final logo red (RGB)


Select Committee on Economic Affairs

Corrected oral evidence: Employment and Covid19

Tuesday 10 November 2020

3 pm


Watch the meeting:

Members present: Lord Forsyth of Drumlean (The Chair); Baroness Bowles of Berkhamsted; Lord Burns; Viscount Chandos; Lord Fox; Baroness Kingsmill; Lord Livingstone of Parkhead; Lord Monks; Lord Skidelsky; Lord Stern of Brentford; Lord Tugendhat.

Evidence Session No. 18              Virtual Proceeding              Questions 149 - 158



I: Fabian Wallace-Stephens, Senior Researcher, Royal Society for the Encouragement of Arts, Manufactures and Commerce (RSA); Tera Allas CBE, Director of Research and Economics, McKinsey and Company.



  1. This is an uncorrected transcript of evidence taken in public and webcast on
  2. Any public use of, or reference to, the contents should make clear that neither Members nor witnesses have had the opportunity to correct the record. If in doubt as to the propriety of using the transcript, please contact the Clerk of the Committee.
  3. Members and witnesses are asked to send corrections to the Clerk of the Committee within 14 days of receipt.




Examination of witnesses

Fabian Wallace-Stephens and Tera Allas.

Q149       The Chair: Welcome to the Economic Affairs Committee and to our witnesses for our first session, Fabian Wallace-Stephens, who is the senior researcher at the RSA, and Tera Allas, who is the director of research and economics with McKinsey and Company.

I will begin by asking the first question, which is not an easy one. How should the Government be judging the best times to withdraw employment support measures so that viable jobs are protected without unduly slowing labour market reallocation when vacancies rebound?

Tera Allas: Thank you for the invitation to join today. It is indeed not a very straightforward question. In general, of course, government needs to consider costs and benefits of these kinds of schemes. There are many on both sides of that equation.

The first thing I would note is that the job support scheme has two primary benefits. We must not lose sight of the second one. The first one is that we maintain jobs and reduce the degree to which there might be labour market scarring because people lose their jobs and perhaps stay out of the labour market for a long time.

The second one is simply to support people’s incomes so that they can continue to spend and live their lives as normally as possible and essentially continue to fuel the demand side of the economy as far as possible. Those two might have quite different timelines. It might also not just be about timing. It might well be about the design of the actual policy.

If I take a step back from those two key objectives, from a labour market perspective it is not viable to lock people into the same jobs for ever. In other words, we probably need to make a big distinction between people and the jobs they are in, and the actual employer and the business they are in. If you are a sales retail assistant, for example, your job could well be entirely viable in the long term, or even in the medium and short term as soon as we get out of lockdown, but the particular employer you are working for may not be somebody you want to for ever spend your time with. We would want to maintain that kind of flexibility in the medium term. As soon as the economy is back on its feet and people can start looking for new jobs, that would be beneficial.

My final point is that that might mean that, when designing these schemes, we pay more attention to the individual, their skills and their ability to look for work than necessarily the firms that are employing them.

Fabian Wallace-Stephens: We have argued that the furlough scheme should be seen as a kind of backstop to prevent mass unemployment. There are also issues with the question. The most appropriate lever to aid labour market reallocation here is not the withdrawal of the furlough scheme. Rather, the Government needs to develop a more robust, active labour market policy alongside furlough support.

This could take the form of an end-to-end transition service or, as we have called it, a job security centre, which could be modelled on Sweden’s job security councils. This would aim to redeploy workers that are most at risk into sectors that are more resilient, by providing information about local labour markets as well as access to career coaching and training opportunities, supporting these workers up to the point of placing them in a new job.

Our analysis has recommended that this be targeted at workers that are at risk of both Covid-19 and automation, and should include more generous financial support than universal credit. Critically, by providing that more generous financial support, you can encourage and empower workers in jobs that are unlikely to be viable in the long-term to find new and better-paid jobs.

The Chair: Do you think that the Government were correct to extend the Coronavirus Job Retention Scheme until March? Will that not create an unnecessary deadweight cost?

Tera Allas: I am not in a position to criticise or otherwise what the Government have done. This is a completely unprecedented situation. We have never seen this kind of recession and short-term shock before, at least not as far as the stats show going back 300 years or so. It is not possible and it is not even helpful to try to optimise policy in those circumstances.

It is probably more sensible to think about worst-case scenarios and mitigate as far as it is relatively affordable. We can get on to the discussion about what is and is not affordable. That is another incredibly challenging set of questions. As Fabian has outlined, perhaps now is not the time to think just about the existing policy instruments but to think creatively about how to build back better.

The Chair: As there is a finite amount of money available, are you not concerned about the deadweight cost?

Tera Allas: I am always concerned about the deadweight cost. I am an economist after all, and I worked for the UK Government for a long time, but I would be more concerned about the longer-term consequences of not having a dynamic, productive and flexible economy that is competitive on the global stage. There are many large megatrends happening globally that are a much bigger risk to the UK right at the moment than some of the deadweight cost and, indeed, the short-term financial issues to do with supporting workers and businesses.

The Chair: It is a price worth paying.

Fabian Wallace-Stephens: Our initial criticism about the job support scheme was that, although we agreed with the notion of viable jobs and we thought that was an important criterion for protecting jobs, not enough consideration was being given to the long-term viability of jobs in sectors, particularly in sectors such as the creative arts and entertainment, which are clearly quite resilient to technological change and often provide good, meaningful work.

Although the return to the original furlough scheme addresses this challenge, it goes too far in the other direction. That is why it will not be very popular with economists, because it now completely eschews that notion of viability. As mentioned earlier, we see this as only part of the solution. If there are other effective policy instruments, perhaps you can mitigate those deadweight costs by getting ahead on the reskilling challenge.

Q150       Lord Tugendhat: I would like to ask about the trends that have been accentuated by the pandemic. The pandemic may well have launched some new trends, but it has also accelerated trends that existed before. I wonder what our two panellists think about that and whether they feel it is possible to assess which jobs are more or less viable as a result of the trends that have been accelerated by the pandemic.

Tera Allas: It is also really important to recognise that there have been pre-existing trends that mean that, had they just continued as was, there would also have been incredibly important structural changes. I will come on to what that means for the viability of different jobs, as Fabian has already implied.

The trends that, according to our research, have been particularly accentuated and accelerated are actually relatively pragmatic, within-business changes in the adoption of technology. The most important one is improving and changing your digital interface with your customers so that you can sell things online and interact with your customers online. The second most important thing has been creating the supply chains to be able to deliver those. There have been many more basic things like cybersecurity, putting your software on the cloud or remote working. The remote delivery of goods, services and labour has been the key trend that has been accelerated.

I would warn not to jump to conclusions about automation. Technically, automation should be accelerated, because if people are becoming more expensive in many ways or it is harder to employ them in their normal way of work, the alternative is to use machines to do that. Because of the uncertainty created by Covid, a lot of businesses are finding it difficult to justify those kinds of large investments. Change management is also much harder; change management is critical to getting any value out of automation, and businesses are having to deal with a different labour force for the moment. Global value chains will not be that substantially impacted, other than what they were already doing, which is shifting to more localised and regionalised activity.

However, the main impact of digitalisation and automation on the labour market is polarisation. We all talk about working from home and whether people will continue to do it, but only about 40% of all the jobs in the UK can even theoretically be done from home. Clearly knowledge workers can, but if you are bus driver, a dental nurse or a retail assistant, you have to be at your place of work to be able to produce the output. That polarisation means, first, that the people who are in digitally deliverable jobs will have more flexibility and will probably have been at less of a risk of losing their jobs, even though there, too, there will be differences.

In the longer term, the automation trends will increasingly eat up work that is either manually or cognitively relatively simple, which machines can basically do. That will continue that polarisation. Previous work that we have done on the future of work suggests that, for the top quintile of jobs, we might be looking at an increase in demand of 20% for healthcare workers, IT professionals and all kinds of professional managers and workers. At the bottom end of the scale, we might be looking at a negative 20% growth rate between now and 2020 for things that are fairly automatable: factory work, perhaps some aspects of kitchen work, lots of warehouse-type work and so on.

Implicit in my answer is that we think that, up to a point, it is plausible to predict or at least to forecast in some manner the key drivers for replacing human work with machines. It is not super-easy to know when exactly businesses will invest in that automation. We know that in the UK, for example, businesses have, on average, invested in a lot less automation than in other countries. How speedy that adoption curve will be, both in the UK and globally, is a big uncertainty.

Having said all that, there are some no-regrets things that one can do in labour market policy. One can already see some shortages of skills that are already required and that will be required more in the future. Basic digital skills, for example, fall into that category. Essentially everybody—20 million people by 2030—will not have the digital skills that they need. Some element of upskilling everybody will be required anyway. Similarly, we already know there are shortages in the health sector, for example, and in some IT sector jobs. One can already lean into the areas that will be in high demand in the future and that are already in high demand.

Fabian Wallace-Stephens: The most significant impacts of technology have related to changes in underlying business models. There has been a significant growth in e-commerce, with online sales and retail peaking at 33% during the first lockdown, which was an increase, up from 20% in February. This meant that, over five or so months, we saw a similar kind of digital transformation to what we saw over the last five years.

Previous RSA research has shown that h the historical shift towards e-commerce, leads to a significant shift away from customer service roles and towards logistics and distribution jobs. In turn, this has tilted the gender profile of the industry towards male employment, thus further entrenching gender inequality in the labour market.

We have also seen a rise in businesses providing online services over Zoom, particularly in health and education but also in real estate and entertainment. The ONS has been collecting some data on this in its Business Impact of Coronavirus (COVID-19) Survey.

Our report also suggested that there are some quite compelling signals that the pandemic could accelerate the pace of automation due to the rising costs of labour in a pandemic, which Tera mentioned, but also public health considerations requiring that businesses introduce technology in a way that creates a wedge between consumers and workers. If you look at the historical trend towards automation and what sectors it has been happening in, generally there has been a trend towards sectors that are highly automatable experiencing a decline in employment growth over the last decade, for example in retail and manufacturing. There are other industries such as hospitality that, although being at high risk of automation, over the last decade have actually experienced quite strong employment growth, between 20% and 30% in terms of new jobs.

My point is that, with the pandemic, we may now be at a turning point, with public health considerations meaning that consumers no longer prefer to be served by a human in these kinds of industries. Potentially, there is now a much stronger business case for accelerated automation in some of these industries. In Hong Kong and across Asia, we are now seeing robots delivering room service in hotels, for example.

Lord Tugendhat: In view of what was said about digital skills, would either of you anticipate that there will be a weeding out of older workers in order to make room for younger workers? We are, of course, very concerned about youth unemployment but, if digital skills are a priority, one might expect that employers would want to weed out some of the older and bring in younger.

Tera Allas: That will depend on the overall labour market balance that we end up with. I am not very clear that, long term, there will actually be a surplus of labour. In the short term, there may well be, and there already is at the moment because of Covid, but once that is recovered from we are heading towards an ageing population. I also would not underestimate the degree to which our younger people do not have the digital skills either. They can look up stuff on Google on their phones, potentially, and they can check their Twitter accounts, but when it comes to actually working with digital tools, their competences are not all there.

My final point is about training. Older people can learn these things too. It would be sad if we thought that older people could not learn skills that are completely within their remit. In the UK in particular, but also in most other countries, the whole idea of lifelong learning is a lost opportunity. We talk about it, but we do not do it and we do not spend money on it, yet we know that employers would be better off if they spent the money training the people they already have rather than getting rid of them and then hiring someone else. That is in 75% of cases, not in 100%.

The Chair: This Committee is quite a good example of older people knowing technology.

Lord Tugendhat: The United States is quite a good example.

Fabian Wallace-Stephens: It is really interesting to think of concrete examples where we might be seeing this play out. Accountancy is the one that springs to my mind. The issue, for everyone, is that the idea of a career for life may no longer be realistic. So, there is need to ensure, through lifelong learning, that people of all ages can be adaptable. What it means to be an accountant now might be very different from what it meant 30 years ago. It is not a case of weeding them out, but the values of people who have been in those roles just mean that they are no longer a really good fit for what that industry requires if it moves towards all accountants being hybrid data scientists.

Perhaps there are other roles that need to be filled due to demand, such as social care workers and teacherspeople with a really rich life experience. They might become more and more attractive opportunities, particularly if there is an expectation that people might have to work for longer and supplement their retirement income with part-time working.

Q151       Lord Livingston of Parkhead: Fabian, could I pick up on a comment that you made? It is not one about the accountant, although I became an accountant 30-odd years ago. You mentioned that, on the one hand, we are seeing that the way certain services are provided will be automated. On the other hand, there have been certain underlying trends in demand for industries—you mentioned hospitality, and social care is another—that actually have been in high employment demand. Will that fundamentally change in most industries? I would be interested in the balance that you see between these two things; between it being provided more capital-intensively and the continued demand in, for instance, hospitality and social care.

Fabian Wallace-Stephens: Again, there are probably different answers for different sectors. In hospitality, changes that have been introduced during the pandemic will likely stick around and possibly lead to reduced headcount in the sector. A really good example of this is self-service apps, like the Wetherspoons app that you have on your smartphone. This technology is really cheap for businesses to adapt. Often, there are business models where they pay a tech company a fraction of their sales so they do not actually have any up-front cost. It is called automation as a service. It is possible that these things will stick around, not least because actually, in a busy pub, sometimes being able to order by an app means that you can get your drinks a lot more quickly. It can offer a more convenient experience.

Hospitality could also be impacted by the shift to remote working, which is likely to have much wider impacts on the structure of the economy.. The economist David Autor argues that telepresence is a form of automation, so this shift to online services and remote working more generally will reduce the need not only for cleaning, security, building maintenance staff as well astransportand hospitality workers that used to serve commuters.

This is exactly what we are seeing with companies such as Pret, which in August announced that it was going to close a ton of its stores, cutting almost a third of its workforce, because its sales were down due to a lack of commuter footfall. In hospitality, there is not just the automation dynamic playing out. There are the broader structural impacts of remote working to worry about.

There are other consumer trends that I think will interact with things. Take air travel, for example. We are seeing a boom in staycations, and that, coupled with growing concern about climate change, will mean that people will look to cut back on their emissions,  even once it becomes completely safe to travel on planes.

In social care, it seems like there are so many positions to fill in that industry that it is likely that we will have to move to having a hybrid workforce, because we cannot actually find enough people to do those jobs because of underlying issues such as the respect we pay these workers, the way we compensate them and the poor working conditions we allow them to experience.

Q152       Lord Stern of Brentford: The question I had in mind was about structural change. That discussion has already begun and I do not want to repeat what has been said. I would like to focus on your views on industries’ activities that are changing as a result of the crisis and the response to the crisis, rather than mostly because of long-term trends in automation. The examples that I had in mind, if you could comment, were mostly about location. You have already begun to talk about transport, but how do you think transport in cities is likely to change as a result of the Covid experience and the response to the Covid experience?

Similarly, on the point about where activities take place in the world, if you can do them remotely there is no difference between me in Sussex and my Indian cousin somewhere else, in India. Again, there is a big issue about the response to the crisis, in this case remote working.

We just started to touch on city centres. If activities stay closer to the suburbs or in the countryside, what will be the future of city centres? These are all questions about the direct effect of the crisis and the response to the crisis on the way we behave and are likely to behave. How big do you think those structural changes are likely to be?

Tera Allas: There are quite a lot of questions there. Starting with the point about how people have changed because of Covid and what of that might stick versus which behaviours will just snap back to business as usual once we are handling the virus and we do not need to be afraid of it anymore, I am sure that the remote working trend has been massively accelerated. Some of that will stick, but, as I mentioned earlier, only about 40% of people can actually work remotely, because, as you said, there are certain jobs where your Indian cousin could deliver service to the UK, but there are many others where you cannot do that.

Also, people have now had several months’ worth of experience of remote working. There is a bit of a backlash against it in many workplaces, partly because people require to get together as people. The social aspect of work is incredibly important for people’s motivation, for their work satisfaction and therefore their productivity in the medium to long term. It is also incredibly important for their health and mental health. We have not started seeing anything yet when it comes to the kind of burden that this isolation-type working has imposed.

In the future, the businesspeople we speak to very much expect that we will retain some of the best aspects of it, such as flexibility and being able to tailor things to individual workers, but we will also go back to the office quite a lot. In broad macro terms, for city centres that might mean a little less congestion and hopefully the peaks and troughs being smoothed out a little more, so that we can use the existing capacity a little more effectively. As Fabian mentioned, some of those jobs and some of the hospitality and entertainment activities—you name it—will perhaps be spread out a little more into the suburbs. Just off the top of my head, I would be quite surprised if the allure and the benefits of agglomeration that we get in cities, especially with knowledge work, were not still quite an important driver in the future.

That is in relation to the UK geography. In terms of international competition, services have already been delivered across borders for a very long time. Services are, of course, a much more important export to the UK than manufacturing is. In net terms, we have a surplus in service exports and imports but not in manufacturing. Services have been growing two to four times as fast as manufacturing trade globally, so that is a pre-existing trend and I have no reason to believe that that would not also have been accelerated quite significantly as a result of Covid. There are also reasons why people will also continue to require local services, such as language barriers and that human factor.

That pre-existing trend of demand shifting more towards services in all economies, as well as being more and more exportable, will continue to play out. Compared to the domestic economies, however, it will probably still not be a big enough force to completely shift the structure.

Fabian Wallace-Stephens: On the economies of place and how we are seeing that play out in the pandemic, there is some quite interesting data from Google. Google mobility data is released almost daily. It has shown that retail and recreation footfall have experienced a particularly sharp decline in London, as well as in other urban centres. We are clearly seeing a doughnut effect playing out. This may be good for suburban high streets, but it is unlikely that we will get like-for-like replacement for demand. We have already seen over the pandemic that supermarkets are some of the biggest winners. Some of the money that you used to spend in Pret is now spent on making your own packed lunch rather than buying a sandwich from the shop down your road. That will obviously have implications for the wider employment composition going forward.

That same Google data shows that there has been an increase in lots of rural tourist hotspots. Places like Cornwall have seen their retail footfall increase. At least, that was the case when I looked at the data in early October before we went into lockdown.

I completely agree with Tera that the future of flexible remote working is likely to be a hybrid model, perhaps with people going into the office for two days a week. That could temper this trend somewhat.

An important point to note about how this fits into the vision is that more people will be able to move to rural areas of the country. That could have a positive impact on local economies. But there is the important issue that people in lots of rural areas in Wales and the south-west, in particular, do not have access to very fast broadband speeds, which could prevent effective remote working hubs.

I have not looked too much into exactly how we expect this to work and international competition, but there has already been a huge trend, particularly with platforms like Upwork, which is delivering highly skilled coding projects for which there is a shortage of people who can do them in the UK. That is one reason why firms might be turning to these platforms; it is not just that they are cheaper. A lot of the work that gets offshored is also tasks that people in the UK definitely would not be that keen on doing, such as call centre work.

Q153       Lord Skidelsky: So far, our witnesses have been giving us projections of the demand for different kinds of services and the effects of these on employment in those sectors. I would like to switch to government policy in relation to this kind of data and projection. The RSA’s report on automation recommended a policy platform consisting of targeted support, transition services and upskilling, but how does that compare to the Government’s current and planned employment skills and retraining policies? In particular, witnesses have complained about the fragmented nature of government services. There are just too many organisations overlapping with each other and not a very clear allocation of responsibilities in government programmes.

Fabian Wallace-Stephens: We talked a little bit about the furlough scheme and my views on that. On targeted support, in the report we were calling for an iteration of the old job support scheme but, as I mentioned earlier, our concerns have largely been addressed through the return to an original furlough.

On the transition services that we recommend, there is definitely no current alternative which the Government have recommended. Ultimately, the transition services that we are suggesting would build on the current Jobcentre Plus infrastructure, but in the long term this would definitely be much more of a universal service, something that all workers would gain access to and benefit from, rather than those who arguably find it most difficult to find work and who go through these low-pay, no-pay cycles. We think that the long-term vision will need to be more of a universal service, just because, as with some of the other examples we have talked about, such as the accountant who may lose their job to technology, people will be cycling through more and more jobs throughout their career.

Critically, it would involve an early intervention system, which means that there could also be changes to employment law to increase the notice period that workers have when they are made redundant, in order to bring it a bit more in line with Sweden, particularly for workers who have been in their jobs for a long time. The notice periods for the UK are really low. It only increases by a couple of weeks for every year you have been in your position.

Critically, as you were saying, we think that this kind of system will need to be both developed and delivered with a range of partners, including industry, unions and local actors. Some of the success of similar schemes in Sweden is down to the fact that they have a very different institutional landscape, where labour markets are governed by a tripartite model of industry, unions and government. It will not be easy to replicate that in the UK, given our history of collective bargaining, but there are ways in which we might be able at least to bring more partners along that journey and for it to be less something that is solely delivered by government.

There might also be a role for local economic partnerships. One of the really interesting initiatives that I have seen emerge in response to the pandemic is the regional redeployment service that was developed by Buckinghamshire, Oxfordshire and some of the LEPs in the south-west, which looked to rapidly reskill workers who were at risk into some of the growing occupations, such as delivery workers and social care workers.

Critically, we think that, in order to have this kind of effective, active labour market policy, you need to have a much more generous welfare system in place. That is why we have called for something akin to a transitional basic income, which would provide people who need to retrain with much more of an incentive, more so than the current system of universal credit provides.

Alongside this retraining programme, we also recommend policies on upskilling. A critical gap in the Government’s analysis is that they have been too focused on short-term viability, on Covid, and they have not really thought enough about the long-term viability and the risk of technological disruption due to automation. Interestingly, in our analysis we identified a group of workers who are at low risk from the pandemic but at quite a high risk of automation. This includes some key workers, such as supermarket workers and delivery drivers.

The pandemic has highlighted the importance of these roles.But too often, they are poorly paid and offer little scope for progression. They are also at risk of automation. To ensure that they are resilient and not replaced  with machines in the next five years, we really need to upgrade these occupations and give workers more opportunities to develop their skills in them. A lot of the research on automation shows that the more highly skilled workers are, even if they are in the same job, the more resilient they will be to technology and the more chances there are that their skills will allow them to complement new technologies rather than their having to compete against them.

Tera Allas: In addition to what we have already heard, I would comment on the fragmentation point. There is a fundamental misconception everywhere, including in the UK, about how you should go about designing policy. If you design policy top down from ivory towers, it will never work on the ground. You need to shift to a customer-centric delivery model, which is why some of these local initiatives are more effective, because you can look at it from the point of view of the worker and not pretend that you know what makes them go find that job or go get that training. We need to work alongside them to figure out what makes it easy for them, and then design a system along those lines.

Another example that has not been mentioned yet is Germany, where its job centres are shifting to becoming skill centres, very much along the lines that Fabian suggested. The target is not to have activity or to announce how many people came in; it is to aim for them being redeployed and employed. That requires a lot of very local knowledge and a combination of analytics and computers sifting through the data, but it very much requires a human interface too.

This work has been particularly successful in charities, because they have been able to make deals with employers to say, “You are currently requiring more people in your data sciences department. You cannot find them. We can find displaced people who have the potential to have these skills. We will employ them and reskill them, but you agree to take them on once we have done that”.

Everybody wins, because the employers themselves tend to say, “These people are as productive as somebody who has [already] been at my firm for two years”. It is about joining up the interim journey for the worker; it may have less to do with government activity and much more to do with grass-roots activity that is centralised on the person who needs help or is looking for the job.

Q154       Lord Monks: Both of you have emphasised the need to promote transition arrangements, which was very much in that RSA report on automation that Lord Skidelsky just talked about.

My question is about the IPPR proposal for a job training scheme to be attached to the extension of the job retention scheme. What do you think of that? Would that be an effective mechanism, or is it too much like the top-down mechanisms that we have just been talking about and perhaps been a bit sceptical about?

Fabian Wallace-Stephens: I cannot comment too much on the specifics of its recommendation. There are definitely some virtues of having something that is linked to the job retention scheme. Critically, when we think about these transition services, they need to have an early intervention aspect. This is one of the main explanations for why they are effective in Sweden..

However, our analysis has shown that it is important for retraining opportunities to be more targeted rather than just for workers that have been furloughed. Critically, they should be targeted in the first instance at workers whose jobs are not viable both in the short and the long term, so workers who are at risk of both Covid and automation, for example in hospitality and manufacturing. In the long-term, once this round of furlough support expires, we might have to think about different mechanisms to protect jobs that we are confident are viable in the long term, rather than ask those workers to transition into roles that might be both lower-paid and more vulnerable to technological change in the future.

Tera Allas: I agree in broad terms with a lot that Fabian has just said. Just to go back to the point about top down versus bottom up, it is more that with top down you still need to have the framework conditions in place for the right kind of activity to take place. Bottom up is more about the day-to-day delivery of outcomes and getting people back to work.

On that front, it would nothing like enough for retraining and reskilling to be attached to just the people who have been furloughed. I would have to agree with Fabian there. Those people are mostly a consequence of Covid, and even though there is some correlation between the jobs that might have been displaced by Covid and those that are more automatable, it is not a one-to-one correlation. It is probably more important to have a broader adult skills policy and redeployment policy as automation and the other global forces start to really take effect.

As I mentioned earlier, in that context, by 2030 we will need 30 million people to be reskilled. Even if we did something with the people who are currently furloughed, that would not go anywhere near far enough. We also need to remember that by far the most effective training for grown-ups is in the workplace, so somehow we need to get employers to pay more attention to what they can do to build skills. Sometimes it will be cheaper for employers to hire someone who has been furloughed and retrain them into a role than perhaps to compete in a very narrow labour market for that person. Expanding the ideas there would be helpful for the whole length of people’s working lives.

Q155       Lord Fox: Following on from what you said, how do you properly incentivise employers, or indeed make employers do this? It is quite clear that they are not, and it is quite clear that the current circumstances do not mitigate towards these kinds of things. Even in manufacturing, you have the value chain where the large companies are poaching the skilled people from the smaller companies on a pretty constant basis. How are you going to drive these behaviours?

Tera Allas: I wish I had a clean answer to that, but there are several different options to be considered. In some ways, it is for the Government to figure out what they think will work. A lot of the business associations will help them with that.

There are two critical factors here, which are the principles that we need to take into account. One is incentives. There are many people who are of the view that we do not currently have a level playing field for investing in human capital versus investing in physical capital. I am not an expert on what these things look like from the business point of view, especially for smaller businesses, but, looking at the numbers, we certainly massively underinvest in human capital in this country. We invest in people when they are young, which is great, but, beyond that, 2% of the Government’s education budget goes on people who have left full-time education. Indeed, the per-person amount of training that people get is nothing like lifelong learning. It is a shame, because we are very much now a knowledge-based, people-based, human-based society, where we want to do exactly what Fabian said, which is complement machines with people who have the right skills and then can have the productivity uplift from that.

Q156       Lord Livingston of Parkhead: Fabian, you keep referring to Sweden as a model. The Swedish unemployment rate, if I am not wrong, has been running at close to twice the UK’s. Are there any other models? The American economy seems to be pretty good at creating jobs if what we are looking at doing is creating jobs. Where else should we be looking, or am I being unfair in thinking that Sweden is the only model that you think we should follow?

Fabian Wallace-Stephens: I would not say that we should necessarily be following the Swedish model for job creation. I do not know enough about the approach there, but when you compare the Swedish employment service with the very piecemeal data we have about the effectiveness of our employment service, the Swedish employment service seems to be much more effective at retraining and redeploying workers. There is very piecemeal data about the effectiveness of Jobcentre Plus in the UK, but there is also lots of evidence that people who use this service go through this low-pay, no-pay cycle.

The Singaporean model is particularly interesting. On the fragmentation point, Singapore has a very unfragmented skills system. Part of that is personal learning accounts, which is something that we tried and failed to introduce in the 1990s in Britain, but the impact of these programmes in Singapore and France suggests that it was an implementation failure rather than that they are a bad idea per se.

One approach, which builds on the previous question, is that, rather than asking businesses to pay for all the retraining, we should rethink some of the existing funding mechanisms that we have in place, like the apprenticeship levy. Reconfigure that into something that becomes more of a general skills levy, which can be used to fund these individual learning accounts that people can use to upskill.

The critical thing that makes personal learning accounts effective, which may not have been in place when we tried to introduce them in Britain, is developing maps for the future of different sectors, understanding the skills needed and communicating that to people so that they can understand what they need to do in order to flourish in this future, more digitised economy.

Lord Livingston of Parkhead: Tera, you made a really interesting comment at the end of your last answer about the UK not supporting learning. The UK, one would argue, does not support investment in technology or in businesses, so it is not an either/or. You made the point that we need investment in technology and in people to operate it. Could you say a little more about that, because a lot of the discussion has been about an either/or, and there is a lot to be said for your comment about it being an and? Could you say a little more about anything that might bind the two things together?

Tera Allas: I am a trustee of Be the Business, which is a charity I am sure you have heard of, which is looking at ground-up, firm-level productivity improvement. In all that work, we always end up thinking ultimately that it is the business leaders and the managers who decide whether and when to adopt technology and whether and how to reskill their people. You find some incredibly innovative and inspirational business leaders, and you find others who are too busy firefighting or have not even necessarily been learning the sorts of skills themselves that would help them.

One thing that definitely brings the two together is just good management and leadership skills. I am happy to expand on how you might go about spreading those skills more widely in the UK business population. Beyond that, we also worked with Be the Business to look specifically at the tech adoption issue, on which the UK could do better. It would benefit from it and it would not necessarily—probably not at all—displace workers but would drive productivity and competitiveness. A critical aspect, again, is making the adoption journey easy. The current bigger [technology sector] players do not necessarily make the adoption journey easy for small businesses, because they are used to selling big enterprise software solutions to large corporations that have IT departments and so on.

From an ecosystem point of view, if those larger companies want to tap into the small business market, they need to change the ways in which they sell to these companies. There are also a lot of competitors, new players, that are tapping into the small business market. As long as we can make sure that they have a level playing field of competition and are encouraged to grow, that will provide those services to our smaller business population as well. Ultimately, if we improve the digital skills levels of everybody, including managers and leaders in small businesses, and that will help adoption a lot.

Q157       Baroness Kingsmill: We have covered off most of the issues to do with the way in which employment will change, not only with Covid but with the speeding up of some of the implications for employment that were already in train, such as AI and automation of various kinds.

The picture that seems to be emerging is one in which there are workers who are working with their brains, are able to work from home and are able to earn quite a lot of money, both for their companies and for themselves.

Then there are the people in the middle, who are not necessarily earning very much but are absolutely essential, such as firemen, teachers, care workers and so on, who do not have those options and who are badly paid, on the whole. We do not have a very good system; despite recognising how important they are as workers, we do not reward them particularly well.

Then there is a layer of people at the bottom who will not be working, who will be in poverty and who will be dependent on state benefits on the whole. We have heard in previous evidence about deep poverty and an increasing number of people who are moving into poverty.

I want to get some sense from you as to how we overcome the almost inevitable increase in inequality that is arising and which has great implications for our society and the future. Much of our discussion has been terribly theoretical. I would quite like some practicality as well. It is very easy to say that we must have more training, we need more skills and all those things, but the practicalities of all that would be very helpful to us when we come to write our report.

Tera Allas: It is a huge question, but I will try to give a fairly brief answer. I am not 100% sure that I fully agree with your characterisation of the different buckets into which people fall. There will be variations to that theme, but broadly speaking, as I have already said, I agree that there might well be a continued polarisation in the workforce.

The good news is very much that, in addition to tech skills and important highly educated brain skills, we will need a lot more human service, human support, empathy and social skills. I do not see that as necessarily being a question of education. Most humans have those skills inherently. The challenge becomes how we make sure that those jobs get paid enough in an environment where quite a few of them are public sector jobs and there have historically been low-paid jobs and not a lot of job security. The same kinds of interventions that have been used in the past will need to be looked at again, such as minimum wages.

Clearly, when you talk about inequality, an economist will immediately think about redistribution. Those are all incredibly politicised and political questions, but they are not difficult in the sense that they are essentially about shifting money from one person to the next. However, it is utterly inadequate to look at inequality just from the point of view of incomes, because ultimately what matters to people is their well-being and their life satisfaction. We know that incomes are a very small part of that. Income inequality is an important component of it, but health—mental health in particular—relationships, communities and all those things that are harder to put a finger on—

Baroness Kingsmill: Housing and food. also.

Tera Allas: Yes. Therefore, looking at it holistically will be incredibly important. It is not just about handing money over. We should not just think about money, because all those other things matter as well.

My final point is about something that we have not really discussed today. We did a report a couple of years ago about public sector productivity. It is our view that, with the right kind of processes, it is probably possible to improve on a lot of the other things that really matter to citizens without spending a huge amount more money. If you put those jigsaw pieces together and people, even on lower incomes, get access to the right kinds of public services, they can have a good life.

Baroness Kingsmill: The answer has not been entirely adequate, but my question is a big one, so that is fine.

Q158       Lord Fox: Briefly, are there any early indications as to how the nature of the adjustments that we have discussed may affect the Government’s levelling-up agenda? We have talked a bit about access to broadband and access to skills, but how do you think those will sum up?

Fabian Wallace-Stephens: We talked about a lot of the different themes. Potentially there are opportunities for some parts of the country. We have not done in-depth research on this, but from the data I have seen I would expect rural parts of southern England to be some of the areas that experience a boom.

Lord Fox: That is not the levelling-up agenda. The Government’s levelling-up agenda is not focusing on rural southern England, is it?

Fabian Wallace-Stephens: Yes, exactly. We have talked about some of the opportunities geographically as well as the fact that the public health situation has been much worse in the north. That may have compounded the impact of automation and the pandemic in general. There have been some interesting specific sectoral explanations for this. Some of the major outbreaks in the north and midlands have been attributed to particular manufacturing sectors in food or in clothing, so there is a challenge there.

Lord Fox: Put simply, will the post-industrial areas, which are the targets for levelling up, benefit from the sort of trends that you are talking about, or will they be threatened by them?

Tera Allas: On the post-industrial trends, if we put Brexit and Covid to one side, the work that we did ahead of either of those events suggested that those areas were not going to benefit from these trends. In other words, all other things being equal, just as we discussed in relation to the inequality between different kinds of occupations, the areas that are least prosperous in the UK and that have been growing the least in the last 20 years are also the areas that are likely to see the least growth going forward, because it is precisely those more manual, lower-skilled, manufacturing-related jobs that are likely to continue to be in even shorter supply on the jobs side. It is impossible to predict how the care economy, which we do expect to grow significantly, might create some new jobs.

It is important to note that Covid and Brexit will have very different impacts. On balance, manufacturing might get a further hit from Brexit, but, at the same time, Covid has hit hospitality and retail more. It is a messy picture. The good news, from a levelling-up point of view, is that there is already data on furloughs and on claimant count by locality. If you look at it, as I happened to yesterday, you see a very varied picture, where it is the big cities that have done really badly. That includes London, which of course is theoretically one of the most prosperous parts in the south-east of the UK, but it also includes Birmingham, Manchester and some others among our second cities. At the moment, from a furlough and a claimant-count perspective, the areas that have had the biggest hit from Covid are cities rather than necessarily the towns that were so heavily discussed as part of the levelling-up debate.

Fabian Wallace-Stephens: A good way of describing it would probably be a levelling down rather than a levelling up. Some of the most prosperous areas have been worst hit, but there may be insignificant gains or no gains to be made elsewhere.

Tera Allas: It is equally some of the least prosperous areas, such as Birmingham and Manchester. It is like a U-curve. I can send you the data if that would be of interest.

Lord Fox: That would be very helpful, thank you.

The Chair: On that note, I am afraid we have run out of time. Can I thank Tera Allas and Fabian Wallace-Stephens, our witnesses? It has been a very interesting session.