Past recessions saw steep rises in unemployment or wage cuts. Tracking the performance of the UK labour market during Covid-19 shows that the most notable effects of a downturn caused by restrictions on mobility are the more dynamic aspects: falls in hours, hiring, job quits and house moves.
Recessions usually happen when people or firms stop demanding things. The Covid-19 recession is not very different in this regard, even if the cause of the demand fall – the restrictions on mobility imposed to try to stop the spread of the virus – was very unusual. Faced with a big economic shock like this, firms can take a hit to their profits, adjust their costs or both.
To a firm, it may be easier to release workers or stop hiring than to sell off land, property or machinery if the downturn is expected to be relatively short. But firms can also adjust their labour costs through changes in wages, hours or personnel (or indeed use a mixture of all three). Workers have some influence on this through unions and their intrinsic worth to an employer.
It is also possible that in a downturn, some people may withdraw from the labour force and do other things (such as stay on in education). This all means that mass layoffs should usually be a last resort, but hiring rates, hours or wages may change as a first response to a shock. Nor is work a uniform activity: self-employment and temporary working are more cyclical, typically falling in bad times and rising in good times, and they are often among the first types of working to decline in a downturn.
UK institutions also matter. There are laws regulating the notice period for redundancies, but there may also be specific interventions during a crisis, such as cuts in interest rates or taxes or job search/placement programmes for the unemployed. The Coronavirus Job Retention Scheme, (CJRS or ‘furlough’) – in which the government pays up to 80% of the wages of workers absent from work but still on the payroll – has never been tried before in the UK. Its presence is almost certainly a major influence on the key indicators of labour market performance.
What can the latest data tell us?
Using the Labour Force Survey (LFS) – the survey used to estimate the official UK unemployment rate in which households are sampled in every week of the year – over the course of the pandemic, we can compare the weekly value of any labour market indicator in the crisis period relative to the average weekly equivalent over the preceding five years.
The dashed vertical lines in the charts show the point of the first registered coronavirus-related death (2020, week 5), as well as the week the UK first went into lockdown (2020, week 12). The light grey shading in the figures indicates the ‘usual’ range of each indicator in any week. Any 2020/21 outcome that lies outside this shading is interpreted as a significant departure from recent norms (highlighted in turquoise).
Not much unemployment
Using the most common metric of labour market performance (unemployment), not much can be observed over the whole of the pandemic that was unusual – see Figure 1. Prior to 2020, the UK labour market had been performing rather well, with the national unemployment rate hovering around 4-5%. But the weekly unemployment rate in the first months of the Covid-19 crisis was well below the average of the past five years.
Since then, it has risen, but not much above rates seen recently at the same point in the year. This is very unusual. Unemployment normally rises a lot in recessions. Even the youth unemployment rate, normally a harbinger of recessionary doom, is similarly unaffected.
Figure 1. UK unemployment
Panel A: Total
Panel B: Youth (ages 16-24)
The employment mix may have changed
While overall employment has not fallen much, there has been some change in working mix – see Figure 2. The share of full-time employment has risen over the downturn at the expense of part-time working and, to a lesser extent, self-employment. This suggests that these latter forms of working have been hit harder in the downturn. The share of temporary working has changed little over the crisis.
Figure 2: Employment mix
Panel A: Full-time employees
Panel B: Part-time employees
Panel C: Self-employed
Panel D: Temporary work
No pause in wage growth
There has also not been any obvious reduction in wages. In the 2008-11 downturn, there were notable wage freezes for many workers, which arguably mitigated the need for mass layoffs. Yet Figure 3 shows that average weekly wage levels (adjusted for inflation) appear to have grown strongly over the crisis.
Figure 3: Average weekly wage levels (inflation adjusted)
Fewer sickness absences
Despite Covid-19, the weekly absence rate from work due to (self-reported) illness throughout 2020 and into 2021 was noticeably below average – see Figure 4. Around 2% of the employed workforce are off sick in a typical week, with fewer sickness absences in Christmas week and the spring and summer, but more in the run-up to Christmas and though the winter. Sickness absence in the year of the pandemic has been a near all-time low.
Figure 4: Sickness absence
Furlough in action
Things do, however, look less than normal across other possible margins of labour market adjustment. Consider first the percentage of workers who say they had a job but were away during the survey week. The weekly norms show large spikes around Christmas, Easter and the summer holiday season. But there is a notable departure from the normal level, beginning in week 10 of 2020 and increasing rapidly to around 24% of the employed working age population by week 16 (four weeks into lockdown). This then slowly falls back for the remainder of the crisis to approximately 10% – a level still way above the norm.
This is the furlough scheme in action – see Figure 5. The UK government’s scheme only compensates employers if their workforce remains at home. A similar scheme applies to the self-employed unable to work through the crisis. Over the duration of the pandemic, there were around 132 million additional person-week absences – 31.5 million employed multiplied by the cumulative percentage of additional absences (418%). This equates to the entire UK workforce doing nothing for four weeks and one day.
Figure 5: Absent from work but retained job
The overall shock to output is made larger because among the majority of employed in work and not away from their job, many reported working fewer hours than usual over the first six months of the crisis. Since August 2020, however, hours worked look closer to normal, perhaps as everyone adjusts to the ‘new normal’.
In a typical week, the entire workforce puts in around 1,000 million person-hours of work. The seasonal fluctuations around this norm are apparent in Figure 6, with lower hours in the summer months and at Christmas. But the chart suggests that there have been around 3,800 million fewer hours worked since the crisis began. This is equivalent to the entire workforce doing nothing for three weeks and four days. Added to workplace absences, this equates to eight weeks of lost output (about 15% of a year’s work).
Figure 6: Seasonal hours worked (millions)
Nobody is hiring
Employers can also save on costs by restricting hiring. And indeed, hiring fell noticeably throughout the crisis. Figure 7 shows that usually between 0.5-1% of the workforce is newly hired in any week, with relatively fewer hires in spring, and autumn being the main hiring period (around 300,000 people a week).
For younger adults, between 1-4% of the workforce are in a new job in any week. Over the first six months of the crisis, hiring stalled significantly and remained below average thereafter. So, the pandemic has compromised this typical feature of labour market change severely.
Figure 7: Hires
Panel A: Total
Panel B: Youth (ages 16-24)
There have been layoffs…
With hiring stalled, employers might have adjusted their workforce costs enough without the need for further changes. This may indeed have been so for the first five months of the crisis (until the initial furlough scheme began to unwind). It is at this stage that the weekly pattern of redundancies (layoffs) in the working age population begins to rise – see Figure 8.
At their peak, redundancies were around 30,000 a week – 0.1% of the total workforce. This is clearly bad for those laid off but is still quite small as a proportion of the workforce and not as severe as the 2008 downturn, let alone the 1980s crash.
And not all of these layoffs will end up in unemployment. Notice periods effectively give people time to find alternative employment. But with hiring also significantly down, we might expect unemployment to have risen as a result. Yet the rise in redundancies does not seem to have filtered into significant rises in unemployment.
Figure 8: Layoffs and quits
Panel A: Layoffs
Panel B: Quits
… but no one is leaving
One additional finding shown in Figure 8 is that the number of people who leave their jobs voluntarily has fallen substantially. While layoffs show little seasonal pattern through the year, quits are typically higher in the autumn (when hiring is also at its highest). But with hiring stalled, there have been much fewer incentives for workers to quit and find new work. In this sense, the UK labour market is much less dynamic than it was before.
No one is moving house either
Often house moves accompany job moves – either new jobs or promotions. As Figure 9 shows, fewer new jobs and promotions are consistent with a significant fall in the weekly share of individuals moving house (in the last month) relative to recent norms.
Figure 9: House moves
What have we learned?
The coronavirus recession is different. Typical metrics of labour market performance were not noticeably affected over much of the pandemic in the UK. Instead, the crisis is marked by higher absences from work, a large rise in short-time working and hiring freezes, rather than wage cuts and mass layoffs. The furlough scheme and associated government interventions almost certainly underpin much of what we see, and they are likely to continue to shore up the viability of many businesses, providing support to the labour market.
A recession caused by restrictions on mobility has had most effect on the labour market features most reliant on mobility – namely work absences, hiring, job quitting and residential moves. The labour market is much less dynamic than in normal times and this matters for a healthy economy. Should this unlock if and when the pandemic subsides, there should be little long-term damage. But if things persist, there could be more serious consequences.
Where can I find out more?
- Coronavirus and the Effect on Labour Market Statistics: This report by the Office for National Statistics looks at the effect of lockdown on collation of labour market information.
- Single Month and Weekly Labour Force Survey Estimates: This article by the Office for National Statistics discusses the possibilities of using weekly information from household surveys
- Labour Markets in the Time of Coronavirus: Measuring Excess: This article by Jonathan Wadsworth discusses how to construct and portray weekly labour market information as used in this article
Who are experts on this question?
- Jonathan Wadsworth, Royal Holloway & Centre for Economic Performance