Furlough: where did the hours go?

The economic fallout from Covid-19 is likely to significantly increase regional inequalities

As we stare down the barrel of a second Covid-19 wave, the government’s plan to end the Job Retention Scheme (JRS) in October is in the spotlight. There is widespread concern for how significant the resulting increase in unemployment could be and which regions and workers might be worst affected. When it comes to looking at the geographical uptake of furlough, the government states there was broad consistency in furlough rates across the nations and regions of the UK at the end of July (provisional data). London had the highest take-up rate of 17% against the UK average of 16%” (Figure 1). But we’ve found that this doesn’t tell the whole story.

With headline job market indicators like employment and unemployment distorted by the JRS, many people are using the rate of furlough to understand the impacts of Covid-19 on jobs. Rapid but less reliable indicators suggest the furlough rate had fallen to around 10% by early September. While the decline will be welcomed by some as signs of recovery, it is important to remember that the scheme isn’t as generous as it once was and many employers have turned to redundancies to balance their books.

A breakdown of furlough rates at the end of July 2020 shows limited regional variation, with the highest proportion of workers furloughed in London.

To understand what is happening beneath these numbers, NEF looked at Office for National Statistics (ONS) data relating to the number of hours actually worked across the economy. Worked-hours represents one of the truest’ indicators of economic activity and of the impact of the public health measures used to contain the virus. Our analysis suggests that at the peak of the crisis (April-June 2020) the total hours worked across the economy fell by around 19% compared to the same period last year. This is around 200 million lost working hours each week. However, this fall was far from evenly distributed.

In contrast with the furlough data, our analysis suggests that London experienced a far smaller economic slowdown than the other regions of the UK. For example, while the total number of hours worked collapsed by a little under a third (29%) in the east of England, total hours in London fell by just 12% (Figure 2). So what might be driving these trends, and the proportionately higher declines in hours worked in places like the East and the Midlands? And why does it look so different to the furlough rates?

Comparing total hours worked between April and June 2020 with the same period in 2019 shows a significant decline in hours which is distributed unevenly across the UK’s regions.

A crude analysis of the furlough scheme over the April-June period suggests that furlough only captured’ (ie paid for) around 150 million (75%) of lost hours. This leaves approximately 50 million hours (25%) unaccounted for. Where did they go?

Claimant count data (the measure of universal credit and job seekers allowance uptake), was where we first looked. The claimant count flags a group of people who have slipped through the cracks of the government’s crisis support schemes. While the proportion of people claiming universal credit has surged across the country, the rate of increase is significantly higher in London and the South (Figure 3). At first reading this output would seem contradictory with our earlier analysis. Why has London experienced the lowest fall in hours and such a large increase in claimants?

The rate of benefit claimants has increased significantly across all UK regions, but the rise has been significantly stronger in southern and eastern regions.

One theory relates to the types of worker being furloughed, and how many hours they previously worked. The data in Figure 4 is less reliable than our headline estimate on lost hours (Figure 2), as it infers regional changes from data on changes in economic sectors. However, what our analysis suggests is that the average number of hours previously worked by furloughed workers is significantly lower in London than the UK’s regions. In other words, London saw a large number of people furloughed who typically worked less weekly hours. By contrast, the Midlands saw a lower number of people furloughed, but before being furloughed they were working much longer hours.

So what can we take from this? We should be concerned for the low-hours, low-paid workers who have been driven to universal credit. But we should also be concerned that the proportional economic costs of this crisis may actually have been higher outside of London, where full-time workers have been furloughed in greater numbers.

Indicative analysis suggests the average total number of hours spent on furlough per-person varied between regions, and was lower in London during the peak of the crisis.

The message emerging from NEF’s analysis is that a premature end to the furlough scheme in October could have highly unequal impacts across the UK’s regions. Differences in the sectoral make-up of these regions, the types of contract typically found in them, pre-existing social deprivation, and the impacts of local lockdowns all must be considered in new policies.

As we prepare for what now looks like a protracted public health crisis and either an extension of some form of employment protection or mass unemployment, we must look for a revitalised package for jobs from government. NEF joins the Trades Union Congrees (TUC) in calling for a proper plan to create and protect jobs beyond October to prevent mass unemployment and deeper regional inequalities. This plan must have three key elements:

  • Create: a jobs-focused fiscal stimulus to create the low carbon jobs of the future. NEF estimates that at least 400,000 green infrastructure jobs could be created by the end of 2021 with a £28.3 billion green investment package. The TUC estimates that 600,000 jobs could be created across the public sector to plug current shortages across health, social care, education, and national and local government.
  • Protect: support for businesses – allowing them to bring people back to work for a shorter working week with government subsidising 70% of wages for hours not worked. Employers will make up the difference to ensure all workers receive 80% of pay for unworked hours, up to £2,500 a month. Similar support must also be put in place for the self-employed, including those who have missed out from previous schemes. Support in both cases would be targeted at those businesses who can show that their trading continues to be impacted by coronavirus restrictions.
  • Reskill: support workers to transition by offering those on shorter hours access to funded training to reskill and upskill during non-working hours. Funded training should also be made available to the unemployed, with everyone receiving a minimum income guarantee while they develop new skills.

It is imperative now that the UK formalises a reskilling programme for its workforce which can ensure that the working hours lost to coronavirus are not wasted. Not only can we emerge from this crisis as a higher-skilled, higher-paid, and more productive nation, but we can also tackle other crises like climate change and the low-carbon skills gap.

Create jobs, protect jobs, reskill workers

Change, in absolute terms, in total actual weekly hours worked in main and second job between Apr-Jun 2020 and Apr-Jun 2019

Region of place of work

Hours lost year-on-year (Apr-Jun 19 to Apr-Jun 20) rounded to nearest 100,000



Northern Ireland




Yorkshire and Humberside


East Midlands




West Midlands










East of England




Source: NEF analysis of ONS Labour Force Survey Data

If you value great public services, protecting the planet and reducing inequality, please support NEF today.

Make a one-off donation

£5 £10 £25 £50 £100

Make a monthly donation

£3 £5 £10 £25 £100