Begin again?

Recessions leave lasting impacts on our labour market. But different recessions leave different legacies: the 1980s downturn left us with stubbornly high unemployment, while the financial crisis led to years of stagnant pay. The scale of the Covid-19 crisis, its cause, and the speed at which it hit, have all been different from previous crises. The highly unequal impact on different sectors, the scale of behavioural change – with reduced travel and the shift of spending online – and the very unusual policy response – including the Coronavirus Job Retention Scheme (JRS) – also mark this out as not just any old recession. So a keen focus on emerging developments, rather than history, may be the best guide to the labour market legacy of this crisis.

Moreover, Covid-19 has brought new challenges for the labour market – many of which could outlast the pandemic. Rather than leaving us with high unemployment, the falls in employment (albeit smaller than previous crises) have shown up in rising economic inactivity as some workers leave the labour force entirely. From remote working to changes in the size of different sectors, Covid-19 has also changed the way that we work. In this report, part of the Economy 2030 Inquiry, we consider the nature of these labour market developments with an eye on the longer-term changes, and the resulting challenges for policy makers over the rest of this decade.

The Covid-19 crisis hit the UK labour market hard through 2020 and early 2021. However, the economic reopening has seen a swift recovery. Vacancies have boomed, with over 1 million vacancies each month since July. This means that, despite a higher unemployment rate, there are only 1.3 unemployed people per vacancy, the lowest level since records began in the early 2000s.

This hiring boom means we are not seeing the rise in unemployment and poor- quality work that has characterised most recessions. Although there is some sign that the share of workers aged 18-24 on zero-hours, temporary or agency contracts is beginning to rise, rates of these types of insecure work remain below pre-crisis levels. Moreover, the prospects for unemployment are far better than past recessions. Unemployment rose by 6, 4 and 3 percentage points respectively after the 1980s, 1990s and financial crisis recessions – but unemployment peaked at 5.2 per cent last winter, just 1.2 percentage points above its pre-crisis level, and as of Q3 2021 was just 4.3 per cent, compared to 4.0 per cent on the eve of the crisis. Additionally, long-term unemployment has been falling since the spring, and youth unemployment is now just below its pre-crisis level.

These staggering success stories are thanks in the large part to unprecedented government intervention, in the form of the Coronavirus Job Retention Scheme (JRS), which protected over 11 million jobs. The remaining uncertainty is by how much unemployment will rise this autumn as the JRS comes to an end. Official forecasts of the unemployment rate at the end of the JRS have improved significantly, falling from over 10 per cent in the summer of 2020 to around 5 per cent in the autumn of 2021 in large part because of (sometimes belated) decisions to extend the scheme until virus-related restrictions were largely lifted. A new survey of workers in the field in mid-October suggests that the overwhelming majority (88 per cent) of those furloughed in September are in work in October, with the remaining 12 per cent – corresponding to 136,000 people – out of work. This is the first direct evidence of furloughed workers’ own experience after the JRS ended, and supports the idea that unemployment is expected to rise by only a small amount this autumn. We should not, therefore, downplay the success of the JRS in holding down unemployment during the worst recession for over 300 years, and one with its roots in the labour market.

This is not to say that policy makers should be complacent. One-in-20 (5 per cent) of those in work in October – and one-in-six (16 per cent) of those who had experienced furlough for at least six months during the crisis – fear for their jobs. Supply-chain pressures, weak household incomes and rising input prices will act as headwinds to the recovery this winter. In addition, the impact of the crisis remains unequal: the lowest-paid workers were over three times more likely to have experienced at least three months of furlough or worklessness over the past 20 months than the highest-paid workers (36 per cent for those in the bottom quarter of the pay distribution vs 10 per cent for those in the top). Labour market disruption over the past 20 months has also disproportionately affected the youngest and oldest workers, and ethnic minority workers.

The self-employed have also been hard-hit by the economic impacts of the pandemic. As a result, many have interpreted falling self-employment numbers in official surveys as a sign of self-employed people being not only more likely to stop working entirely, but also to leave self-employment for the safety of being an employee. But we find that there has been no increase in people saying that they have made the switch from self-employed to employee, suggesting that much of this fall represents changes in classification, rather than a genuine change in economic activity. Overall, the evidence suggests that high unemployment will not be the primary lasting challenge for the labour market post-pandemic. But another form of worklessness still may be, with economic inactivity rising after years of progress.

The success of the JRS has been one factor in holding down the unemployment rate. But, crucially, not everyone who has lost their jobs has moved into unemployment. The first year of the Covid-19 crisis led to a larger fall in labour market participation among working-age adults than any other crisis in the last four decades, in stark contrast to the long-run trend of rising participation prior to the pandemic. Although participation has begun to rise slightly in recent months, the fall remains larger than at comparable points in both the 1980s recession and the financial crisis, despite a smaller fall in employment. Overall, the number of people who are economically inactive – that is, not working or looking for work – has increased by 586,000 since the start of the crisis, of whom 364,000 are working-age. This is considerably larger than estimates of the number of migrants who left the UK during the pandemic, but has received far less attention in discussions of current reported labour shortages in some sectors. Rising inactivity on this scale matters not only for the individuals concerned, but also for the economy. Higher employment was a key contributor to what little income and GDP per capita growth we did see in the 2010s, and both the Bank of England and the Office for Budget Responsibility have suggested that at least some of the rise in inactivity could prove to be permanent, contributing to economic scarring in the longer term.

But this overall fall in participation hides a complex picture. There are two groups whose changes in participation have been both particularly large and distinct in this crisis. First, older working-age people (aged 45-64) have reduced their labour force participation during the pandemic. This is unusual: their participation rose during the financial crisis. Among 55-64-year-olds, for example, participation has fallen by 1.2 percentage points during the Covid-19 crisis, whereas it rose by 1.4 percentage points after the financial crisis. Moreover, although both men and women in their 50s have seen similarly large falls in participation during the pandemic, the Covid-19 crisis has been particularly bad for older women’s employment relative to previous recessions. By the three months to August 2021, the employment rate of women in their 50s had fallen by 2.1 percentage points. This puts the trend of the past decade into reverse – in the decade prior to the crisis, adults aged over 50 accounted for 88 per cent of the increase in the labour force – and contrasts with rises in the 1990s recession and the financial crisis. If these older workers do not return to work as the economy rebounds, this reduction in labour supply could outlast the pandemic.

Second, there is a sharp gender divide to labour force participation among younger workers. Among adults aged 25-44, labour force participation among men has gone down by 1.1 percentage points, but it is up by 1.8 percentage points among women, particularly among mothers (almost three-quarters (74 per cent) of mothers of 0-3-year-olds were in the workforce in 2021, compared to around two-thirds (68 per cent) in 2019 and 2017). This means that women now make up almost 48 per cent of the workforce, up from 47 per cent in 2019, and 44 per cent in 1992. It also goes alongside a shift of about half a million women from working part-time to full-time since the pandemic began.

Some have put this down to the effect of homeworking, and we find some evidence of that, particularly among mothers: one-in-ten (13 per cent) coupled mothers said that remote working has enabled them to increase their labour supply, compared to 5 per cent of both coupled fathers and coupled women without children. But there are two other drivers of increased female participation, both of which continue pre-crisis trends. One is that, similar to the experience after the financial crisis, second earners have been increasing their labour supply in response to labour market disruption experienced by their partner. One-in-seven (15 per cent) of adults in couples whose partners were furloughed on less than full pay have entered the workforce or upped their working hours during the pandemic, compared to one-in-ten (9 per cent) of those whose partners were furloughed and received all their earnings. The second is the sectoral shifts over the last two years, which, on balance, have favoured female-heavy sectors – such as public administration, IT and communications, and professional and other services – at the expense of male-heavy ones, even allowing for the declines in hospitality and retail.

Some of the rise in inactivity could also be linked to specific features of this crisis: long Covid, or fear of the virus. Our survey suggests that around 600,000 adults have either left the workforce since the pandemic started, or are working fewer hours, for these reasons, with around half of those (260,000) expecting this to last for at least six months. The age profile is not particularly slanted to older workers: it is more reflective of the age profile of those currently reporting symptoms of long Covid, something that is skewed towards younger people and reflecting recent Covid-19 caseloads. The number of people who report being economically inactive because of long-term health reasons has risen since the start of the pandemic, but by only 150,000 (and the pre-pandemic trend had been upward too). Similarly, there has been little change overall in the number of people leaving the workforce each year for health reasons more broadly – suggesting that any exits from the workforce due to Covid-19 have been offset either by reductions in other health problems or by changes to work (such as remote working) supporting those with other conditions to remain in employment. Policy makers should nonetheless remain alert, as there are clearly risks that long Covid cases increase, or that the continuing presence of Covid-19 keeps those who are worried about the virus out of the labour force permanently.

The pandemic has not only changed the amount of work people are doing – it has changed the types of jobs being done in the economy too, and such change has been unusually rapid. From the very start of the crisis, the pandemic has been sectorally uneven, as lockdowns and social distancing restrictions closed customer-facing sectors like hospitality while demand ramped up in sectors like health and care. Although some of these trends have reversed rapidly as the economy has opened up, others may prove to be more permanent. As of October 2021, employee jobs in retail, hospitality and leisure combined remained 146,000 below pre-pandemic levels, even as overall employee jobs were higher than the start of 2020.
The scale of this reallocation of workers, and process by which it happens, are important. First, the speed of the recovery in the labour market this winter will depend on how efficiently vacancies can be filled. Second, a successful recovery for household living standards needs not just vacancies to be filled quickly, but also that the labour market efficiently sorts workers into roles where they can be most productive. We consider both of these, looking first at how the sectoral change has been achieved, and then at what that means for the match between workers and the jobs they are doing.

First, we find that it is not the case that sectoral reallocation has happened through workers moving from hard-hit industries to booming sectors. At the height of the crisis, the number of workers moving jobs each quarter fell to record lows. Even by Q3 2021, when job-to-job moves recovered to the highest level on record, just 43 per cent of job switchers moved to a new industry – the lowest since at least the early 2000s. There has also been little change in the share of workers who change sector within a given year, which has been stable at around 5 per cent of the workforce since 2014. This suggests that changes in the relative sizes of different sectors are happening primarily by new entrants to the labour market moving into growing sectors and people leaving shrinking sectors into unemployment or inactivity.

This meant that at the start of the crisis there was a substantial mismatch between the industries where jobseekers have most recently worked (largely the shrinking sectors, like leisure and hospitality) and those with available vacancies (particularly health and social care). But as the economy rapidly opened and hiring reached record levels over the summer, mismatch was no longer a problem as jobseekers were able to return to the previously hard-hit sectors. But policy makers should be alert to the risk that mismatch could become more of an issue in the longer term – especially because, in future, we will not be able to rely on a rapid economic reopening to resolve it. For example, there is little relationship between the industries where unemployed people are looking for work and the sectors with the highest number of vacancies: despite the hospitality sector having 151,000 vacancies in October, for example, unemployed respondents to our survey were no more likely to be looking for jobs there than in finance and real estate, which had just 62,000 vacancies.

As well as considering the speed of labour market adjustment, we are concerned about its quality. Here there are worrying early signs that, as sectors like hospitality and leisure have reopened, workers are moving into occupations that do not match their skill level. The share of workers who are in a higher-skilled job than a year earlier has fallen back to levels last seen in 2012, and moves into lower-skilled jobs have increased since before the pandemic, while the share of people who are over-educated compared to the average level of educational attainment for the occupation they are in has reached a record high. As the recovery continues to take hold and a wider range of opportunities become available, policy makers need to support workers to move into jobs that match their skills and experience; such reallocation could be a fairly easy way to boost productivity.

The other unique change wrought by the pandemic has been the large-scale move of some parts of the workforce (mostly those in higher-paying occupations) to remote working. Our survey shows that workers expect the increase in hybrid work – with some hours in the office and some at home – to outlast the pandemic, but they expect that the share of workers who work exclusively from home to fall back to pre-pandemic levels by October 2022. But we also find that workers have got a taste for homeworking, and would ideally like to spend around one-tenth (8 per cent) more of their working hours at home than they were doing in October 2021.

Survey evidence also confirms that higher earners are still most likely to be working from home – and, as a result, those in higher-paid places, London, have the highest reported rates of homeworking. Although there are only small differences between ages, pay and geography in the fraction who expect to be working entirely from home next October (between 9 and 14 per cent of workers in different groups), there is considerable variation in workers’ expectations of working in a hybrid manner: less than one-third of the lowest-paid (31 per cent) and oldest, aged 55-65 (32 per cent), workers expect to take part in this new working model, compared to almost two-thirds of Londoners (65 per cent) and the highest paid (66 per cent).

Remote working is also having an impact on people’s work-life balance. Despite remote work enabling some parents (particularly mothers) to enter work, mothers are significantly less likely to expect to have a hybrid working pattern in the future than women without children (52 per cent vs 38 per cent). But this seems to reflect the constraints of their jobs rather than their preferences: given the chance, mothers would like to spend significantly more time working from home (and more so than other groups).

Overall, the accelerated move to hybrid working, amid wider variation in firm practices and worker preferences, may increase the volume of reallocation of workers as people renew their preferences about where they live and work. Educated professionals will have more options about where to live within, but not between, regions. And the proliferation of hybrid work, with workers in the office and city centres some of time, will likely not mean the end to cities’ roles as the centre for production and basis for agglomeration. Finally, while many are focused on the change that hybrid working brings for higher earners who now have greater lifestyle choices, the more important lasting effect will be on low-paid workers who will need to find new jobs in new places as they respond to the demand of higher-paid workers.

In this report, we have provided insights into how the labour market is evolving as we begin to emerge from the pandemic. It remains to be seen exactly which changes from the pandemic stay the course, but it is already clear that there are lasting changes on the horizon. The Covid-19 pandemic is very different from previous downturns, with larger (and more complex) falls in participation, a unique sectoral impact, and the emergence of new working models. As policy makers consider the challenges for the post-pandemic labour market – in combination with other challenges such as Brexit and the net zero transition – they will need to take the unique nature of the Covid-19 crisis into account.

• Unemployment in Q3 2021 was just 4.3 per cent, compared to 4.0 per cent on the eve of the crisis. Long-term unemployment has been falling since the spring, and youth unemployment is now just below its pre-crisis level. This is a much better performance than in previous crises.
• The end of the JRS has not led to a spike in unemployment: survey evidence suggests that 88 per cent of those furloughed in September are in work in October.
  Instead of leading to unemployment, the fall in the number of workers through the crisis is showing up in higher numbers of economically inactive people – those not working or looking for work. This has increased by 586,000 since the start of the crisis, of whom 364,000 are working-age.
• Among 55-64-year-olds, participation has fallen by 1.2 percentage points during the Covid-19 crisis; it rose by 1.4 percentage points after the financial crisis, This reverses the trend of the past decade– in the decade prior to the crisis, adults aged over 50 accounted for 88 per cent of the increase in the labour force.
• There is a sharp gender divide to labour force participation among younger workers. Among adults aged 25-44, labour force participation among men has gone down by 1.1 percentage points during the pandemic, but it is up by 1.8 percentage points among women. 74 per cent of mothers of 0-3-year-olds were in the workforce in 2021, compared to 68 per cent in 2019 and 2017. Women now make up almost 48 per cent of the workforce, up from 47 per cent in 2019, and 44 per cent in 1992. This is partly due to homeworking allowing some carers to work, partly due to second earners working more to offset labour market disruption experienced by partners, and partly due to sectoral shifts which, on balance, have favoured female-heavy sectors.
• The large-scale sectoral reallocation over the past two years has not happened through workers moving from hard-hit industries to booming sectors. Instead, changes in the sizes of different sectors are happening primarily by new entrants to the labour market moving into growing sectors and people leaving shrinking sectors into unemployment or inactivity. Although vacancies are filling quickly, a worrying future sign is that there is little relationship between the industries where unemployed people are looking for work and the sectors with the highest number of vacancies.
• There are signs of growing mismatch in the labour market, too. The share of workers who are in a higher-skilled job than a year earlier has fallen back to levels last seen in 2012, and moves into lower-skilled jobs have increased since before the pandemic. The share of people who are over-educated for their occupation has reached a record high.
• Survey evidence suggest workers expect the increase in hybrid work to outlast the pandemic, but not fully remote working. Hybrid working will give educated professionals more options about where to live, and more opportunities to balance work and family commitments. But the more important impact will be on low-paid workers who will need to find new jobs in new places as they respond to the demand of higher-paid workers.