Labour Market Outlook Q2 2021

As Covid-19 restrictions begin to ease and more sectors begin to reopen, there are positive signs of recovery in the labour market. The number of payrolled employees has begun to increase (although it is still 700,000 lower than it was in early 2020); furlough rates are falling; and vacancies have recovered to pre-crisis levels. Despite the swift roll-out and encouraging results of the vaccines, however, concern over new Covid-19 variants and rising cases mean that the outlook for reopening the economy is still uncertain.

There may be further labour market challenges to come: both the Bank of England and the Office for Budget Responsibility expect the peak of unemployment to come later this year, when the Government plans to end the Coronavirus Job Retention Scheme (JRS). Even as the majority of furloughed workers are expected to have returned to work, any lingering social distancing restrictions or reduced demand could lead to businesses letting go the remaining minority of staff when support is withdrawn. And rising unemployment could have lasting consequences for job quality, particularly for the lower-paid workers who have borne the brunt of Covid-19-related job losses and pay cuts.

In this Outlook, we focus on the impact of the Covid-19 crisis on women, unpicking the reasons for the smaller-than-expected labour market impact on women relative to men, and examining the impact on parents. Our Lifting the Lid section looks at the uneven employment recovery in different areas of the UK, long-term youth unemployment, and the recovery in job vacancies in different sectors.

Spotlight

The economic impact of the Covid-19 crisis has not been evenly-distributed. As customer-facing sectors have borne the brunt of the shutdowns, the impact has fallen disproportionately on the low paid and the young (and, to a lesser extent, older workers). But one area where the impact has been surprisingly equal is the impact on men and women. Not only is that at odds with what many (including the Resolution Foundation) expected, it also contrasts somewhat with countries like the US and Japan, where there have been widening gender gaps in labour force participation, with women hit harder than men. In this Outlook, we explore why gender differences in the labour market shock caused by Covid-19 in the UK are much smaller than expected, and whether this also holds among parents.

Early on in the crisis, one of the first pieces of evidence suggested that women were significantly more likely than men to have lost their jobs. But subsequent evidence has consistently shown a far smaller gender difference in labour market outcomes like employment, furlough, and hours worked.[1] In fact, Figure 1 shows that according to the Labour Force Survey, the employment rate among men has fallen by more since the start of the crisis than it has among women (by 2.4 per cent and 0.8 per cent respectively), driven by large falls in self-employment. On the other hand, increases in female full-time employment – with the largest rises in education, public administration, and professional services – have counteracted much of the fall in part-time work and self-employment among women.[2] And there have also not been substantial falls in women’s labour force participation: economic inactivity among women aged 16-64 actually reached a record low in the three months to February 2021, while men’s economic inactivity reached a joint record high.

Figure 1: Men’s employment has fallen more than women’s

Figure 2 focuses on outcomes in March 2021 of people who were in employment before the crisis. At that point, there was a small gender gap, driven by women being more likely than men to have lost hours due to the crisis (4 per cent compared to 2 per cent).[3] But the gradient is still far smaller than those seen across different sectors, age groups, and pay quintiles: in March 2021, for example, the lowest-paid workers were three times more likely to have experienced a labour market hit than the highest-paid. This reinforces Resolution Foundation findings throughout the crisis (in May 2020, October 2020, and February 2021), which have consistently shown roughly-equal impacts on men and women when it comes to job losses, furlough, and cuts to hours and pay.

Figure 2: In March 2021, women were slightly more likely to have lost hours and pay due to Covid-19

Early expectations that women would be hardest hit in the crisis were based on the fact that women are more likely to work in the lower-paying, face-to-face sectors, like hospitality and retail, that have faced the biggest impacts from social distancing restrictions. But in fact, across the whole economy, there has been little relationship between the sectors where employee numbers have fallen fastest and those which employ more women than men (see Figure 3). There are three key reasons why these early concerns were unfounded. First, the impact of the crisis on sectors like manufacturing, which are overwhelmingly male, was underestimated in early modelling: in reality, more businesses than expected stopped operating over the first lockdown despite not being mandated to close, or faced demand shortfalls. Second, although women do dominate in some of the hardest-hit sectors, women also make up more than 70 per cent of health and education workers, sectors that were protected from the worst economic impacts of the crisis – although many of these frontline key workers were, of course, at far greater risk from the health impacts of catching the virus. More generally, women in employment were twice as likely as men to be in the public sector before the crisis (30 per cent of women in 2019, compared to 14 per cent of men), providing further employment protection.[4] And finally, it is important to note that Figure 3 only covers employees, and that men were around twice as likely as women to be self-employed before the crisis, a group that has been affected much more than employees.

Figure 3: The crisis has hit both male- and female-heavy sectors

So the gender mix in the sectors that have been hit hard during the crisis – and those that have expanded – helps explain why the overall labour market impact varies little between men and women. But, of course, this crisis has affected our lives in ways that materially affect gender inequalities. For example, as schools and childcare providers closed to most children, parents had to balance work and home-schooling – and the evidence is clear that women took on the majority of the burden.

Figure 4 shows how the working hours of men and women have evolved over the course of the crisis, split out by whether or not they have children under 18. At the height of the first lockdown in April 2020, the average fall among all groups was more than 30 per cent: women with children had the largest average fall in their working hours, at close to two-fifths (37 per cent), compared to between 32 and 34 per cent among non-parents and fathers. More strikingly, however, by July 2020 – when much of the economy had opened up following hospitality’s reopening on 4 July, but schools remained closed – the average mother’s working hours were still down by almost a quarter (24 per cent) on their pre-crisis level, a fall almost twice as large as non-parents (13 per cent) and four times the size of fathers (6 per cent). The pattern of working hours over the course of 2020 suggests that school closures had a far bigger impact on mothers’ working hours than those of fathers, consistent with evidence from time use data.

Figure 4: Mothers’ working hours fell the most while schools were closed in 2020

The return to home schooling for many at the start of 2021 does not seem to have had as big an impact on how many hours parents are working, perhaps linked in part to higher attendance rates among those children who were allowed to attend school in person, and there was a far smaller gap in mothers’ and fathers’ working hours (Figure 4 shows that, in January 2021, the average mother’s working hours were down 11 per cent on pre-crisis, compared to 8 per cent for fathers). Although this might be encouraging for those concerned about gender inequalities, just counting how many hours are being worked hides the fact that many mothers were shouldering the burden of home-schooling as well as working close to their pre-crisis hours. In January 2021, although mothers were only slightly more likely than fathers to say they had reduced their hours due to childcare or home-schooling (13 per cent and 11 per cent respectively), they were 5 percentage points more likely to say they had adjusted their working patterns for the same reasons (a further 18 per cent, compared to 13 per cent of fathers).[5] Mothers’ mental health was also significantly worsened by school closures, while fathers, on average, appear to have faced no impact.

Looking at non-parents’ working hours in Figure 4 shows something more surprising: by January 2021, the average woman without children was working more than before the crisis, consistent with the growth in full-time work shown in Figure 1. Overall, the fall in women’s total hours worked (which reflects both changes in employment levels and average hours worked by those in work) has been around one-third smaller than that of men’s.[6] This pattern of some women increasing their hours during the crisis is consistent with changes in working hours after the financial crisis: many women (particularly second earners) increased their labour supply in response to the income shock.

Figure 2 and Figure 4 show data just before the early 2021 lockdown ended. Since then, retail and most of hospitality have opened up, payroll employment is beginning to rise and vacancies have recovered to pre-crisis levels. It is difficult to predict the short-term prospects for the recovery, but there is little sign that there will be a gender skew in any further rises in unemployment. The numbers of men and women on furlough were similar at the end of April 2021 (1.64 million and 1.69 million respectively), a narrowing of the small gap that existed at the start of the year. Unless there are substantial changes to the distribution of furloughed workers over the summer, this suggests that any rise in unemployment when the JRS comes to an end might be gender equal.

Figure 4 showed that in March, the average woman with children was still working 7 per cent fewer hours than at the start of 2020 – a fall similar in size to that experienced by fathers, and smaller than that of men without children. But others are already out of work. Figure 5 considers how quickly people who lost jobs during the crisis are moving back into work. It suggests that women were re-entering work at less than half the rate of men in March 2021 (15 per cent and 36 per cent respectively). While it is difficult to know exactly why this is, one explanation could be that the burden of home-schooling halted some women’s job search or led them to put off looking for work while they looked after children. (Sample sizes prevented a further split into parents and non-parents by gender, but job entry rates fell among parents and rose among non-parents at the start of 2021.) Women were, however, more likely than men to re-enter work over the summer and autumn of 2020, suggesting that moves into work could pick up going forward.

Figure 5: The job entry rate among women has fallen in 2021

The headline gender differences in the labour market impacts of this crisis, then, have been much smaller than many expected, and smaller than suggested by work at the very start of the crisis. The sectoral hit of this crisis has hit both male- and female-heavy industries, while women have been relatively protected (from the economic impacts, if not the health impacts of the virus) by their over-representation in the public sector. Women without children who kept their jobs are now working more than before the crisis, and the large fall in self-employment has hit men much more than women. But closures of schools and childcare providers have introduced a labour market hit to women that would not otherwise have existed, bringing with it a mental health hit and risking compounding the pay penalty that mothers already face in the labour market. And looking further ahead to a ‘new normal’ in the workplace, there is a risk that a move towards hybrid working could disadvantage women: fewer women than men say they want to return to the office full-time, which could harm women’s career progression if offices become more male-dominated. As we enter the recovery, policy makers should keep a watchful eye on any further rises in job losses among women (and mothers in particular), and support those who have left employment to get back into work. And now more than ever, policy makers and employers alike should support working parents, no matter their gender, to balance work and childcare without detriment to their longer-term career prospects.

[1] In contrast, the financial crisis affected men more than women: men’s unemployment peaked at 9.1 per cent, compared to 7.8 per cent among women.

[2] Source: RF analysis of ONS, Labour Force Survey.

[3] The lower rates of furlough, job losses, and hours cuts among parents overall tend to disappear if we account for the worker’s age: parents are less likely than non-parents to be in the (hardest-hit) youngest or oldest age groups, and less likely to work in the sectors that have faced the tightest restrictions such as hospitality, retail, and leisure.

[4] Source: RF analysis of ONS, Labour Force Survey.

[5] Source: RF analysis of ISER, Understanding Society.

[6] ONS data suggests that women’s total hours worked were down 5 per cent between January-March 2020 and February-April 2020, compared to 8 per cent among men. Exact levels should be used with caution, as the ONS are in the process of reweighting their data to take account of population changes since the start of the pandemic, but here we assume that the reweighting process will not have a significant effect on the difference between men and women.