Post-furlough blues

Going into the autumn, most official forecasters expected unemployment to be little affected by the ending of the Coronavirus Job Retention Scheme (JRS), which closed in September 2021. However, there were still 1.1 million workers on furlough as the scheme ended, skewed towards sectors like aviation that are still far from reaching pre-pandemic levels of activity, and increasingly skewed to older workers, who tend to struggle finding work after periods of worklessness. Understanding what has happened to those furloughed workers is key to assessing the health of the labour market this autumn. This spotlight presents the first evidence based on data collected from workers after the JRS closed, supplementing what is already known about furloughed workers’ expectations (measured in September), and businesses’ assessment of their furloughed employees’ outcomes (measured in early October).

Our survey suggests that 88 per cent of respondents who were furloughed in September were in work in October, with 8.5 per cent being inactive and just 3.4 per cent being unemployed. If this is representative of the population, this implies around 136,000 of the 1.1 million on furlough in September were not working in October. Those furloughed in September had worse outcomes than those employed normally in September, with furloughed workers six times more likely to be out of work in October than those employed normally, even after accounting for factors like age and sector. This is driven by fully furloughed workers, as partially furloughed workers were no more likely than those who were regularly employed in September to be out of work by early October.

If confirmed by official statistics, this would be a good result for the UK labour market – albeit not for the 136,000 former workers – showing the value of continuing the JRS several months after restrictions on economic activity had been lifted. But it also highlights that policy makers should focus a little less on headline rates of unemployment, and instead focus on the elevated rates of economic inactivity, and on wider challenges to the strength of the recovery through the winter, as seen in the disappointing growth rates for the third quarter of 2021.

The UK’s labour market is in a strong position, at least when it comes to unemployment. Last month, the Office for Budget Responsibility (OBR) forecasted unemployment to peak at 5.2 per cent following the end of the JRS, down from a peak of 6.5 per cent forecast in March (see Figure 1). Improved forecasts have been led by a more positive set of labour market indicators than in previous forecasts: unemployment has been falling throughout 2021, buoyed by record job vacancies, and planned redundancies were falling over the summer even as firms faced the imminent end of the JRS.

Figure 1: The OBR has hugely revised down its forecast for unemployment

Still, as the scheme came to end in September 2021, around 1.1 million workers (about 4 per cent of the workforce) were either fully or partially furloughed (see Figure 2), with rates of furlough remaining very high in sectors such as passenger air transport (36 per cent) and travel agencies (35 per cent). In addition, the profile of those remaining on furlough became increasingly skewed towards older workers in the final few months of the scheme (rather than being concentrated on younger workers, as was the case during the various lockdowns): the share of employees aged 60-64 (5 per cent) and 65 and older (7 per cent) on furlough in September were both larger than the share of employees aged under 35 (3 per cent). This is worrying given that older workers tend to take longer to return to work than their younger counterparts, and they often return on lower pay.

Figure 2: At the end of the furlough scheme 1.1 million workers were furloughed

The biggest uncertainty about the UK labour market this autumn is the fate of those who were furloughed at the end of September. Similar to the OBR’s recent forecast, the Bank of England is also expecting only a small rise in unemployment this quarter, with the Governor Andrew Bailey saying “while the unemployment rate is projected to pick up in Q4 as the furlough scheme ends, the rise is only small as most employees on the scheme are assumed to return to work.” (see also the Bank’s November Monetary Policy Report press conference). This assessment was based in part on the finding that around 10 per cent of furloughed respondents to the Bank of England/ NMG survey expected to lose their job at the end of the scheme.

There is already some indirect evidence about what happened post-furlough. Research by Indeed shows only a small increase in the proportion of workers who are urgently searching for a new job after the end of furlough. Evidence from the Office for National Statistics Business Insights and Conditions Survey (BICS) suggests that 3 per cent of workers left voluntarily and 3 per cent were made redundant (see Figure 3). Although the BICS study has proven to be an accurate guide to the number of furloughed employees throughout this crisis (as seen in Figure 2), this number is extremely low, and may reflect that the survey is completed by businesses, rather than employees.

Figure 3: Business survey estimates suggest that between 3 and 6 per cent of those furloughed in September were not working in October

To help fill this gap, we fielded a survey in October that asked people about their current and previous labour market statuses.1 This is the first survey that we know of that asks employees themselves about their post-furlough labour market outcomes.

In line with expectations, our survey does not show a big rise in unemployment: it suggests that the rate of unemployment among working-age adults was just 0.5 percentage points higher in October than it was in August 2021. Looking directly at respondents who were furloughed in September, our survey finds that the vast majority (88 per cent) were in work in the first half of October (see Figure 4), leaving 11 per cent who were not in work (3.4 per cent were unemployed, and a further 8.5 per cent were not active in the labour market); this is broadly consistent with the Bank of England/ NMG survey of households in September that showed around 10 per cent of furloughed respondents expected to lose their job at the end of the scheme.

Applied to the 1.1 million we know were furloughed at the end of September, then that would mean 136,000 of those furloughed in September were not working in the first half of October (with a small rise in unemployment of around 39,000, and a larger rise in inactivity of 97,000, something that matches the wider trend of rising inactivity over the course of this crisis).

Figure 4: Our survey of working-age adults suggests that around 136,000 of those furloughed in September were not working in October

Part of the reason why a relatively high fraction of those furloughed in September were in work in October is that some had already found another job: estimates from the four surveys we have fielded during the pandemic (shown in Figure 5) show that around 15 per cent of workers on furlough since June 2021 were furloughed in one job and working in another.2

But the more important reason is that post-furlough outcomes vary considerably between partially and fully furloughed workers. Regression analysis shows that, even after controlling for age, education, industry group and pre-crisis pay, those furloughed in September had worse outcomes in October than those employed normally in September, with furloughed workers six times more likely to be out of work than those employed normally.3 This result is being driven entirely by fully furloughed workers, with partially furloughed workers – who comprised 44 per cent of the total on furlough at the end of September – no more likely than those employed normally in September to have lost their job in October.

Figure 5: Around 15 per cent of furloughed workers have had another job since June 2021

Overall, it appears that the end of the furlough scheme has brought poor labour market outcomes for some (fully) furloughed workers, particularly as this coincides with the cut to core levels of Universal Credit by £20 a week. For the labour market as a whole, though, the impact is small, chiefly reflecting the decision taken by the Chancellor in March to continue the JRS well after all restrictions on economic activity had been lifted.

But there are two important considerations for policy makers. First, much of the focus on worklessness during this crisis has been on unemployment, with those leaving the labour force entirely and becoming inactive receiving much less attention. It seems that furloughed workers who have lost work have moved mostly into inactivity; elsewhere, Resolution Foundation has highlighted that economic inactivity has increased for groups like the retired and students. So inactivity – rather than unemployment – is what policy makers need to focus on. Second, having deliberately focused attention on post-furlough outcomes when presenting last week’s Monetary Policy Report, the Bank of England should instead be more mindful of wider challenges to the strength of the recovery this winter. Last week’s GDP data, in particular, showed that combined growth in July, August and September was no better than its pre-financial crisis average, suggesting that supply-chain disruptions and shortages have derailed the recovery.

______________________________________________________________________________________________________________________
1 The survey was designed and commissioned by the Resolution Foundation, in partnership with the Health Foundation (although the views in this note are not necessarily those of the Heath Foundation). It was conducted using an online interview administered to members of the YouGov Plc UK panel, which is made up of 800,000+ individuals who have agreed to take part in surveys. The total sample size was 6,100 adults, aged 18-65 and fieldwork was undertaken during 14-25 October 2021. The figures presented here have been analysed independently by the Resolution Foundation and are not the views of YouGov.
2 Neither the HMRC’s administrative data nor the BICS survey is able to provide an equivalent estimate of the fraction of furloughed employees who had another job.
3 Predicted out of work (either unemployed or inactive) rates are based on the results of logistic regressions looking at furlough status, holding constant the following personal and job characteristics: age, industry, education and pre-crisis pay. Having controlled for these, the underlying risks were 6.6 per cent for furloughed workers and 1.1 per cent for regular employed workers.