1. Main points

  • The ratio of unemployment to vacancies increased between January and April 2020, indicating growing imbalance between labour supply and labour demand; the increase was driven more by falling demand for labour than by increasing supply.

  • The number of vacancies in the wholesale and retail trade (67,000), accommodation and food service activities (27,000) and information and communication industries (21,000) reached their lowest levels on record in the period March to May 2020.

  • The number of vacancies in small and medium-sized enterprises declined by 48.9% on the quarter to 180,000 in the period March to May 2020.

  • Experimental online job adverts data imply that the worst impact of the pandemic on labour demand was experienced in May and some sectors increased hiring in June 2020.

  • Average actual hours worked, adjusted for workers who were temporarily away from work, show that accommodation and food services (negative 8.4%) and education (negative 4.9%) experienced the largest declines in effective average actual hours.

Nôl i'r tabl cynnwys

2. Introduction

This article analyses the early impacts of the coronavirus (COVID-19) pandemic on the UK labour market. It offers additional analysis to what was covered in the Labour market overview for June 2020. It presents an overview of the UK labour market and illustrates how the coronavirus pandemic is impacting the labour market by analysing changes in vacancies and in actual hours worked. The analysis takes an industry approach to gain greater insight into the different industry-level impacts of the pandemic.

More about coronavirus

  • Find the latest on coronavirus (COVID-19) in the UK.
  • All ONS analysis, summarised in our coronavirus roundup.
  • View all coronavirus data.
  • Find out how we are working safely in our studies and surveys.

  • Nôl i'r tabl cynnwys

    3. Labour market overview

    The UK labour market estimates for the period February to April 2020 showed employment increased by 6,000 on the quarter to 33.0 million, but there are notable indications of the negative impact of the coronavirus (COVID-19) pandemic showing through falling vacancy numbers and actual hours worked and slowing earnings growth. Pay as you Earn (PAYE) Real Time Information (RTI) earnings estimates show that median pay declined by 3.5% between March and May 2020. Figure 1 shows the trend of PAYE-RTI paid employees. Between March and May 2020, PAYE-RTI paid employee estimates reduced by 612,000 to 28.4 million, and between April and May they reduced by 163,000.

    The effect of the pandemic on UK output is reflected in the falls in gross domestic product (GDP). In Quarter 1 (Jan to March) 2020, GDP is estimated to have fallen by 2.2%, largely reflecting the fall in output in March 2020 of 5.8%. The GDP monthly estimate for April 2020 showed that output fell by 20.4% in April 2020. The level of employment remained high partly because of the government's job retention schemes. In the period to 21 June, the Coronavirus Job Retention Scheme (CJRS) had provided income support to 9.2 million employee jobs and the Self-Employment Income Support Scheme (SEISS) had provided income support to 2.6 million self-employed jobs.

    The estimates of the number of employee jobs covered by the CJRS and SEISS are from HM Revenue and Customs' (HMRC's) experimental statistics. HMRC data give an all-industries employee jobs furloughing rate (28.0%) for the period up to 31 May 2020 that broadly aligns with that from the Office for National Statistics' (ONS') Business Impact of COVID-19 Survey (BICS) where the workforce furloughing rate was 29.8% for the period 18 to 31 May 2020.1 Redundancies as a result of the pandemic have so far remained low as firms were encouraged to furlough their workers rather than make them redundant. The BICS for the period 18 to 31 May 2020 showed that less than 1% of the workforce had been made permanently redundant across all sectors. This compares with the normal redundancy rate, which averaged 3.4‰ (per mille) in 2018 and 3.8‰ (per mille) in 2019.

    The figures for the ONS are in line with the International Labour Organization (ILO) classification of workers. The ILO classifies people whose absence from work is temporary, such as those furloughed under the existing government schemes, as employed. Thus, the measured UK unemployment rate remained relatively low and the employment rate remained relatively high, despite the pandemic causing major disruptions to the economy.

    The outlook of the labour market started changing in May 2020 as the government began to loosen some lockdown measures that had been introduced to reduce the spread of the virus. Some businesses started taking back workers they had furloughed, and others that had stopped trading because of the pandemic resumed trading.

    Notes: Overview of the labour market
    1. It should be noted that the two data sources are not fully comparable.
    Nôl i'r tabl cynnwys

    4. Vacancies and unemployment overview

    Vacancies represent the unmet demand for labour. Unemployment represents the excess supply of labour. The demand and supply of labour were impacted by lockdown measures and public health restrictions that were introduced to reduce the spread of the coronavirus (COVID-19). Business closures and restrictions to non-essential travel caused the demand for labour to fall.

    We can analyse the demand and supply of labour by calculating a ratio of the number of unemployed persons per vacancy, as shown in Figure 2. At its highest in the period September to December 2011, the ratio was 5.8. A higher ratio means there are more unemployed people per vacancy, which loosely translates into growing supply over demand for labour. A lower ratio indicates greater demand for labour than supply.

    The ratio was at its lowest (at 1.5) in the period January to May 2019, indicating how tight the labour market was before the pandemic. Since the beginning of 2019, the ratio has been broadly flat. The latest increase of the ratio (of 23.5%) to 2.1 in the period November to January 2020 and February to April 2020 was linked to the impact of the pandemic on vacancies, which experienced a large decrease of 166,000 (or 21%) to 641,000, while unemployment reduced by 8,000 to 1.34 million.  

    Nôl i'r tabl cynnwys

    5. Vacancies by industry

    The latest data available for Office for National Statistics (ONS) vacancies are for the period March to May 2020. In this period, the number of vacancies in the UK economy decreased by 42% on the quarter. External surveys like the latest KPMG and REC UK Report on Jobs for May 2020 showed continued decline in both permanent placements and temporary billings following record drops in April. The survey gave the coronavirus (COVID-19) pandemic as the reason recruiters' clients had cancelled or postponed recruitment plans, as the clients opted to wait until the outlook improved.

    The fall in the number of vacancies varies across industries. The Vacancies, jobs and public sector employment bulletin for June 2020 showed the extent to which vacancies declined in different industries. The industries showing the largest percentage decreases in the number of vacancies on the quarter, in the three months to May 2020, were accommodation and food service activities (negative 71%), art, entertainment and recreation (negative 55%) and construction (negative 54%). There was greater disruption of hiring and activities in some industries that rely on close interaction between people and that could not take advantage of technology to continue trading (for example, accommodation and food services). The closure of non-essential businesses would also have impacted or delayed hiring decisions. The human health and social work activities industry experienced the second-lowest decrease in vacancies (negative 18%) behind public administration and defence (negative 14%). The low decrease in vacancies in the human health and social work activities industry is partly associated with the increase in demand for treatment of COVID-19 patients. In this analysis, we focus on the growth trends of vacancies in industries that had percentage changes of 50% or more in the three months to May 2020, as shown in Figure 3. The rest of the industries are classified as "other industries". Figure 3 also includes an all-industries trend.

    In the selected sectors, the collapse in the number of vacancies in the current period is faster and deeper than during the 2008 to 2009 economic downturn. For example, in the three months to May 2008, vacancies in the accommodation and food services industry fell by 3%, while in the same period in 2020 they fell by 71%.

    The number of vacancies in the wholesale and retail trade (67,000), accommodation and food service activities (27,000) and information and communication industries (21,000) reached their lowest levels on record in the period March to May 2020. In the wholesale and retail trade industry, the worst affected subsector was motor trades. Disaggregating the wholesale and retail trade industry shows that in the motor trades and the wholesale subsectors, vacancies reached their lowest levels (3,000 and 10,000 respectively) during the 2008 to 2009 financial downturn, while in the retail subsector, vacancies reached their lowest level of 46,000 in the March to May 2020 period.

    Analysis of the impact of the pandemic on vacancies in businesses of different sizes shows that there has been a larger impact on vacancies in small and medium-sized enterprise (SME) firms (that is, those employing between 1 and 249 workers). In these firms, vacancies reduced by 48.9% on the quarter to 180,000 in the period March to May 2020. In large firms (employing 250 or more employees), vacancies decreased by 36.5% to 296,000 in the same period. In 2019, SMEs employed 60% of all workers, and HM Revenue and Customs (HMRC) data show that between April and 31 May 2020, they contributed 60% of the employee jobs furloughed. They, like large firms, have experienced falling vacancy numbers. The ONS' vacancies estimates for the period January to May 2020 show the fall in vacancies in SMEs is intensifying, particularly for businesses employing 10 to 49 workers.

    According to the Business Impact of COVID-19 Survey (BICS) for the period 18 to 31 May 2020, of the surveyed businesses that had not permanently stopped trading, 46.4% of SMEs had less than or up to six months of cash reserves compared with 35.2% of large businesses. A further 3.9% of SMEs and 3.3% of large businesses did not have any cash reserves. A recent Organisation for Economic Development and Co-operation (OECD) publication on SME policy responses to the coronavirus discusses how SMEs are affected by the coronavirus pandemic. The OECD reports that SMEs in industries like transport, construction, accommodation and food services, and "other" personal services were particularly adversely affected by the crisis as they also faced liquidity challenges. It also reports that across OECD countries, SMEs may have less resilience and flexibility in dealing with global demand shocks as well as there being higher costs to smaller firms for changes in work processes as a response to the pandemic.

    Nôl i'r tabl cynnwys

    6. Experimental online job adverts

    The extent of the impact of the coronavirus (COVID-19) pandemic is shown by leading experimental indicators. In this article, we use the Adzuna online job adverts data to complement the official statistics by analysing online job adverts for different types of jobs. The Adzuna data are published by the Office for National Statistics (ONS) on a weekly basis as part of the faster indicators. The data are a compilation of online job adverts that give early indications of what is happening to vacancies and recruitment. These data are not directly comparable with ONS Vacancy Survey data because of methodological differences, so they should not be considered as a direct reflection of the labour demand situation. They are also subject to volatility but still provide timely indications of early changes in labour market trends.

    The Adzuna data make it possible to analyse online job adverts in the economy. The Adzuna categories used do not correspond to Standard Industrial Classification (SIC) categories, so the job categories are not directly comparable with industries. However, the job categories in Figure 4 have been found to have high correlation with corresponding official statistics.

    The coronavirus pandemic affected industries and job categories differently. Industries containing non-essential businesses that were forced to close because of restrictions were the most impacted. For example, retail and wholesale, and catering and hospitality were the most affected categories in terms of online job adverts lost.

    Figure 4 shows that the index of job adverts in the catering and hospitality, and wholesale and retail categories declined steeply between January 2020 and 26 June 2020. On 27 February 2020, online job adverts in the wholesale and retail category were 12.2% lower than the average for 2019; on 26 June, they were 69.7% lower than the 2019 average. This shows that the loss of job opportunities intensified in the category between February and June 2020. On 26 June 2020, the catering and hospitality category online job adverts were 71.3% lower than the average for 2019. Online job adverts in education also declined between March and May 2020, before showing an upward trend in June 2020.

    In the health care and social care category, job adverts have remained broadly consistent with the 2019 average. On 26 June, the health care and social care job index was only 2.1 points below the 2019 average.

    Lockdown measures began to be relaxed in May 2020, with more businesses re-opening in June. These developments contributed to increasing hiring activities in some categories. The early signs of optimism were captured by the Adzuna online job adverts data. The indices in Figure 4 imply that online job adverts are now gradually trending upwards. The online job adverts index for all businesses bottomed out on 1 May 2020, at 41.8 points, and has been increasing to reach 50.5 points on 26 June 2020. The change for the health sector is likely to be small, as most of the human health and social work activities were not impacted by the coronavirus restrictions. Other industries were also hiring but at a reduced rate.

    Experimental online job adverts declined at a slower rate in some sectors in June 2020

    The decrease in online job adverts has been declining in some job categories. Figure 5 compares the changes in online job adverts on the last recording day of the month in March to June 2020 with the average for 2019. It shows by how many points the index for each job category differs from the baseline (the average for 2019). We look at the cross-sections of the categories in Figure 4. We comment on the trends rather than the specific points.

    Job adverts in wholesale and retail, catering and hospitality, and education job categories decreased the most at the end of May compared with the end of April. The three job categories reduced at a lower rate at the end of June, which may be linked to easing of lockdown restrictions. The all-businesses index followed a similar pattern. This suggests businesses’ hiring confidence is beginning to increase even though it is still lower than the pre-lockdown level. It is important to note that the categories in Figure 5 are volatile and not directly mapped to occupations or industries separately. The job adverts can be in any industry.

    Figure 5 shows that on 26 June 2020, of the selected categories, the smallest decrease in online job adverts was in the health care and social care sector1. The change for the health sector is likely to be small, as the human health and social work activities were not impacted by the coronavirus restrictions. As such, the online job adverts would not have had a large fall, and so they would be unlikely to rise by much. The human health and social work activities industry was hiring during the coronavirus restrictions. Other industries were also hiring but at a reduced rate.

    Notes: Job adverts
    1. The increase in online vacancies for health care and social care jobs was smaller than in other industries, but the ONS Vacancy Survey showed a decrease in the health care and social care industry between March and May 2020. There has been lower correlation between Adzuna and ONS vacancy data for this sector, so caution needs to be taken when referring to the sector.
    Nôl i'r tabl cynnwys

    7. Effective average actual hours worked

    The coronavirus (COVID-19) pandemic is affecting the number of average actual hours worked across industries. The Labour market overview for June 2020 used the experimental weekly Labour Force Survey (LFS) to show that the number of people who were temporarily away from a job trebled in the middle of March and at the end of April 2020. The people who were away from work include those who were shielding, on maternity or paternity leave, on furlough, or away for any other reason but were expecting to return to work. We cannot explicitly single out workers who were away from work because of furloughing. According to the rules governing the Coronavirus Job Retention Scheme (CJRS), furloughed workers were not allowed to work for their employers. They could work for other employers only if the workers' existing contracts of employment permitted them, in which case they would have positive hours.

    We analyse the estimates of effective average actual hours worked in each industry to show how the pandemic affected the number of hours worked by workers who were in work, to avoid overstating the fall in actual hours worked when we include workers who were on furlough. We differentiate between effective average actual hours worked and average actual hours worked that have not been adjusted for workers who were temporarily away from work (which we call unadjusted average actual hours worked).

    We calculate effective average actual hours worked by dividing the total actual hours worked by the number of people who were in work. The calculation excludes people who were temporarily away from work, who numbered 4.4 million in the three months to April 2020. Over half of this period relates to before the lockdown, which means it gives us a partial picture of the overall impact. In the three months to April 2019, there were 2.1 million workers who were temporarily away from work.

    Figure 6 shows the annual percentage change of effective actual hours worked by people who worked between February to April 2019 and the same period in 2020. It also shows the annual percentage change of unadjusted average actual hours worked in the same period. The comparisons of the same period in 2019 and 2020 allow us to identify the industries with the largest differences between effective actual hours worked and unadjusted actual hours worked. The differences are hours attributed to people who were temporarily away from work for various reasons, including because of furloughing.

    Figure 6 shows that effective average actual hours worked decreased in all industries except in the agriculture and fishing industry. The largest decreases in effective average actual hours were recorded in the accommodation and food services (negative 8.4%), education (negative 4.9%), energy and water (negative 4.8%) and construction (negative 4.6%) industries. Workers in these industries are working fewer hours on average in February to April 2020 compared with the same period in 2019. Many hotels, pubs and restaurants are not operating at their usual capacities because of the pandemic. Working patterns have also changed in the education sector as between 23 March and 1 June 2020, schools were closed to most students and most learning in schools and universities was taking place at home and online.

    The lowest decreases in effective average actual hours worked were in the health and social work activities (negative 0.5%) and “other services” (negative 0.5%) industries. Other factors that contributed to the decrease in actual hours worked were the introduction of reduced working hours in some industries and the impact of lockdown measures that restricted non-essential travel.

    Accommodation and food services had a 16 percentage point difference between effective and unadjusted actual hours worked. This was followed by the “other” industries category (which includes arts, entertainment and recreation and “other services”) with a 13.4 percentage point difference. In these industries, not accounting for workers who were temporarily away from work overstates the changes in actual hours worked. The industries with hours least affected by people who were temporarily away from work were energy and water (with a difference of 0.4 percentage points) and banking, finance and insurance (with a difference of 1.7 percentage points).

    Nôl i'r tabl cynnwys

    8. Actual hours worked in the UK and US

    The UK and US responded differently to the coronavirus (COVID-19) pandemic. The UK's job retention schemes allowed firms to put their workers on furlough rather than make them redundant. Therefore, the employment rate remained relatively high at 76.4% in the three months to April 2020. Between October 2019 and April 2020, the UK employment rate reduced by 0.1 percentage points. The UK unemployment rate increased by 0.1 percentage points to 3.9% between October 2019 and April 2020. Economic inactivity increased by 0.1 percentage points in the three months to April 2020.

    The US employment rate decreased from 70.7% in March 2020 to 62.4% in May 2020. The US unemployment rate increased from 4.4% in March 2020 to 13.3% in May 2020. Around half of the states had short-time work schemes in place, which allowed firms to reduce staff working hours according to business needs with a proportion of the shortfall in wages subsidised by the government. However, many workers were laid off, resulting in a large increase in unemployment and demand for unemployment support. In addition, economic inactivity increased by two percentage points between March and April 2020 to 14%.

    To understand better the impacts of the pandemic on the UK and US labour markets, we analyse the changes in actual hours worked in both countries. International comparisons should be interpreted with caution, in part because of differing measurement methods of some variables. For example, in the UK, furloughed workers were not allowed to work for their employers and could only work elsewhere if their contracts of employment allowed them to do so. These conditions restricted the number of furloughed workers who were working. Thus, we assume that broadly, furloughed workers and those who were temporarily away from work for other reasons caused a reduction to total actual hours worked because they did not work but were still counted as employed. Since they were still counted as employed, they caused the average actual hours worked to decrease. In addition, the unemployment rate remained low at 3.9% between January and April 2020.

    The laying off of workers in the US resulted in an overall reduction in actual hours worked. The calculation of average actual hours worked does not include the laid off workers who become unemployed.

    Figure 7 helps us to understand the overarching trends of monthly actual hours worked in the two countries. The UK data are experimental weekly average actual hours aggregated to monthly. They are less timely than the US data, so the datum point for May is missing.

    Figure 7 shows a clear impact of the pandemic on average actual hours worked in the UK. Monthly average actual hours worked deviated from trend by declining from March onwards and much more than in the US. Between February and April, the actual hours worked in the UK declined by around 21.4%.

    In the US, there has been no big change in the trend of actual hours worked. The average monthly actual hours have largely been steady in 2020, with a slight increase of 1.5% month-on-month in May 2020. A possible explanation for this is that despite the reduction in employment numbers, workers who remained in work may have had to put in extra hours to meet labour demand, which increased average actual hours worked. The comparative analysis of the impact of the pandemic on the UK and US labour markets shows the importance of analysing different indicators to get a full picture. Different labour market policies in the two countries result in different outcomes.  

    Nôl i'r tabl cynnwys

    9. Conclusions

    Although headline labour market indicators show that employment increased slightly in the three months to April 2020, some indicators are already showing the impact of the coronavirus (COVID-19) pandemic. The number of vacancies and average actual hours worked fell sharply in different industries. The job retention schemes in the UK look to have shielded the labour market from rising unemployment and increasing redundancies.

    This article shows that the impact of the pandemic has been most noticeable in the accommodation and food services industry where many labour market indicators deteriorated the most. It has shown that vacancies in some sectors reached their lowest levels in March to May 2020.

    This article has highlighted that most vacancies were lost in small and medium-sized enterprise (SME) firms and that SMEs contributed the largest proportion of workers who were on furlough. It compared the UK and US labour market outcomes and showed that average actual hours worked fell sharply in the UK but not in the US and that there was higher unemployment in the US than in the UK.

    The impact of the pandemic on the labour market is still evolving. The future performance of the labour market will be shaped by what happens when the government support schemes comes to an end and how quickly business confidence builds up for firms to bring more workers from furlough and to increase hiring.

    Nôl i'r tabl cynnwys

    10. Authors

    Abby Lucas, Marina Romiti, Ellie Dennison and Blessing Chiripanhura.

    Nôl i'r tabl cynnwys

    Manylion cyswllt ar gyfer y Erthygl

    Marina Romiti, Blessing Chiripanhura
    marina.romiti@ons.gov.uk, blessing.chiripanhura@ons.gov.uk
    Ffôn: +44 (0) 207 592 8698