21
Examination of witness
Huw Pill.
Q89 The Chair: Good afternoon. Welcome to this hearing of the Economic Affairs Committee. I am delighted to welcome Huw Pill, the chief economist from the Bank of England. Thank you very much for making time to come at relatively short notice; it is very much appreciated. I know that you are under time pressure. We have 75 minutes with you, I think I am right in saying, so we will endeavour to use that time very precisely.
The Government today published a White Paper, Get Britain Working, about inactivity. For those who might be watching, we are not going to get into detail on that because it has literally just come out. We are not going to put Mr Pill under pressure to answer questions about a document he has not yet had the chance to read. I will start by asking two questions about where we are in a macro sense, and Baroness Wolf likewise, and then we will get on to the main substance, which inactivity in the labour market.
I will start by asking about the Budget. As you know, a few weeks ago this committee published a major report into the UK’s debt sustainability. I am interested to know your views, as things currently lie, on the impact of the Budget on debt sustainability overall and whether it has in the medium to long run increased the risks to our debt sustainability.
Huw Pill: First of all, thank you for inviting me today. It is a pleasure to be here.
We have obviously made an assessment of the Budget through the lens of our own mandate, which as a member of the MPC is a mandate to preserve price stability and achieve the 2% inflation target. That rests on fiscal sustainability being maintained over the long term. But our focus is obviously a more cyclical frequency of developments and we leave it to our colleagues in the Office for Budget Responsibility, who have the main responsibility for assessing the long-term sustainability of the public finances.
If you ask me to focus on the long-term fiscal sustainability point, there have been some changes to fiscal rules. That probably is where the focus should lie because, if those fiscal rules are respected, we would have confidence that fiscal sustainability will be maintained. That is the sort of issue on which we rely to be able to run monetary policy in an independent way so as to achieve our mandate. Those changes in fiscal rules, from that perspective, are relatively minor and modest in character. We would be confident that as long as those fiscal rules are respected, fiscal sustainability will be maintained.
The Chair: I will come on to Baroness Wolf in just a second to pick up on the fiscal rules point, because I think she might want to come in on that. Obviously, the OBR has made a prediction on this and you may say that that is all you want to say on it, but what impact will the measure that is getting a lot of attention, NI and employers, have on the labour market from your perspective?
Huw Pill: If you go into the specific measures, again viewed through the lens of an MPC member, I should perhaps start by saying that the overall view—the overall macroeconomic outlook and the overall outlook for monetary policy and inflation developments—is still governed more by the unwinding of the effects of those very big global macroeconomic shocks that we have faced in the last few years. I am thinking particularly of the pandemic and the invasion and its impact on food and energy prices.
I understand why in this building and in the media there may be an orientation towards looking at the impact of the Budget, but for us the big story remains that big macroeconomic story. For us on the MPC, it is important to remember that we are not doing a forecast, if you like, pre the Budget measures and post the Budget measures. We are doing an August forecast and a November forecast. Many things other than the announced government fiscal policies have changed between August and November, not least the shape of the UK yield curve. Crucially, as I have already said, our forecasts are based on fiscal policies announced by the Government. In terms of the news they impart to our overall assessment of the inflation outlook and therefore for monetary policy at the cyclical frequency, I would probably not want to overstate the role that Budget measures in general and the NICs changes in particular have played.
Having said all that, and as I think was discussed in the Treasury Committee and in our own published Monetary Policy Report, when we look at the overall impact of the Budget, we are probably looking, at peak, at an increase of about half a percentage point to inflation in the shorter term. That does tend to wash out as we go to the more policy-relevant medium-term horizons. Part of that comes from stronger growth and therefore more pressure on resources. The slack in the economy that was previously anticipated to open up is opening up less quickly, and that leads to a higher profile for inflation than we otherwise expected. There are a number of direct measures to prices, changes in VAT—more on where it is applied and not to the rate, of course—but also changes in excise duty and fuel duty. Some of these are innovations, based on the way we do our forecast, but the supposedly temporary reduction in the fuel duty has been rolled over and rolled over and rolled over again. So to some extent our forecast is perhaps a little bit of a misleading guide to what our assessment of the impact of that is because we had anticipated that in coming to our policy conclusions.
Finally, to get to your question, the NICs is another layer on that. Our baseline view in our published forecast is not dissimilar to what we have from the OBR, and that means a small positive impact on inflation but not one that is really of note from a macroeconomic monetary policy point of view. We do recognise in our analysis that an increase in national insurance contributions widens the wedge between the producer wage, the wage paid by firms, and the consumer wage, the wage received by wage earners. That is a disincentive to employment at some level. In the context of our discussions this afternoon, at the margin that probably is weighing against participation in the labour market. We do not think that it is an enormous effect, given the quantities involved.
The key issue for us from a more cyclical point of view is whether the higher costs implied by the rise in NICs are being passed through by firms directly into prices, which would be of obvious concern to us because that would be a perpetuation of higher than target inflation, and whether they will result in some reduction in wages—which would also have implications for us to the extent that that impacts on demand—and/or in reduced levels of employment, which similarly would have an impact on reducing demand. Of course, one would wish to see higher productivity be a source of being able to cover these costs, but as you know all too well, that has not been a feature of the UK economy for some time.
The Chair: We may well come back to productivity in a moment. Lord Griffiths, do you have a very quick question? We are very tight for time.
Q90 Lord Griffiths of Fforestfach: It is a very quick question. Given your inflation objective and if you take Dave Ramsden’s speech, there is an awful lot of tackling inflation by looking at the breakdown of individual items, whether it is airfares or energy prices or whatever. What assurance do we have that as you are doing this you also have a target for bringing down demand by setting a target for monetary growth at the same time? Sometimes I feel there is huge emphasis on the immediate and on a relative price effect, but on the other hand you are in danger of monetising that price effect. I just wonder how you manage to keep the balance between the two.
Huw Pill: I have tried to emphasise in my previous remarks that from my perspective—I should not speak for my boss, Sir Dave Ramsden, one of the deputy governors—we have a medium-term focus and a focus on aggregate inflation. Of course, we are trying continually to learn about how to extract the signal over the medium term about developments in the CPI. As you know, that is the variable we have been asked to target as the MPC. I think that does require us to interpret developments in individual components in so far as they give us information about what the medium-term developments in the aggregate inflation are.
One of the things that we have been doing of late, as I am sure you are aware, is that we have tended to put a little bit more weight on developments in services price inflation and, although this is not a component of the CPI, in terms of economic drivers, domestic pay growth. These are two things that we think have a more persistent impact on overall inflation, whereas developments in food and energy prices can be very transient. They are driven by global factors. On developments in goods prices, we are increasingly of the view that we are a price taker, even on goods that are domestically produced, because most UK manufacturers are serving a global market and are therefore pricing to international markets. Therefore, monetary policy has relatively little influence on those and we have to look through some of those developments. That is how we would see it in terms of trying to interpret price and cost developments here in the UK.
You make a good point that in terms of looking to the medium term and thinking about our framework, we do need to have a nominal framework for monetary policy. You might talk about that as looking at nominal activity, nominal GDP, or perhaps equivalently to some extent looking at developments in measures of money, broad money. These are not things that are emphasised in the discussions of the MPC, but they are certainly things that I look at quite regularly and sometimes introduce into the discussions at the MPC. Those have shifted a little of late. There was very strong broad money growth in the period after the onset of the pandemic, associated in large part with Bank of England measures such as quantitative easing. Subsequent to that, when quantitative easing came to an end, we saw quite slow developments, in fact contracting developments, in broad money. Those have now also played their way out of the system and broad money growth is beginning to expand at modest and relatively stable rates of around 3.5%. From my perspective, that is quite an encouraging sign. Developments in nominal GDP in the UK economy are certainly not strong—and, of course, we would like them to be stronger because of stronger productivity and stronger real growth—but in terms of having a sustained expansion of the nominal economy, consistent with achievement of price stability over the medium term, I think we are in better shape than we have been.
The Chair: Excellent. Thank you. We have done the macro. I will move straight into inactivity now as we have one hour left with you.
Q91 Lord Lamont of Lerwick: In the Mansion House speech the governor made a lot about the problems with the LFS, the Labour Force Survey. What are your assumptions—I call them assumptions but I might say conclusions, because you have to have a view—about the likely trends in labour market inactivity? How sensitive are the Bank’s forecasts to these? The second part of the question about the sensitivity is as important as the first part.
Huw Pill: In making these assessments, we undertake an annual stocktake of the supply side of the economy. We are about to enter that process right now, so I cannot give all the details of that. That will be discussed at the end of this week. In that regular annual supply stocktake we analyse developments in productivity and developments in other aspects of the supply-side capital deepening and so forth. In particular, we look at developments in labour activity and labour participation. That has been a source of quite a lot of noise and disturbance to the economy, particularly in the period since the onset of the pandemic.
Our current assumption has been that we expect to see a trend decline in labour participation. That is what we have seen over the last four or five years. The pace of that decline in participation, though, has been moderating of late. These are the things that we will be reviewing as a result of the ongoing stocktake. I should say that when we do that assessment, of course, we look at the labour force statistics, and when we make our forecasts we publish some forecasts for participation to be in line with the labour force statistics. But, as we will no doubt get to, there are question marks over the labour force statistics, which the governor referred to in the Mansion House speech. Our analysis is augmented by a whole set of other sources of data; we are trying to enrich and deepen our understanding of what is going on there. Perhaps I will leave that, because I imagine we might come back to that.
On our policy assessment, a little bit related to what I said to Lord Griffiths, one part of our policy assessment is to try to understand inflationary pressures emerging as a result of pressure on resources. We have a pretty good insight into the demand side of the economy. We can see the amount of spending in the economy through data on GDP, consumption, investment and so forth. We do not have such a direct view of the supply side of the economy. We are continually having to make and update our assessment of how strong growth in the supply side of the economy is, and what the level of potential activity or potential output in the economy is, in order to construct measures such as the output gap, capacity utilisation or the unemployment gap, which we all see as measures of trying to understand whether there is pressure on resources in the labour market, the goods market or the capital market. The greater the pressure on resources, the more likely it is that we will generate inflationary pressure. Understanding how the economy’s productive potential is evolving through time, one important driver of which is the participation rate in the labour market—the number of people who are active and therefore, at least potentially, contributing to producing things in the UK economy—is pretty central to that.
When we look at cyclical developments in the economy, we try to look quarter to quarter, year to year, trying to understand where there may be sources of slack in the economy. You might think about whether the number of hours worked per worker has reduced. People who are in jobs could work longer hours. That would be a source of more production that would not affect the labour participation rate. We also look at the number of unemployed people in the economy: people who are unemployed becoming employed and therefore producing something.
There is a third element, which has been in focus of late and which we call the labour participation gap. That is the gap between how many people are participating in the labour market and what we might think is the equilibrium, normal trend rate of labour participation. That has been a source of slack in the period since the pandemic. In other words, given our assumptions about where we think labour participation will evolve through time, on the basis of demographics and trends, labour participation has been lower. What we are now seeing is that that has corrected through time. Our assessment, partly based on looking at data outside the LFS, is that now we do not see that as such a source of weighing on the degree of slack.
The Chair: In layman’s terms, in this debate about data—I want to come back to the second part of Lord Lamont’s question—I am still trying to gauge in my mind how critical this is to you when you are making decisions about rates. Is it something that is a factor and, in the list of factors, how high is it? I am asking the question in a slightly different way, but I am still trying to clarify in my mind how big a problem it is for you, not having the clarity on the data.
Huw Pill: I would distinguish two things in response to the question. I hope I understood Lord Lamont’s question correctly. I was trying to answer that question in the context of labour participation. I think that is important, although it is probably not the key source of slack and pressure on resources and so forth. We are looking more at things such as unemployment and the ratio of vacancies to unemployment. Those are probably more key measures of slack in the labour market. We have a whole set of constructions about measures of capacity utilisation, the output gap and so forth, when we are looking at it.
The labour force statistics—statistics including from the Labour Force Survey, other than participation—are all constructed simultaneously, of course, because labour force participation drives unemployment and is also associated with employment in the economy. This has been particularly the case over the last few years, at a time when, because of all the big disturbances to the economy—I listed some earlier: the pandemic, the invasion and so forth—there have been big shocks to the supply side of the UK economy. As I think we have demonstrated, and other central banks and other forecasters in the UK have demonstrated, it has been very difficult to forecast the UK macro economy of late. One of the things that the Bank of England and the MPC have been forced to do is perhaps to put less weight on forecasts two or three years hence, because those have proved more volatile and uncertain, and we have had to put more weight on what we have called late-cycle indicators. We have tended to emphasise developments in services price inflation, developments in pay growth and indicators of tightness in the labour market. We think that those three things are associated with domestically generated, more persistent movements of inflation, which are precisely beginning to address the medium-term underlying nominal dynamics, which I think is the true measure of whether we are achieving our mandate to maintain price stability and achieve the inflation target.
In that context, what we are looking at a lot is employment growth, unemployment rates, employment rates and, almost by residual, participation rates. The participation rate is perhaps not the crucial thing but the labour force statistics, and particularly the Labour Force Survey, are crucial elements.
The Chair: That is a very good segue into Lord Blackwell’s question.
Q92 Lord Blackwell: Mr Pill, as you have just been saying, you are on the record as having pointed out some of the deficiencies in the Labour Force Survey with the shrinking number of households replying. One of the other areas where we have been looking at statistics is trying to understand the source of inactivity and the transition between employment, unemployment, sickness and long-term sickness, where again the statistics seem to be difficult to interpret. Do you want to comment on that particular aspect of the statistics? Do you have suggestions for what better statistics we could and should be collecting to help understand these issues?
Huw Pill: The challenges in collecting these statistics are formidable. I think we should be careful about suggesting that one can pick and choose. The Labour Force Survey is a single survey that leads us to have information on employment, unemployment, participation and reporting on the causes of inflation, including long-term sickness. Basically, you are left with two approaches. Do you improve the Labour Force Survey? There are efforts under way to do that, which I think are quite admirable. One of those in the shorter term is to raise the response rates, and resources have been placed into the Office for National Statistics in order to raise those response rates, to have more face-to-face interviews, to have more people go and conduct those interviews. It is challenging because the basis of that survey is calling people up. My teenage children never answer the phone when I call them up. I complain to them and then they point out to me that I never answer the phone when my parents call me up either. There is an element of people not answering phones in the way we used to, and I think that was at the heart of what the governor said in his Mansion House speech. There is something perhaps distinctive about the UK that people do not answer their phone. I cannot really comment on that, but that certainly is reflected in these results.
There is an effort under way to create something called the Transformed Labour Force Survey. This is all under the auspices of the ONS, so I am reporting its work here. That is an internet-based survey. I think that probably is the way to improve participation and response rates and ultimately lead to better statistics, at least in the intermediate case.
Then there is an alternative way of looking at it. If we cannot make the Labour Force Survey better, and that will take some time, we could also look at other sources of data. There are a number of reports, including a recently published report by the Resolution Foundation that is quite consistent with work that we have done internally at the Bank. We have collaborated with colleagues at HMT and at the ONS in conducting that work. The people who work in all those institutions are all generally people who have worked in others of those institutions at some time, so it is a very fluid and integrated set of economists.
We do look at other sources of data. A pool of data there is the administrative data and particularly what is reported in the data collected by the Department for Work and Pensions. When you ask people, “Why are you not active in the labour market?” and people say, “That is because I am long-term sick”, the increase in that number since the pandemic, if you go through the LFS route, is quite significantly higher than the increase in the administrative data, as reported by the DWP, on people applying for and able to claim various disability and sickness benefits.
That creates two things. It probably says the direction of travel is towards higher long-term illness and sickness being a cause of lack of participation. I think that is quite a robust conclusion. It is a significant number, but there is uncertainty about it. We do not really have a way of resolving that, given the information that is available to us.
Lord Blackwell: In this administrative data, the system knows the names and identities of people registered as unemployed and claiming unemployment benefit. The system knows the identity and names of people claiming sickness benefit and long-term sickness benefit. In terms of understanding that coherent picture of who is moving from what benefit over time, whether the long-term sick are coming straight out of the labour market or whether they are unemployed for a period, is that just a question of throwing the resources at that databank in order to get a better analysis or is there an inherent problem in the incompatibility of the data?
Huw Pill: We have a whole machinery that has been created of late, in recent times, of big data analysis. There is tremendous potential within the Government to use administrative sources of data to do analysis of the type you are proposing. Some of that analysis has been done. It is not always easy to do, to match up different datasets and get access to different datasets. This is something that many of my colleagues at the Bank would like to do for analytical and research purposes, and there are challenges to that. Other countries have made more progress in integrating their data and using it more effectively and efficiently.
I also think it is important to keep in mind that there is a danger in that, because you have lots of data and you can analyse it in a certain way, you should not lose sight of what the question you are asking is. If you are asking the question, “Are people shifting from unemployment benefit on to sickness benefit, and what are the incentives to do that?”, the approach you are describing does make sense. If you are asking the question that I think is more relevant to a member of the MPC undertaking the macro-type analysis that we were talking about earlier, which is, “Is someone active in the labour market, looking for a job, pushing down the pressure on resources that otherwise would lead to higher growth in wages, higher growth in costs and higher growth in prices?”—in other words, an inflationary dynamic—you have to think about the economic set of issues there and the incentives underlying that.
For example, if we were to see changes in the relative attraction of unemployment benefit versus sickness benefit or the application of rules through time, and we know that during the pandemic there have been variations of those things—partly for practical reasons because certain assessments could not be made in person, for example—those perhaps might generate more news in what we are doing than understanding from a macroeconomic point of view.
I guess the point of what I am trying to say is that I do not think the use of the administrative data is necessarily a panacea for all problems. It is a question of being clear what the problem you are trying to address is, being clear how you are going out to analyse it, and then using an appropriate set of techniques and statistics in order to have a view on that. For many of the macroeconomic questions that I am dealing with in my day job, as it were, it is important to have at least a functioning Labour Force Survey and, I hope in the not-too-distant future, Transformed Labour Force Survey data, internet based, which has higher response rates. I think that will give us an insight that we will not get just from the administrative data.
Q93 Baroness Wolf of Dulwich: I want to follow up a bit on that. I think I heard you say earlier that you had expected increases in activity rates and they were increasing more slowly than you had expected. What was the basis for that? That is obviously the macro question. What drove those judgments?
Huw Pill: Of course, it depends on over what horizon you are talking about. If we look back over the last decade, say, participation rates were rising up until the pandemic and then falling from the pandemic onwards. It is a peak-shaped relationship. What were the main driving forces ahead of the pandemic? There are essentially two factors here. This is accounting in some sense. One factor that we should not lose sight of is that we have an ageing society. As people get older, other things equal they are less likely to participate in the labour market. That would suggest, other things equal, a downward trend in labour participation. Offsetting that in the period from 2014 to 2019, participation was rising in each bucket of population. A big part of that, for example, was the increase in the female retirement age towards male levels, which increased participation in the 60 to 64 bucket, if you want to use that language. These are the main forces that are driving our longer-term assessments.
What we have seen since the onset of the pandemic is that other than for prime working-age females where participation has gone up in the bucket, and for over-65s where population has gone up in part because the state pension age has risen, in all the other buckets the participation has now decreased. You have an accentuation of the downward trend coming from the overall demographic profile. The UK population getting older is being added to by declining participation in many of the buckets. That is what leads overall to this spike shape that we see.
Of course, what I am doing is describing data. This is all known with the benefit of hindsight. In terms of understanding the demographic trends, we are making assumptions and forecasts, but demography is one thing that, in principle, ought to be relatively straightforward to forecast. People get older; there is a cohort effect and so on. What has proved harder to forecast—so this is where the impact of the pandemic did come initially as a surprise to us—is how the response was going to be within each cohort. Crucially, I think there is still an open question, at least in my mind, whether the decline in participation we have seen is now everyone as they reach a certain age, say 60. Are we still going to see lower participation rates for people reaching 60 on an ongoing basis, or is it just that group of people who became 60 during the pandemic and maybe were locked down? Is it just one vintage, if you like, that had lower participation rates and we will see a recovery, or will it be a more permanent thing? That is an open question.
Q94 Lord Burns: You have described how you have to look now at a range of information and try to piece together the various messages that are coming from that. What conclusion do you reach about the state of the labour market today? Some of the statistics seem to show that there is slack in the economy, yet many of the stories we read are of shortages of labour in many areas, particularly in hospitality. What do you think is the uncertainty range surrounding this question of the tightness of the labour market, and where do you put it today? How difficult has this made your job over the last year or two?
Huw Pill: I can quote you from the Monetary Policy Report, which is the official assessment of the MPC. That says the labour market has eased over the last year or so but remains reasonably tight by historical standards. That seems perhaps to be having it both ways, and I think that is a reflection of the level of uncertainty that we face. Unemployment, which is probably an imperfect measure for some of the reasons we have talked about, has remained remarkably low through this cycle.
Vacancy data does not come from the Labour Force Survey. It is not independent but it is a distinct measure that comes from a different survey. Partly thanks to the work of Christopher Pissarides at LSE, we think a lot about matching in the labour market and, as a result of that, the ratio of vacancies to unemployment has become an important indicator of labour market tightness to complement the traditional unemployment gap—unemployment relative to some measure of natural rate or NAIRU-type measure. We are looking at all these things. The ratio of vacancies to unemployment has fallen very significantly over the last year and a half, albeit from very high levels, essentially unprecedented levels, and is now, roughly speaking, at the level it was in 2018-19 prior to the onset of the pandemic. Those measures are the basis for thinking about why we have eased but are still relatively tight by historical standards. We do not think there is enormous slack in the labour market.
The open question here, and this goes to some of the things we are talking about, is that those measures do not really capture the possibility that people left the labour market but, in the language that at least we use, they are marginally attached. In other words, they are people who are no longer actively seeking a job but if a job were to become available they would take it. That worker would not count towards being unemployed but would potentially be offering some restraining pressure on wage growth and therefore on cost dynamics and therefore on inflation.
When we look at that, we do not see—this is a question that is asked as part of the Labour Force Survey—an unusually high level of people who say, “I am not participating in the labour market now but I would accept a job if one became available”. That is consistent with the idea that many of the people who are no longer participating are no longer participating, shall we say, for more lasting reasons, including the health reasons that we just discussed.
Q95 The Chair: Just to build on that, and maybe this is more a layman’s question for policymakers, I am thinking about how we use the term and definition of unemployment. There is obviously the unemployed as you have just been describing it, and then there is a group of working-age people who for some reason or other are not working, maybe for very valid health reasons or other reasons, but are still potentially, as you say, marginally attached. Do you think therefore that the definition of unemployment is something that we are using to camouflage this big economic and social issue of inactivity?
Huw Pill: I hesitate to use the word “camouflage” because that perhaps suggests a bit more consciousness about how these statistics are constructed.
The Chair: I would agree on that.
Huw Pill: In a previous life, I looked not only at UK labour market statistics but at labour market statistics for other jurisdictions. The US has a very rich set of labour market statistics. There was a big debate about what mattered for pressure on resources in the labour market and ultimately for wage growth and inflation, and whether it was U-3 unemployment or U-6 unemployment. The fact that you have a three and a six suggests, and it is true, that there are at least six different measures of unemployment. Which one is most relevant will depend a bit on circumstances and may vary through time. The hierarchy of measures of unemployment going from one to six had exactly the character of what you are describing.
When we talk about unemployment and what we are trying to measure in the Labour Force Survey, it is the International Labour Organization standardised global definition of what unemployment means, which is people who have jobs and people who do not have jobs but are actively seeking a job. The notion of “actively seeking” is that they have tried to look for a job in the last two weeks, as I recall. That would be consistent with U-3 in American language. The U-6 measure of unemployment is a broader one and would include what I described earlier as those who are marginally attached to the labour market.
In our own analysis, to the extent that we can rely on the data—a deeper issue here is that we have problems relying on the data—the people who are inactive are asked, “Would you take a job if one became available?”. We can construct these measures and our analysis encompasses those things. It is important to emphasise that at least at the Bank we are looking across these measures and trying to draw information from them all.
The Chair: Before I turn to Lord Turnbull, I have another question to pick up on what you said. I think that you said—correct me if I am wrong—that in the Resolution Foundation’s work that was published last week its findings are broadly consistent with what the Bank and Treasury have been finding. Is that right?
Huw Pill: I do not want to comment for the Treasury. I guess you should ask the Treasury.
The Chair: Yes, fair enough.
Huw Pill: From the Bank’s point of view, the short answer is yes. My former colleague Ben Broadbent, the ex-deputy governor of the Bank, gave a speech almost a year ago, I think, at the London Business School. That was in the context of the ONS suspending its publication of the labour force statistics at that time. The accusation had been made by some journalists at the Bank of England and MPC press conference that we were flying blind. Ben went out of his way to say that even in the absence of these data being published, there are other sources of data. There are things we know better and there are things we know worse, and he went through that in some detail. In particular, there are alternative sources—and these are the ones used in that Resolution Foundation report—which are measures of employment, and the workforce jobs and survey measures give us an insight into that.
What Ben argued was that, despite not having the labour force statistics in the way we traditionally would, we still have a pretty good read on the growth of employment in the UK economy. You might think that if you know employment then you know unemployment by residual, but the challenge we face is that the margin between unemployment and participation, or inactivity, which is exactly the margin that we were just discussing about marginal attachment, is much more difficult to measure. Of course, compounding all this—this is one of the requests that we as the Bank put in; it was top of our list in the letter I sent to the ONS, which someone mentioned earlier; it was not published but was “freedom of information”-ed and then published, so there is a distinction here—was that we obviously have different population estimates and uncertainty about migration. One of the features of the labour force statistics as they currently stand, which we hope will be corrected next month, is that they will embody the latest population estimates for the UK. At a time when migration has been quite strong, that can have quite a big effect.
The Chair: Just to be clear, I think the Resolution Foundation estimates that the growth in employment has been underestimated by 930,000 workers since 2019 and, secondly, that the true rate of inactivity is considerably lower than in current estimates. Did both of those statements ring true to you, therefore?
Huw Pill: The analysis we have done is similar. We would see those perhaps hinting towards upper bands of a range. To get to those numbers or the direction that is leading to, you have to assume upper bands on the level of employment and upper bands on the employment rate. We would say that probably somewhere between the published data and those data lies the truth—the truth, of course, is a slightly loaded term—and that is based not just on the sources there; there are other sources too.
To go through the list of things we have done, we have made our own population estimates in anticipation of what the ONS will do when it reweights the labour force statistics on the basis of the latest population estimates, which they will do in December. That raises the employment rate for us, marginally but in that direction. We have also looked at other sources of data. One source of data is the Understanding Society survey. This is a completely independent survey run through the University of Essex. A question in that survey essentially asks, “Do you have a job? Are you in the labour market or are you not?” It is effectively a participation, employment and unemployment question. Unfortunately, because of the pandemic, that survey was not run in 2020 or 2021—there were many problems in life that we had to face. That survey would also suggest that employment is higher and participation rates are higher than is in the LFS series.
If you then add in the data that is more focused on in the Resolution Foundation analysis, which is the data coming from the workforce job survey and the HMRC real-time indicator, which is essentially tax data, the workforce job survey would definitely say there are more jobs. One caveat you have to put in that place is that the workforce job survey measures the jobs, not the people. If you have people with more than one job, you will not be capturing participation correctly. Crucially, if you have a reason to think, which is not implausible, that during the pandemic and its immediate aftermath there was more reason to have more than one job, the workforce job survey would tend to overestimate the participation and employment.
The Chair: Does the data give you, in the work you have been doing, a number or a rough estimate of how many people are marginally attached who are within the inactive workforce?
Huw Pill: I cannot remember that off my head. I can send you that afterwards.
The Chair: That would be very helpful.
Huw Pill: I think it is important to say that that number is not higher than was the case pre-pandemic. That has been quite a stable number, and that is quite key.
The Chair: Just to be clear, are you saying the marginally attached number has been quite stable?
Huw Pill: Yes. If we try to understand why inactivity has risen, the main reason why inactivity has risen in the reported data is that people are reporting higher long-term sickness, not that people are saying, “I am just not looking for a job this week but I would take a job if one came up”.
I also want to mention that the HMRC data, the tax data, in principle runs into the pros and cons of the administrative data that we discussed earlier. You need to keep the economic issue in mind, from our point of view. There is an issue there around self-employment. Again, during the course of the pandemic, moving into self-employment may have been more or less attractive for different people, and that is quite a hard thing to keep track of.
In the analysis that we have done, we have used the same sources. We have had discussions. In the group of people at the working level, as they say, looking at these issues across ONS, HMT, Bank of England, OBR, IFS and Resolution Foundation, there is a very intense discussion and a lot of sharing of information. Of course, the interpretation of the results then comes in different ways. Given our responsibilities, perhaps we are slightly more cautious.
Lord Layard: Following on, the HMRC data presumably would allow you to capture someone who has more than one job.
Huw Pill: Yes.
Lord Layard: Maybe it even allows you to calculate how many jobs there are. How has that trend compared with the workforce job survey trend?
Huw Pill: That HMRC data gives you the highest level of employment. If you do the ranking, the labour force measure of the level of employment is lowest, and then you go to the workforce jobs, which is higher, and then you go to HMRC, which is higher again. The difficulty in using the HMRC data is the treatment of self-employment.
The Chair: Lord Turnbull has been waiting very patiently beside me.
Q96 Lord Turnbull: I come back to this question of causation. You said that higher long-term sickness is a cause of inactivity; I would call this a medical explanation. Suddenly people are becoming sick and then they cannot work so much. It seems to me that there is equally another explanation, which is that they are becoming sick because they are becoming inactive: the stress, particularly for younger people, in getting a job and holding on to that job is the thing that is dragging them down. Whether they are sick depends on how they present themselves, how they describe themselves and how they frame the applications that they are making to DWP for benefits of various kinds. I do not think we should look at this thing as an external variable that we cannot do much about. I suspect it is very much interactive and forms a loop with the state of the labour market. If we go for a more sociological explanation, I think that what Duncan Smith on the radio was saying was causing distress rather than sickness is really what lies at the back of this.
Huw Pill: You are stretching my expertise or knowledge, probably beyond their limits. The point you make is potentially very profound. We have not done work on that at the Bank. The suggestion you make that there may be causality in both directions requires us to think quite hard about how to investigate those two effects and come to a deeper understanding.
I will make two points. The first point I make, going back, is that you are 100% right that what matters here is our understanding of the economic forces. If your description of what is happening is correct, an unattached worker who has gone through the dynamic you are describing is less likely to imply some downward pressure on wage and cost and therefore price developments. From our point of view, a worker in that situation who has left the labour market would be quite different from someone who has left the labour market for other reasons. It is important that we understand that, and we have work to do.
Secondly, if we look at the demographics of the people who report long-term sickness as a reason for inactivity, what we saw in the initial phases of the pandemic is that it tended to be older people. There was a discussion about whether this is long Covid and whether long Covid would disappear. Then we shifted into a discussion that maybe this is to do with chronic health issues that were not being treated because hospitals were being overrun, which tended also to be an older demographic. I think it is true, and this would be very consistent with what you are saying—it is not a demonstration of what you are saying, but it would be consistent—that we are seeing more longer-term health issues in younger generations, and a lot of that is associated with mental health issues.
Lord Turnbull: It also strikes me as strange that we define the labour force from 16 to 64, yet NEETs are supposed to be in education or training. Should we not start to recast these statistics so that it is basically 19 to 67? At the top end, I think you were indicating that we would have more people who are sick, who have difficulty working or who do not have the physical strength to do the work that they used to do. At the moment we just seem to be stuck. Although the pension age has changed, and changed markedly for women, we have not changed the way we collect the statistics.
Huw Pill: Those are all good points. Consistency through time is quite important for statistics too, as is international comparability. Both of those things are driving where we are.
I can reassure you—perhaps reassure is too strong a word, but I can tell you—that the issues you are raising are issues that we look at quite closely. As I was saying earlier, when we look at participation by each cohort, or each bucket as I was describing it earlier, it is interesting to see that the 65 and above bucket is one of relatively few buckets where we have seen an increase in participation during and since the pandemic. It is the only cohort where we have seen an increase in participation for both men and women. It is not rocket science, I guess, based on what you said, but that is because we have seen an increase in the state retirement age.
Those things are being incorporated in our analysis, because for our analysis it is precisely what you say: how much is this demographic dynamic and participation dynamic by demographic weighing on the pressures in the labour market, wage dynamics, cost dynamics and ultimately inflation?
Q97 Lord Rooker: Looking at it from the employers’ point of view, why should the employers be queuing up to employ people who look as though they have opted out of the labour market until something comes along but are not looking? Why should they seek to employ when there has clearly been a major change in the use of mental health issues in terms of long-term sickness? The employers will take these things into account, surely. Therefore, we could end up with a cohort of people who might never work again because the employers will pick and choose, and rightly so. The employers’ actions will surely change as a result of this. Is that taken into account at all?
Huw Pill: I think it is taken into account in our analysis. The behaviours you are talking about are ones that we do try to explore, not least when we talk to our agents, who literally go and talk to businesses. Indeed, I accompany them on occasion. I was in east Lancs a few weeks ago and south Wales last week. These are the types of questions we ask. Are you having recruitment difficulties? How do changes in regulation in the legal framework—national insurance, the minimum wage, the national living wage and so forth—affect your choices on exactly these dimensions? Often unprompted we will hear the types of issues you are discussing. Certainly, that is something that we are very cognisant of.
When we are making this assessment about how much slack, if you like, or how much useable labour, if we want to call it that—that is slightly loaded—the pressure on wages, costs and ultimately inflation stemming from the labour market, these issues are ones that really are at the front of that. Inevitably, we have to include that in our analysis, given our mandate, our job and our pursuit of the 2% inflation target.
It is a slightly different question to ask how we should measure these things. Whether you are participating in the labour market is a choice of the individual. It is a separate question from, “If you are participating, will someone else give you a job?”. It is probably useful from a statistical point of view to keep that distinct, but the issues you are discussing are definitely ones that we have to think about.
Q98 Lord Davies of Brixton: You made an interesting point about the over-65s. I have occasionally seen people talking about labour market statistics where rather than, for example, the 80% target—the Secretary of State mentioned it again in her statement today—they talk about a 60% target, including everyone over 65. Is that a meaningful way of thinking about it, do you think? I have seen people using it.
Huw Pill: I have already strayed further than perhaps I should have from our mandate, which is 2% inflation, so setting targets for other policymakers is probably not something I should be engaged in.
Lord Davies of Brixton: Do you think that it is a valid target, a way of thinking about things?
Huw Pill: I go back to Lord Turnbull’s comments earlier. A lot of this discussion is a reflection of the fact that the labour market is changing. Behind that, the demographics are changing. We are living in an ageing society, and we cannot be a productive society as we age if there is a cut-off point at 64 when the state retirement age is 66 and so forth. I do not want to sound prejudiced, but I do not think that many people in this room are in the prime age group as would have been seen in the past. Understanding the labour force in the whole, realistically, is something that we probably need to think about in public policy for labour markets. We think about the implications of that for monetary policy.
Q99 The Chair: Before I turn to Lord Londesborough, can I ask you a question about your forecasts for migration? Some argue that we have become overreliant on migration, especially for filling low-skilled, low-wage jobs. If the immigration system was changed quite dramatically to control migration, as some argue, obviously that would have an impact. I am interested in your views on what impact that would have on the issue of the labour market.
Huw Pill: From the perspective of an MPC member with the mandate we have, the 2% inflation target, and understanding pressures in the labour market, in some sense everything claps back to the demand for labour and the supply of labour. I know that sounds very trivial, but it is true. One source of supply of labour is migration from overseas. There is an issue around whether we see a segmentation of the labour market so that there are certain segments where we do not have people here who are willing or able to do certain jobs. If we do not allow people to come from outside, things will not be done and/or the wages of those who are able to do them will go up fast. From the point of view of the monetary policy of the Bank of England, the latter is something that imparts inflationary momentum to the economy. From the point of view of the overall economic performance of the UK, which also matters for the MPC, the lack of anyone to do that job will be a constraint.
The Chair: What is the data suggesting to you as you look ahead on that point?
Huw Pill: On which point, sorry?
The Chair: On the willingness of people within the workforce to do the low-skilled jobs.
Huw Pill: I think that is the question about labour force participation. We now feel we can explain the dynamics of labour force participation largely on account of these demographic trends, which are unfavourable in that we are ageing, making assumptions that are, if you like, post-pandemic assumptions about participation by cohort group: higher participation for the 65s and over, for the reasons we have discussed, but essentially stagnant or slightly lower participation for other groups, other than prime-age women for whom working from home and better provision of childcare have maybe allowed higher participation.
There is a set of issues that we judge. Given all that, we think this has now reached a point where the level of labour force participation we are seeing is broadly in line with what we think a natural level of labour force participation would be. Is that higher or lower? It is certainly lower than what the Government’s announced target has been, so it would require policy measures to improve that. Would that be a good thing? It would be a good thing in terms of making more productive human resources available in the UK economy, but also from a more narrow MPC perspective it would ease the constraints in the labour market in a way that would allow us to be somewhat less concerned about labour cost pressures and their inflationary impact.
The Chair: I will bring Lord Londesborough in for a final question but, Lord Layard, do you have a question?
Q100 Lord Layard: Just going back to what you said about demand and supply, presumably the migrants who get employed also spend their money, so they contribute to the demand.
Huw Pill: Yes.
Lord Layard: There is also that whole chunk of migrants who are students, who are big contributions to demand, alongside with export. I would have thought that, on the whole, immigration raises the excess demand in an economy.
Huw Pill: I think that is true. I would distinguish two things. I am answering several questions. I think that the pressures specifically in the labour market can be eased by workers coming in. I agree with you, though—and this is the underlying assumption of what we have at least over the medium term in the framework we use at the Bank of England—that migration in terms of overall pressure on resources tends to be a wash, precisely for the reasons you are describing. The point you make, which is well made, is about the composition of migration. Is it worker intensive? Is it family intensive? Is it non-working family member intensive? Is it student intensive? We look at these things very closely, precisely because they have the potential different effects you outlined.
The Chair: I am sorry to cut you off. I am very conscious of time. We have one big question left from Lord Londesborough.
Q101 Lord Londesborough: It is a little bit cheeky, but before we hit the big question—and we promised not to get into the Get Britain Working White Paper—there is a very interesting stat in relation to the state of the labour market, which the media were pre-briefed on, coming from the Prime Minister, saying that 9 million adults in this country lack the skills needed to get into work. That is an interesting stat, because we understand that overall economic inactivity is 9.4 million. Is that 9 million a figure you recognise? How do you factor in the skills gap, coming back to your point about the participation gap, and indeed measure the lack of skills? I am not sure where this 9 million figure has come from. It is the first time I had heard of it.
Huw Pill: I have not read that report so I do not want to comment on whether that is a number I recognise, because I do not really know where that number is coming from.
Lord Londesborough: It is coming back to the skills gap itself.
Huw Pill: The point about the skills gap is very relevant to us. A lot of what I was discussing earlier takes the labour market as a single integrated market that is fungible. That probably does not really accord with most people’s personal experience of the labour market or the economically relevant aspects of the labour market—that was true of Lord Rooker’s question—but more generally it is true.
We do have analyses. We have done quite a lot of trying to understand whether there are mismatches in the labour market. Those may be geographic mismatches, but increasingly we would see them as skills mismatches. Again, when I go around and I talk to firms, as I do through our agency network, and when we talk about whether the labour market is tight, the message we get from our agency contacts is that the labour market overall has eased. It is still not loose, but recruitment difficulties in general have eased. Certain professions and certain skills are difficult to find. Those may be IT experts, accountants or welders. These are places where in some cases we are reliant on immigrants. From the perspective of understanding that segmentation and skills mismatches, given the structure of the UK economy, that is something we are very cognisant of.
Q102 Lord Londesborough: Thank you. I will come to the main question now. Can we get your views on the impact that long-term sickness has on the economy and the likely cost of failure to remedy the problem? Specifically, we are looking at three buckets. The first is the benefits bill itself, the second is the lost output and productivity, and the third is the loss in tax revenue. How do you look at those three areas?
Huw Pill: In a way, this is back to where I started. At a high level, for people who are not participating in the labour market and therefore to a large extent are not in the productive part of the UK economy because of long-term sickness, there is potentially a big cost to them as individuals. That was the point Lord Turnbull made and, for sure, it is true. But from a more macroeconomic point of view, it is reducing the potential productive capacity of the UK economy. Those are people who could be producing things but, because of their absence from the labour market, they are not producing things. When we look at the long-term potential growth of the UK economy, this is detrimental to that.
That has an impact on the state of the public finances. To the extent that those people are receiving benefits, they are not paying taxes and they are not being productive, that would naturally increase the numerator of the debt ratio—so we might see bigger deficits and ultimately bigger debts as a result—but it will also constrain the denominator of the debt ratio, namely the overall size of the UK economy as measured by gross domestic product or something like that. Of course, those are all things that are challenging for the performance of the UK economy in general.
If it is okay, I will leave it at that in terms of overall perspective, but I am a member of the Monetary Policy Committee. I work at the Bank of England and we have a specific mandate to maintain price stability, the 2% inflation target. It is crucial that the economy functions and that we have sustainable public finances, and this is certainly influencing those things, although hopefully we have mitigations and measures in place to ensure that we do not run into problematic situations. The tenor of this whole discussion, at the cyclical level where we are focused, is about understanding where there are cost pressures being generated through higher wages because there is a tightness in the labour market because the supply of labour is impaired through the lack of participation. That requires us to monitor, for sure, and, if necessary, if those cost pressures are such that domestic inflationary pressures become stronger than is consistent with our 2% target, act. It is something we need to be very much abreast of but, to the extent that it is constraining the potential output of the UK economy, it is also something that we will have to factor in. Ultimately, it leads us to have to run a tighter monetary policy than we would otherwise have to do.
Lord Londesborough: Finally, do you feel the Government’s target, which is to achieve an 80% employment rate from the current 75%, is a credible target? What period of time do you think would be required to achieve it?
Huw Pill: I do not know whether this is a “get out of jail free” card, but I am reluctant to set targets for other policymakers or comment on the targets of others. We focus on our own target.
What I can observe is that the pre-pandemic behaviour of labour participation between 2014 and 2019 has an upward, favourable trend in it. If you extrapolate that, it will reach 80% in the foreseeable future. That period between 2014 and 2019 was a period when, as I said earlier, despite the negative demographic trends, meaning the ageing population, the participation for most buckets was rising. In particular, female participation from 60 to 64 was rising because of the rise in the retirement age for women. If you think 80% is attainable because you could extrapolate that trend, that is a reasonable thing to think.
Since the start of the pandemic we have seen a reversal in that trend, for reasons I have also discussed. There is a very broad set of policies about retirement age, the structure of the welfare system, labour market regulation and, of course, the broader dynamics of the economy that will influence that. I think you asked whether it was a credible target. I am not sure what credible means in that context.
Lord Londesborough: Achievable.
Huw Pill: Most things are achievable if you try hard enough, I guess.
Q103 The Chair: In the two minutes we have remaining, can I clarify one point arising from the Bank’s own internal analysis of the data points? I just want to be absolutely clear. The Resolution Foundation says that “the true rate of inactivity is considerably lower than in current estimates”. I think you said you agree with that. We do not need to get into numbers, but the first question is: do you agree with that, yes or no? Secondly, it goes on to say, “A rise in health-related benefit claims and long-term sickness can be consistent with little to no rise in the overall inactivity rate”, and what we are seeing within the inactivity rate in terms of people claiming that they are inactive for long-term sickness. Is that what your data is also showing when you pull together all this data that you have access to?
Huw Pill: It would be fair to say that the answer to your first question is directionally yes, across the broad set of information we look at. But that is a directional answer; it is not an answer that says the level of the participation rate or that we think we are back to where we started and so on. We would be a little bit more cautious than the Resolution Foundation’s upper estimates.
On your second point, of course, this latter question follows partly from the first question. One of the reasons for that caution is that we would not fully embrace that. There are reasons, which we discussed, why the reporting in the Labour Force Survey perhaps overstates the role of long-term sickness, and the use of administrative data gives a different read on that. Again, directionally it is not a different read but in magnitude it is a somewhat different read. That is probably how we would see it.
The Chair: On the employment rate, if there is this gap, that would imply that productivity growth, as the Resolution Foundation also states, has been even weaker than thought.
Huw Pill: I think that by a question of arithmetic, that is true.
The Chair: By definition. Just to end on a very gloomy note, these statistics would suggest that things are worse in many ways.
Huw Pill: The level of output or gross domestic product of the UK is a number that also gets revised. Indeed, it gets revised substantially very frequently and is partially based on assessments of population. Population dynamics are things that are also influencing all these numbers in ways that we need to think about. Yes, obviously, if you take the output as given and you say there are more people working, productivity is lower.
The Chair: Yes, quite. I just wanted to make sure everyone was getting that on the record. We have done that in record-breaking time. Thank you very much indeed, Huw Pill, for coming in.
