broader group who have poorer health and are inactive. It was growing very gradually pre-pandemic, but that is coming through.
The extent to which the pandemic has actually changed that and the extent to which the flows into early retirement potentially continue into future quarters is quite uncertain. It looks as if that has slightly dropped back, but these things are moving around between quarters because of the sample size, essentially.
Lord Fox: That is a helpful cue. David, you might want to pick up on this. We heard from someone else giving evidence that there was a suspicion that people were retiring early because they could. They were actually sick, but it was the easiest way of moving from where they were to where they wanted to be. Is there any sense in the data that early retirement is masking sickness, if you follow my drift?
David Finch: I would say that it is hard to tell, because of the nuanced way in which you can interpret the data. Historically, people have reported reasons for inactivity such as caring for family members, but that individual also has a health condition. It is hard to know what a person perceives as their main reason and wants to report as their main reason for not working or leaving work. It can be that people change their view over time and it can also be that perhaps they have a health condition and are looking to retire but do not quite equate the two things; they just assume that they need to retire. I think there is a difficulty in interpreting some of the data.
The Chair: We are going to come on to pre-pandemic and long-term sickness with Baroness Noakes. I do not want to start jumping around too much; otherwise, we shall end up messing up our question order.
Viscount Chandos: This is a quick step on from Lord Fox. I think what I am hearing from you, Mr Burn-Murdoch, is that some of the data is pretty fuzzy and the causes are multifaceted. Do you think that people’s experience through the pandemic, when they were able to trial a different lifestyle from what they had probably had previously in their working lives, was an additional kicker that pushed or accelerated people to say that their financial circumstances were not going to be the same because they were previously working from home or possibly furloughed? Would that have led to a jump in economically inactive people?
John Burn-Murdoch: It is very hard, if not impossible, to get at that from the data. It feels plausible, and certainly colleagues of mine at the Financial Times have spoken to people who said some variation of that. They were perhaps, as Lord Fox just said, already considering retiring. They spent a couple of months being unable to continue with their business. They may have felt they had sufficient independent financial security from their private pension that meant that they could shortcut that process by a few years. When we look at the disruption of the last couple of years and the fact that many people were, in essence, temporarily retired involuntarily, it feels plausible that some of them would have felt that they could make it work.
The Chair: Very interesting research from the University of Essex looks specifically at men versus women and shows that, with women, it is a rise in retirement that is driving levels of inactivity. To what extent does that compute with what you are saying? Is that masking inactivity driven by ill health among those women, or is it something else? You mentioned a multitude of issues, with the pension being one of them. Is that what we should be looking at?
John Burn-Murdoch: The pension stuff is really interesting. I started looking at it today by looking at people who flowed from activity to inactivity at different points on the income distribution and whether that flow out of activity had been driven by the highest earners, which might indicate that these are people who have sufficient private pension wealth to take that kind of leap.
I spoke to someone today who said that the difficulty is that someone may individually not be a high earner but may be in a household that is wealthy and it therefore affords them that. That will be a really interesting area to look at—who is retiring and what their financial circumstances are.
If we are able to show that the bulk of that retirement is being driven by people of a relatively low socioeconomic status, it would suggest that this has not been born out of financial comfort and is therefore probably more likely to be masking something like health, whereas if it is people higher up the wealth distribution that would suggest that maybe it is people whose health is fairly reasonable but they have decided simply to take the buy-out, as it were.
Lord Stern of Brentford: Again, I have an interpretation issue. You mentioned the tick-up around the beginning of Covid. Do you see any incentive to being seen as vulnerable or long-term sick in the Covid context in relation to treatment or vaccination?
John Burn-Murdoch: I think that is possible, but possibly more likely is simply that, as we all started talking about health more and thinking about whether we were feeling sick on any given day, people may have simply had that higher in their mind. When someone was filling out a survey, a month earlier they might have said that they were inactive through retirement, whereas now they are saying that it is illness. The shock to society that the pandemic created certainly could explain how people began thinking about their health more explicitly.
Lord Stern of Brentford: There is an advantage at least in vaccination—although I appreciate that would not be Q1 of 2020—from being seen as vulnerable.
John Burn-Murdoch: There is, but nobody used the Labour Force Survey to determine who got vaccinated. Even if a small number of people were doing that, I do not know whether it would necessarily show up in the data. As you said, to the extent that we see an uptick, it is in Q1-Q2 2020 rather than when vaccines came online.
Q91 Baroness Kramer: Is the fuzziness and noise in what is happening now changing the long-term trend or is it a short-term pattern that is specific to the moment and circumstance? Within the overall general population numbers, are there some patterns that we should be paying attention to? Is there a difference in health status between immigrant and non-immigrant populations? Could a shift in the pattern of immigration be playing into any of this? What about regional differences? We have tended to deal with high-level general population numbers, and they are probably masking a lot of difference.
David Finch: The changes that we have seen since the pandemic have highlighted this as an issue. It is a bit early to see whether things go back to where they were, but I think it has highlighted what was already an underlying trend, which itself was slightly hidden by a decrease in inactivity in the years before the pandemic among the 65-plus age group. At a high level, you saw inactivity falling, which gave you an overall positive story, I suppose. Then, to some extent, people looked less closely at some of the finer detail.
In looking at sickness trends across the whole population, when we have looked over the decade before the pandemic—you run into some data issues as some of the definitions within the surveys have changed—there was a pre-existing increase in the prevalence of reporting of long-term health conditions across some of the younger age groups, particularly in mental health and depression. That was happening pre-pandemic and we have seen that in some of our analysis of administrative datasets. What is hard to tell is how much, as John said earlier, that reflects increased diagnosis and awareness, in that people are reporting it more because it is being identified more within the population.
As well as trends, there is the scale of the inequalities, which you mentioned, between different geographical areas. You very broadly tend to find that parts of the north-east have higher shares of the population with long-term health conditions. There is variation across the country, but you tend to see that very broadly characterised as a north/south divide.
It is also between different socioeconomic statuses in different areas. For example, a 60 year-old woman living in the most-deprived decile has the same level of diagnosed morbidity as a 76 year-old woman in the least-deprived area. You have a 16-year gap in the extent of diagnosed illness between women living in the least-deprived and most-deprived areas. It is about a 10-year gap for men. It is also important to consider whether that is getting any better, and in general it had not been getting better before the pandemic. The impact of the pandemic would suggest that it has probably got worse.
Baroness Kramer: May I ask something related to that? Looking at the data, is there evidence that income distribution is having more of an impact today than it did historically?
David Finch: I cannot say that it necessarily is, but we have seen less progress. In the last decade there was less improvement in general health—a lot of this is taken from life expectancy and reported good health—but among people in the least-deprived areas we see things at least remaining flat, if not slight improvements, whereas among people living in the more-deprived areas we are seeing some deterioration. It is hard to say exactly what is driving that, but we are seeing more deterioration in the more deprived areas.
Baroness Kramer: Do you have any comment on the differences in health status of immigrants versus non-immigrants?
David Finch: No.
John Burn-Murdoch: I will absolutely look into that, but I have not looked at it yet. Again, one of the tricky things is that, as you reduce the sample size, some of that data will get even fuzzier, but we certainly know that there were significant demographic differences in outcomes from Covid, so it would not surprise me if there were some differences. I do not know exactly what they are, and I would not guess at the direction of travel either.
Baroness Kramer: Are you able to cut the data by postcode, region or whatever?
John Burn-Murdoch: By region, yes, but if we go much grander than that in the LFS, the samples get too small.
The Chair: Is it possible for you to do that by region over time? I was very struck by the fact that there are certain regions where inactivity is markedly higher, but I have seen only a snapshot and would love to see what the trend is pre-Covid.
John Burn-Murdoch: Yes, it is possible. As I said, I have not done it yet but, whether it is me or the folks at the IFS and the Resolution Foundation and so on, I am sure that someone will look at this.
Baroness Kramer: Has anybody asked before about whether there is a difference in health status for immigrant and non-immigrant populations?
John Burn-Murdoch: It has certainly come up before; it is just not something that the people looking at this have had time to get to yet.
Q92 Lord Livingston of Parkhead: I would like to return to the question about the UK compared with other countries. It is obvious that the UK is a big outlier from the OECD survey. Canada and the United States are broadly flat in the employment rate among 55 to 64 year-olds. In Italy, Germany, Japan and France participation in the workforce has increased since before the pandemic. One might say that Italy was struck very badly by Covid, and it is a surprise to see it there.
You mentioned the list of who is least dissimilar to us. When you look at the health status, or indeed the state of health services, in these countries, do you have any idea why certain countries such as the US would be closer to us, but Italy would be very different?
David Finch: We have not made international comparisons between health and employment outcomes. The one thing we have looked at, which I suppose is a broader proxy, was the slowdown in life expectancy improvements—
Lord Livingston of Parkhead: I was looking at general health and health services.
David Finch: We saw that the UK compared quite poorly to other countries—we had one of the biggest slowdowns in improvement, and the US had a similar change. Everywhere had a slowdown in improvements, which suggested to us that, from a broad health perspective, although this was an international phenomenon across other countries, we had a particular worsening, which came back to poorer underlying health within the UK.
In particular, when we looked at mortality by age compared to France, Belgium and the Netherlands, we had similar slowdowns in mortality improvements for older age groups, but the UK was an outlier for some of the younger ages. If you are seeing higher mortality among 50 to 64 year-olds, that is an indicator of poorer health within the underlying population.
John Burn-Murdoch: Echoing what David said, it is interesting that the US has not seen the same sustained increase in people dropping out of the workforce as we have seen in the UK, but it is probably the closest to us in terms of not being anywhere near where it was when the pandemic hit.
It feels plausible to me, at least, that, given what we know about population health and narratives of things such as deaths of despair in both those countries, that would fit with the idea of people deciding to leave or not rejoin the labour force, whether for acute health reasons or because of chronic malaise.
I looked at some more detailed examples from other countries. A comparison of the UK and Spain, for example, showed that they had a very similar trajectory in the timings of various peaks in the pandemic, with both seeing an initial spike in people dropping out of the labour force. But Spain has returned not only to normal but to higher participation, lower inactivity and, in particular, there has been no uptick in people reporting inactivity due to ill health. Again, that at least suggests that there is something specific to population health and, perhaps, the healthcare system in countries like the UK and the US that is causing them to have worse outcomes and slower recoveries.
Lord Livingston of Parkhead: Do you have any cross-country information on waiting lists? Is waiting list information published by different countries, or is the UK, because we have an NHS, pretty unique in reporting it?
David Finch: Sorry; I am not aware that we have been looking at that. I can check with colleagues at the Health Foundation.
The Chair: It would be very helpful if you did.
Lord Livingston of Parkhead: To understand it: is it the state of people or the lack of treatment, or indeed neither of those things, that are causal factors? It is not obvious from the Covid rate why you have been getting very different outcomes.
David Finch: The way we would characterise it is that there was underlying and pre-existing poor health within the population before the pandemic, but backlogs and waiting lists in the NHS are exacerbating the situation. That is extending the period in which people already in poor health are in poor health, and that is deteriorating, or it is potentially leading to a risk that, if people who become sick are not treated fast enough, they could lose employment and then become inactive—
The Chair: We will come on to that in a moment with Baroness Noakes, so I am sorry to cut you off. John, do you have anything to add?
John Burn-Murdoch: Regarding waiting lists, there is a European survey that is done every year, which, unfortunately like many things, the UK has not been part of for the last couple of years. The last time the UK was surveyed, in 2018, around 8% of the population said that they had needed either examination or treatment in the last year and had not been able to get it. That was almost the highest in all of Europe, with only Poland, Greece, Latvia and Serbia behind. This was 2018 and it was already looking as though there was a bigger issue.
Lord Livingston of Parkhead: But over that period we had seen increasing participation in the workforce. The trend changed only in 2020.
John Burn-Murdoch: Yes, we had. I just had a survey carried out using exactly the same wording and methodology in a selection of these countries, and it looks like the UK may have gone from having around 8% with unmet needs to almost 16%. Although it looks as though that figure has increased in other countries due to the pandemic pressures, the gap between the UK and peer nations in people’s inability to access healthcare seems to have widened. Again, it is absolutely not an open-and-shut case, but there is certainly a suggestion of an issue there.
On the same topic, the reason I think that is at least suggestive of something is that, according to the Labour Force Survey data, the incidence of chronic pain—musculoskeletal issues—in the population has not increased, yet inactivity among that group has significantly increased. Again, to me at least, that hints that people had a condition, were unable to get treatment for it and that made them less likely to participate.
Lord King of Lothbury: Many of the comparisons that you have given have been between immediately prior to the pandemic and where we are today. I am struck by the fact that the behaviour of inactivity and labour force participation in the decade prior to the pandemic was very different in different countries. The US saw a sharp fall in labour force participation; we saw an increase, especially among older people. At one level you could say that we had a lot more people waiting for an excuse to retire or, at the margin, suffering a health condition that led them to retire. Is there anything in the experience of the decade prior to the onset of the pandemic that you think is relevant to the conclusions that you are tempted to draw from the comparisons you have made?
John Burn-Murdoch: Yes. I have looked back at this specific point several times. A few people have said, “Well, inactivity rates were already so low here that maybe this is just an adjustment towards the norm for developed countries”. The one push-back I give on that is: we were not exceptional; we had higher participation rates and lower inactivity than the average OECD country, but countries such as Denmark, the Netherlands and Sweden, which might be considered our peers, had similar if not lower rates of inactivity yet have still seen recovery back to the pre-pandemic trend.
That does not at all refute the thesis that you are putting forward, but nor does it, to me, confirm that that is what we are seeing happen. It is certainly possible that we had, in essence, a cohort of people, perhaps in their early 60s, who in many other countries would have retired, and this was the nudge they needed, but there is nothing in the data per se that points especially strongly at that.
Lord Fox: You may have answered this question, but I will ask it in a different way. Either there was something stored up in the workforce that somehow the Covid epidemic unleashed, or the Covid epidemic introduced a new condition or situation that caused its exit. You seem to be saying that there is some evidence that, in a sense, there was bottled-up pressure for these people to leave the workforce before Covid, and Covid may have been the excuse or the stimulus they needed to leave. Is there any evidence to say that, or is it merely supposition? David gave a similar answer to a previous question.
John Burn-Murdoch: Yes. We have used the word “fuzzy” a few times today, and it goes in that category. There is a clear inflection point and that naturally leads people to theorise about what might be driving it—whether it is actual healthcare or people rethinking their employment circumstance. When you have a shock to the system and that produces a result such as this, it feels as though there must be some mechanism driving it, but the data is certainly not sufficiently clear-cut to say that it is or is not any of these things.
David Finch: I broadly agree with John that there is probably a combination of factors going on, and we cannot clearly say from the data exactly what is driving it, but clearly it is such a shock. People are taking time away from work and are potentially furloughed for extended periods. We talked about people accessing pension pots. There are some clear mechanisms through which you can see that people who were already close to leaving the labour market would have been accelerated by the pandemic, but I do not think that that is the sole explanation. It is probably one of several different reasons.
Q93 Lord Layard: If I understood what you said about five minutes ago, John, looking at the trend in different countries and the percentage of people saying that they are unable to access health support, Britain is further off-trend than any country. That seems to be one of the most important pieces of evidence we have had that could possibly link to your previous evidence that we are also off-trend in our inactivity rates—and it points the finger at health.
Let me ask you a different question, which might go the other way. Can you tell us about the comparative trends in mental and physical health conditions in the population, including those in work and inactive people? Most of the NHS waiting list problems are in physical health, and if the main change is in mental health and there is not a big change in the incidence of physical ill health, that would tend to question the emphasis on sickness and the explanation of inactivity. Will you both say something about the comparative trends?
John Burn-Murdoch: Absolutely, and I think David might have interesting things to say on this as well. I believe I am right in saying that it is not uniquely or especially in physical health that we have these wating list problems. I believe I saw something saying that the waiting times for people who turn up to A&E with a mental health condition, which does happen if someone has had an episode, are worse than for those with a physical condition. Certainly, mental health provision from community teams was severely disrupted during the pandemic, so I do not know whether it necessarily follows that the disruption and waiting times that we are seeing are more likely to affect physical conditions than mental.
I broadly agree with your point about unmet need in healthcare. It underlines what most of us already believe: that even back in 2017, 2018 and 2019 we had a waiting list problem in the UK. We were way off target from where we were a decade ago. As your colleague mentioned a few minutes ago, the fact that we already had that issue in 2018 suggests that, although the situation has certainly worsened, it is not necessarily an open-and-shut case that the worsening of the waiting list scenario is what caused that inflection point, because we were already on a worsening trend and most countries have also seen a worsening over the last couple of years. It will have been a factor and certainly will not have helped, but whether it is sufficient to explain an inflection point remains to be seen.
David Finch: There are two different ways in which the delays for treatment could be affecting employment. First, if you are already inactive and waiting for treatment to go back, your inactivity or barrier to going back to work is, in effect, deepening, because your existing health condition is not being dealt with.
Secondly, people who are becoming sick for the first time are off work for longer, but they are still in employment. We have not particularly looked at that here. Some recently published data showed that the duration of sick notes has been increasing recently, but those people are still in employment. The risk is that, if that continues too long, it will eventually lead to further inactivity.
On the point about disruption to services, again from the data that I am aware of, that is not limited purely to physical conditions; it has also been seen for mental health conditions.
Q94 Baroness Noakes: I would like to look at trends in long-term sickness, starting with the pre-pandemic era and then looking at what impact the pandemic has had, including on inactivity. Will you describe long-term trends in sickness in the UK? Had they been increasing before the pandemic? To what extent do we understand the drivers for whatever was happening in the statistics? May we start with that and move on to what impact the pandemic has had?
John Burn-Murdoch: Straight out of the gate, long-term sickness in the population had been increasing. Part of that, we would expect, is due to an ageing population—even the working age group is slightly ageing. There was an upward trend, but that has steepened significantly. The steepening in overall long-term health conditions happened in at least one quarter, if not a few quarters, before the start of the pandemic.
The very unhelpful area of “other” health conditions is where there was most clearly an acceleration and inflection point closer to 2018 and 2019, whereas when we come down to more specific things, such as issues with the circulatory system, blood pressure and breathing, you see a much clearer and sharper rise as the pandemic hits.
Similarly, as we might expect, although levels of depression and anxiety have been increasing very steadily, there was an uptick and acceleration at the start of the pandemic. Again, the question is: how much of that is prevalent and how much is increased reporting? As Lord King mentioned, we may have been prompted to think more about those things.
Diagnosis rates of some learning disabilities are improving, which is part of this as well. There are different patterns in different areas, but it is certainly true that for many conditions, the increase—the acceleration—started just before the pandemic. I am sure that David will have a lot to say on this, too.
David Finch: I do not have a huge amount to add to what John said. Part of the difficulty is trying to understand how much of this is greater awareness of conditions and increases in diagnoses against actual worse prevalence. Looking at both survey and administrative data, what we tend to pick up most are changes in mental health conditions being reported, which gets very entangled with awareness and diagnosis. It is quite hard to unpick precisely what is happening.
Our broader concern is the demographic shape of some of this. We have a more ageing population, and the risk with the older part of the workforce is that, if they have poorer health, they are more likely to become inactive, so it becomes a bigger problem. Before the pandemic, there was a rising population of people inactive with long-term conditions and, as John said, that has slightly accelerated more recently, but it appears to be part of a longer-term historical trend.
On the flip-side to that, when we looked at some of the employment outcomes and patterns of hours worked, we saw that some of the outcomes for people with reported mental health conditions have not particularly worsened for those in employment, which suggests that this increase in prevalence is partly about greater awareness among the people being surveyed.
Baroness Noakes: Are we able to determine to what extent the increase in inactivity is due to things that were already happening in health trends and what the specific impact of the pandemic was?
David Finch: As we were discussing at the start, the actual pandemic-specific impact appears to be coming through a couple of channels: people retiring earlier and people with poorer health.
When people have looked at flows around this, they have seen that, although the total population of people who are inactive and in poorer health has increased since the start of the pandemic, some of the flow analysis points towards there being more people who were already inactive with poor health coming through into the population. It points towards that being a pre-pandemic issue that has been made worse by some of the impacts on health during the pandemic.
Baroness Noakes: John, in answering will you say whether long Covid is having any measurable impact?
John Burn-Murdoch: Long Covid is really interesting. The tricky thing is that it has to go under the “other” category in the Labour Force Survey, which means that it is lumped in with many other things and, as I said, that category as a whole started accelerating before the pandemic. Of course, that does not mean that long Covid has not played a role, but we cannot see that from this data alone.
Interesting work has been done by the IFS, among others, looking at the impact of long Covid. I believe its finding was that, although it has disrupted people who are in employment, by reducing their hours or leading them to take sick leave, it has not tipped people out of activity, at least in the aggregate level. The best evidence is that long Covid will be playing a role, but it does not seem to be a key driver of people moving from activity to inactivity.
The Chair: May I build on all these questions? John, I have one of your excellent articles in front of me, and it shows the really sharp rise in chronic illnesses during the pandemic. We have that evidence. I also come back to the University of Essex work, which shows that the rise in inactivity is slightly higher among women than men. It shows that rising sickness is the most important reason that men are not looking for a job, while rising retirement, relative to trend, is the most important reason that women are not looking for a job. Given the massive rise in chronic illnesses, has anyone disaggregated the men and the women in these charts?
John Burn-Murdoch: I suspect that I did at one point, and it is buried on the cutting room floor, as it were. This may be another example of why this is so fuzzy, because there are so many reasons why this could be playing out differently for men and women. It could be to do with specific health reasons. There have been various things with Covid. I believe that self-reported long Covid incidence is generally higher among women, whereas mortality risks from Covid are higher among men. Of course, many health risks are higher among men, but this is not clear. This is a classic example of how there are both physiological and sociocultural reasons that could each explain this, and we do not know which it is.
The Chair: If you have a chance to look on your cutting room floor to see whether it is among the pile of evidence, that would be very helpful.
Lord Livingston of Parkhead: I have a data question for you, John. I do not know whether you have seen the Labour Force Survey work that looked at where the rise in inactivity is happening between different types of occupations and different wage levels. Some people have taken their pension and retired, so you would expect to see a greater rise in activity at the higher end. Some were working from home and could not be bothered going back; again, you would expect to see a higher rate. Yet what we have seen is no rise in inactivity in the lowest wage set quartile, but a huge rise in inactivity in the lower-middle wages, and over half of that appears to be accounted for in the areas of wholesale, retail, transport, storage, manufacturing and health. People have to go to work, but they are in declining industries. We have had one witness saying, “I don’t recognise the data”, but can you correlate that with everything else that has been said?
John Burn-Murdoch: Who did that analysis?
Lord Livingston of Parkhead: It was the University of Essex looking at the Labour Force Survey.
John Burn-Murdoch: I have not seen that data, but I have been meaning to look at the occupational split of where people have dropped out of the workforce.
Lord Livingston of Parkhead: I will send you the information in return.
John Burn-Murdoch: Again, it would fit with the idea that this is being associated with—I was going to say driven by, but that is too strong—a sense that people have had enough, and that is why they are leaving activity. It may be that they have had enough due to health reasons or stress, or because they have been working longer than their peers in other countries, perhaps. It could be all sorts of things. It would fit with that, but, again, it is very difficult to know. It may simply be that those industries were more disrupted during the pandemic, and therefore—
Lord Livingston of Parkhead: They were all on furlough.
John Burn-Murdoch: Yes and that could play out in different ways—it could be that that gave people a taste of what it would be like not to work every day. It could equally be that the act of not being at work meant that people’s situation worsened, both physically and mentally. It is interesting, and I am sure that more work can be done on that.
The Chair: Lord Fox is going to ask the smallest question ever.
Lord Fox: If that was the case, we would expect to see a burst of people leaving and it going back to the long-term trend. Is there any sense of that?
John Burn-Murdoch: Not yet, but it looks as though the sharp increase in activity that we saw over the first couple of years may be slowing. It is a very good point. Nobody here would expect this to continue indefinitely. When the increase slows, we will start to get an idea of whether this was a two-year shock or something more steady.
The Chair: Lord Livingston said that if this is a group of people who do not have a large amount of savings, who retired for whatever reason and who are facing the full ferocity of a cost of living crisis, you might not see it stop; you might see it unwind. The question comes back to what Lord Livingston was getting at, which is: if they have left occupations in declining industries, where are they going back to? We do not know the answer to that question.
John Burn-Murdoch: Yes, and that may be another cause of a further increase in ill health among the already inactive. These may be people who left while being of reasonable health and now their health may deteriorate.
Lord Monks: John, you mentioned the data being rather fuzzy at times. Trying to distinguish a clear pattern through it all is not easy. One thing that is becoming clearer in this great problem is that the NHS is absolutely central to it—its performance, its ability to get people back to work, it being able to offer some people a positive future who might not think they have one, and so on. It is also a major employer. It must have suffered—I am sure it has suffered—great shortages of labour as doctors have been retiring early. We seem to read about that quite often. I am trying to pinpoint, when we come to our recommendations, which ones we should focus on.
Do you agree that the performance and condition of the NHS are absolutely central to that? Often the NHS finds itself in a vicious circle—its staff get sick, or leave, and it is short. The staff who remain are more vulnerable, and so on. Do you agree that the NHS’s performance, as an employer and service provider, is absolutely crucial to the future?
John Burn-Murdoch: It is, certainly. There is also a danger that we are too NHS-centric in the policy solutions. A lot of what we are talking about could be broader social determinants of health, with rates of poverty increasing and what we talked about in terms of deprivation. A lot of these issues can be the fundamental reason why people slowly see their health deteriorate; maybe later they need to interact with the NHS and cannot do so.
I absolutely think that the NHS needs to be enabled to work better, to work more effectively, whether that is through more funding or through a different allocation of funding and resources. This is a society-level issue.
It is about occupational health, as well. A key is to enable people who may have health conditions to stay at work or to help those who recently left the workplace to get back into it. This is multifaceted, but of course the NHS is part of it.
David Finch: Clearly, the NHS is very important in helping to deal with some acute health issues. We have not seen it as the main reason why you are having these increases or why you have the inactive population with poor health. We would point towards longer-term issues, as John was saying. There are wider or social determinants of health leading to people having poor health in the first place. That is happening to people over the long term in their experiences, through their lifetime—some of the big inequalities such as deprivation. Effectively, the conditions people are living in point towards that as a key driver of some of the poorer health being seen.
There is a risk that focusing on the NHS means that more employment-focused types of activation policies are overlooked. Some of the existing DWP schemes, although they are talking about people with poor health, or helping people who have been out of work for longer, are for the longer-term unemployed rather than people who are inactive. Historically, it has been really hard to get an effective policy in place to help overcome barriers to work for people who are economically inactive and have health issues. That is partly because they have health barriers and those health issues can vary massively even between specific conditions. People can have different severities and have different barriers to their health. By being out of work for an extended period, they have quite a significant distance to get back to the labour market.
There is a risk that an overfocus on the NHS from an employment perspective means that some of that very hard work to get better at helping people with poorer health back into the labour market is overlooked. That should be a priority focus, looking forward.
The Chair: I am very conscious that we are over time, but I need to ask one final question because I am very struck by what you say. I hear what you both say, but I come back to the Essex University research. It says, “Almost all of the rise in inactivity is from workers who think that they will definitely not work again. That is completely unlike what we saw during the Great Recession”.
That is a real red flag. One of the policy recommendations could be that government need to do more to help people back into the workforce. If people are saying that they are not going to work again, for whatever reason, we now have a cohort who is out of the workforce. Have you, John, seen any evidence of that finding supported by your excellent work?
John Burn-Murdoch: Again, it is not something I have seen but it is not something I have seen evidence against, either. Think of this as a chart. We have been going down and we have come back up. We now go back to being parallel to where we were, but perhaps inactivity is always around one percentage point higher. That would not be out of kilter with what we see in a lot of other countries. Maybe we have reset to an inactivity level that other countries see. Again, it is too early to know.
It equally may be that, over time, this turns out to be very specific to the pandemic cohort and we end up in the medium term—five to 10 years—back on the trend we were. It is very striking hearing that, but how exactly it will pan out remains to be seen.
The Chair: I am conscious that I have tried everyone’s patience. Thank you very much to you both for the work that you and your organisations have done. You have made a fantastic contribution to the debate. Your articles and your publications have been very good. Many thanks and thank you for sparing your time. Sadly, as we now know you, we will come back with further questions.
