ILG0003
Written Evidence submitted by Dr Christoph Görtz, Dr Danny McGowan and Dr Mallory Yeromonahos.
1. The research underpinning this written evidence has been undertaken based on a UKRI/ESRC funded project (ES/V015958/1) on policy effects during the COVID-19 pandemic. The research has been conducted at the University of Birmingham and studies the effects of the Coronavirus Job Retention Scheme (CJRS) on household finances.
2. Our analysis uses probit regression analysis applied to survey data from the Understanding Society Database (USD), provided by the UK Data Service. This data set tracks over 16,000 households since 2010 and is representative of the UK population. We study the effects of the CJRS during the pandemic by using responses to a novel USD COVID-19 survey that comprises eight waves between April 2020 and April 2021.
3. Our study derives unique inferences about the CJRS and its effects on household finances. In particular, we establish how participation in the CJRS influences households’ finances and its effectiveness in preventing financial distress. Our findings show
● The CJRS was effective in preventing financial hardship for the vast majority of workers: only 3.8% of the UK workforce experiences financial pressure due to the CJRS.
● 61% of furloughed workers are not in financial distress.
● Increasing the generosity of CJRS payments would have had limited effects on protecting more workers from financial hardship, suggesting the policy was well designed.
● The CJRS was necessary to prevent large-scale household default.
How does the CJRS affect household finances and how effective is it in preventing financial distress?
4. 29.7% of eligible individuals are furloughed at some point during the pandemic. On average, a person is furloughed for 4.8 months and experiences a 16.54% reduction in their net monthly income, equivalent to £292.
5. The majority (61%) of furloughed individuals are not in severe financial difficulties. This share may decline in future if individuals are on the CJRS for a long time, because their ability to draw down savings diminishes the longer they are furloughed.
6. Being furloughed implies severe financial difficulties for two groups of people - these comprise 39% of all people on the CJRS.
Group 1: people whose hourly income declines below the Office for National Statistics’ (ONS) definition for low hourly pay due to being furloughed. 33% of furloughed workers receive an hourly compensation below the ONS low pay threshold and are likely to be in severe financial difficulties.
Group 2: people with little financial headroom prior to the pandemic who experience particularly large income reductions from being furloughed. Prior to being furloughed they typically have high expenses and low savings relative to their income. Furlough forces these individuals to either reduce spending or draw down savings. For those with particularly low financial headroom, these adjustments may not be sufficient to avoid financial distress. Figure 1 shows that the probability of late payments on mortgage, rent or other bills, increases when a person’s share of income lost from being furloughed rises. The average reduction in net monthly income from being furloughed is 16.5% (indicated by the red line in Figure 1) and the probability of late payments is 3%. In comparison, for individuals with a 50% income loss the probability of late payments triples to 9%. Individuals that faced an income loss of 50% or more face a probability of late payments of between 10% and 25%. We classify these individuals as being in severe financial difficulty - they comprise 6% of all furloughed people (Group 2).
Figure 1: Share of Income Lost due to Furlough and Probability of Late Mortgage, Rent or Bill Payments
7. The design of the CJRS protects the majority of furloughed workers from serious financial hardship. Owing to the scheme’s design, only 3.8% of the UK workforce falls into either of the two groups of people with severe financial difficulties. In comparison to the situation prior to the pandemic, the CJRS increased the overall share of people in the UK workforce with an income below the ONS low pay threshold (group 1) by two percentage points from 12.0% to 14.0%. People that are in financial difficulties due to the combination of a particularly large furlough-induced income loss and low financial headroom prior to the pandemic (i.e. Group 2 with income losses larger than 50%) comprise 1.8% of the UK workforce.
8. Higher CJRS payments would have had limited effects on preventing financial distress. Figure 1 shows that higher CJRS payments would have a limited effect on mitigating financial distress. Among furloughed individuals, the average monthly income reduction is 16.5% (indicated by the red line). At this point the probability of late payments is rather flat. For this reason, higher CJRS payments that shift the average income reduction leftwards to, say, 10%, would not substantially reduce the probability of late payments. The successfully targeted design of the CJRS is demonstrated by the fact that the vast majority of individuals face income reductions that are concentrated in the flat area of this curve. Figure 2 displays the distribution of furloughed workers and the income reduction they experience from the introduction of the CJRS. Notably, Figure 2 shows that the vast majority of individuals (87.5%) on the CJRS face an income reduction of less than 30%. This corresponds to the flat part of the curve in Figure 1 where reductions in income imply only very small changes in the probability of late payments. This evidence is specific to people in Group 2, i.e. people who ended up in financial distress due to the combination of high expenses and low savings relative to their income loss. It indicates that higher CJRS payments would have had limited effects on reducing the number of people in Group 2.
Higher CJRS payments would have helped to reduce the share of people in financial difficulties due to an income below the ONS low wage pay threshold (Group 1). However, as highlighted in point 7 above, the introduction of the CJRS increased their overall share of the workforce by only 2 percentage points.
Figure 2: Distribution of Furloughed Workers and their Income Reduction from the CJRS
9. Other government policies mitigate the adverse effects from being furloughed. Mortgage and rent payments tend to be the largest single expense item for individuals. Being furloughed increases the probability of late payments on rent by 4.1% percentage points to 7.4%. For individuals with a mortgage, the effect is less pronounced. The probability of late mortgage payments rises only by 0.7 percentage points to 3.0%. The reason is that many mortgage holders use the Mortgage Holiday Scheme to relax potential financial constraints. Being placed on furlough more than doubles the probability that a person takes a mortgage holiday from 8.9% to 19.0%. Hence, the salary loss and uncertainty surrounding future income from being furloughed made it more than twice as likely for individuals to opt into the Mortgage Holiday Scheme to relax financial constraints.
10. The CJRS was necessary to prevent large scale household default. What would have happened in a situation without the CJRS, where furloughed people would have instead been on unemployment benefit? We conservatively assume that people would have received the maximum unemployment benefit. For the time they had been furloughed, they would have had to use savings to cover the difference between unemployment benefits and the income they received from the CJRS.
45.7% of furloughed individuals could not have covered the shortfall at all, since they do not have any monthly savings prior to the pandemic. They would be in substantial financial difficulties without the CJRS.
Among the other 54.3% of individuals with positive monthly savings, half saved less than 13.4% of their take home pay prior to the pandemic. As described in point 4 above, the average person is furloughed for 4.8 months. For this time, in a situation without the CJRS, the average person would have had to use £4,779 from pre-pandemic savings to cover the difference between the unemployment benefits and the income they received from the CJRS. They would have needed 22.7 months prior to the pandemic to accumulate these savings.
Figure 3: Distribution of Months Required to Accumulate Savings to Cover the Shortfall between Furlough Payments and Hypothetical Unemployment Benefits
Figure 3 shows the distribution of months required to accumulate savings to cover the shortfall between total furlough payments received and hypothetical unemployment benefits. It is evident from this figure that in absence of the CJRS, a large number of people would have required several years of savings to cover the income shortfall from being unemployed. 40% of furloughed people with pre-pandemic savings would have required more than two years of savings; 13% of people would have needed to accumulate savings for more than 5 years. It is unlikely most individuals would have had sufficient savings. In the absence of the CJRS, not only furloughed individuals without savings, but also a large number of individuals with low savings before the pandemic would have experienced severe financial difficulties and household default.
Characteristics of people on the CJRS and of furloughed individuals in financial difficulties.
11. Job type, education and gender are important factors in determining whether a person is furloughed. Workers in routine or intermediate jobs are 9.9 percentage points more likely to be furloughed relative to a person employed in a managerial position. Individuals with an educational qualification below a university degree face a 14.1 percentage points higher likelihood of being furloughed compared to those with a university degree. Men are 3.1 percentage points more likely to be furloughed than women.
12. People with higher pre-pandemic income levels are less likely to be furloughed. A £1,000 increase in net monthly income reduces the probability a worker is furloughed by 0.08 percentage points, conditional on their job type, education level and gender.
13. Ethnicity and geographical location have, on average, no influence on whether an individual is furloughed.
14. Who on the CJRS is unable to pay their mortgage, rent or household bills? Household savings are key to buffering the adverse effects of being furloughed. Furloughed individuals who are late on payments tend to have lower pre-pandemic savings to household income ratios. The pre-pandemic savings rate was 6.03% for those who are late on payments, compared to 9.31% for furloughed workers who are current on their payments.
Among furloughed individuals, people with an educational qualification below a university degree are approximately 6.70% percentage points more likely to be late on payments, conditional on their pre-pandemic savings, gender, ethnicity, income and whether they are in a couple. On average, ethnicity, gender and geographical location do not affect the likelihood that a furloughed worker is late on payments.
June 2021