Household debt and Covid

Overview

Covid has had an unprecedented impact on the UK economy. And despite a huge degree of public policy support, the crisis has hit incomes and employment opportunities.

In the past, households with high levels of debt have amplified shocks to the economy by cutting back on spending. The latest evidence suggests that during the global financial crisis countries with larger initial debt relative to income saw larger falls in consumption, putting downward pressure on economic activity (see the December 2019 Financial Stability Report for more information).

But, so far at least, there is little evidence that household debt has amplified the Covid recession. There are likely to be several reasons for this.

Unprecedented public policy support, in particular the Coronavirus Job Retention Scheme, has supported household incomes through the pandemic. Social distancing measures have also meant that many households were able to increase their savings, as they were effectively forced to reduce their spending on certain categories of goods and services, regardless of their level of debt. Temporarily, at least, this is likely to have increased their resilience to future shocks. In line with Financial Conduct Authority (FCA) guidance, lenders have also supported households by offering temporary deferrals of debt repayments, also known as ‘payment holidays’.

UK households also entered this shock in a stronger financial position than before the global financial crisis. In part, this is a result of post financial crisis regulatory reform. The Financial Policy Committee’s (FPC) mortgage market Recommendations have helped to guard against a significant increase in the number of households with high debt burdens. And the banking system is better capitalised than before the global financial crisis, enabling lenders to offer more support to households.

But the share of households reporting being in financial difficulty has started to increase, particularly for households with unsecured loans – who tend to have lower incomes and are less likely to be in employment. These households are also less likely to have savings and have been less able to accumulate savings through the Covid crisis. This may make this subset of households more vulnerable to future shocks.

The extent to which household debt may yet play a bigger role in the Covid crisis will depend on the evolution of the pandemic and the resulting pace of economic recovery, which remain uncertain. It will also depend on how governments, households, businesses and financial markets respond to these developments.

Introduction

The Covid crisis has been unique in the UK’s recent history. The spread of Covid and the actions taken to contain it have had a large impact on UK households. The lockdown measures put in place by the Government brought cities and towns to a halt, and pressed pause on entire industries. Despite a huge degree of public policy support, the crisis has hit incomes and employment opportunities.

In the past, households with high levels of debt have amplified shocks to the economy. This is because households struggling with high debt repayments or concerned about their future financial situation may make drastic cuts to their spending to avoid defaulting on their debt or to bolster their savings. And, in the case of default this can lead to losses for lenders, reducing their ability to supply credit to the wider economy. This can amplify an economic downturn and also pose risks to financial stability.

This article explores whether household debt has amplified the impact of Covid on the economy, and whether it might in future.

To answer this question we examine the impact of the shock across different households, how they responded, and what implications this has had for household debt burdens so far. To do this, we complement aggregate statistics with several household surveys (Box A) collected throughout the crisis.

Household debt and financial stability

Historically, household debt has amplified economic shocks due to the pressure it puts on household finances.
Households with debt may be more vulnerable when faced with an income shock, as the drop in income can make it difficult to manage regular repayments as well as their ongoing spending. There are two channels through which this can amplify an economic downturn, and in turn pose risks to financial stability:

• Households with high levels of debt relative to income may cut consumption by more, amplifying any economic downturn. Evidence from the global financial crisis showed that households with higher pre-crisis debt relative to their incomes made sharper adjustments to their spending (see the December 2019 Financial Stability Report for more information). This may be because they struggle to continue making debt repayments, or become more concerned about their future financial situation and their ability to borrow. This is known as the ‘borrower resilience’ channel.
• Households with high levels of debt may also be more likely to default on their debt, resulting in losses for lenders. This could compromise the ability of lenders to keep supplying credit to the wider economy, further amplifying any downturn. This is known as the ‘lender resilience’ channel. This is likely to be more material for unsecured debt, as borrowers are more likely to default on unsecured debt and lenders do not have collateral to cushion losses.

Both these channels can amplify an economic downturn. They can also interact with each other, creating negative feedback loops. For example, a rise in defaults that results in losses for lenders would likely also result in a tightening of credit conditions. This could make it more difficult for households to borrow to smooth consumption, exacerbating the impact on the real economy. Similarly, if households are unable to borrow, this can affect the housing market, causing a drop in prices and worsening losses for lenders in the case of default.

Evidence suggests a strong link between household debt and the size of consumer spending cuts in a downturn. During the global financial crisis, countries that had higher levels of household debt relative to income initially saw larger falls in aggregate consumption. Analysis of household-level data suggested that individual households with higher mortgage debt relative to income adjusted spending more sharply in response to shocks.^{footnote [2]}

Households with higher mortgage and unsecured debt-servicing costs relative to income are also more likely to go into arrears. Indeed, historically, there has been a strong correlation between unemployment and consumer credit loss rates.^{footnote [3]}

Households entered the Covid pandemic in a stronger financial position than before the global financial crisis.
There are reasons to believe that household debt may play less of a role in the Covid crisis. UK households entered this crisis in a stronger financial position than before the global financial crisis. The total stock of UK household debt (excluding student loans) was equivalent to around 123% of total household income, materially below its 2008 peak of around 145%.^{footnote [4]} Of this, mortgage debt accounts for around four fifths and consumer credit about a fifth of the total.

Interest rates are significantly lower than in 2008, meaning that borrowers had lower debt-servicing costs on average compared to before the financial crisis. Post financial crisis reforms, such as the FPC’s mortgage market Recommendations, put in place in 2014, have also helped to guard against a significant increase in the number of highly indebted households.

The UK banking system is also better capitalised than before the global financial crisis.^{footnote [5]} This has helped lenders to avoid any large contractions in the supply of credit. This is in contrast to the ‘credit crunch’ experienced in 2007–08. And has meant that, alongside government policy support, UK banks and building societies have continued to supply households with additional credit to support their finances, as well as offering temporary breaks on debt payments (known as payment deferrals or ‘payment holidays’).

Impact of Covid on households’ income and savings

Policy support has helped households manage their finances through the Covid crisis.
The first national lockdown in March 2020 led to a record 13.5% fall in GDP in 2020 Q2, as spending in the economy dropped dramatically. But the policy response has meant that not all of this impact has fed through to households’ incomes. Government measures, such as the Coronavirus Job Retention Scheme (CJRS) and the Coronavirus Self-Employment Income Support Scheme (SEISS), have supported incomes and employment over the last year, particularly for lower income households. A cumulative total of 11.5 million jobs have been supported by the CJRS since it began, while the SEISS supported 2.8 million people as of May 2021.

Thanks to this support, the fall in households’ gross disposable income was limited to around 3% (Chart 1), significantly less than the fall in GDP. But lockdown restrictions meant that the fall in aggregate spending was significantly larger, at around 20%. Consistent with this, the household savings ratio rose sharply, peaking at over 25% in 2020 Q2 – around four times higher than before the global financial crisis.

Chart 1: Household consumption fell by a lot more than household income

Household gross disposable income, individual consumption expenditure and the household savings ratio 2007 Q1 to 2020 Q4 (a) (b) (c) (d)

Footnotes

• Sources: ONS UK Quarterly Sector Accounts and Bank calculations.
• (a) The global financial crisis ran from 2008 Q3 to 2009 Q4.
• (b) The period shaded dark purple/medium purple represents where a lockdown was in place. The first lockdown ran from late March to the end of June 2020 and the second lockdown ran from the start of November to early December 2020. See more information on lockdown timelines.
• (c) The period shaded pale purple covers the period in which social distancing restrictions were in place.
• (d) The household savings ratio measures household savings as a proportion of household income.

But the impact on incomes and finances has varied across households.
Despite the unprecedented levels of support in place, at least a quarter of households experienced a fall in income because of Covid in April 2020. The initial hit to incomes was fairly broad-based (Chart 2), with the furlough scheme and increases in universal credit providing greater support to those on lower incomes. However, the impact has generally been more persistent for lower-income households, who were less likely to see a recovery in incomes through 2020 and 2021. The impact on the self-employed, particularly those who did not meet the criteria for the SEISS, has been particularly marked.

The impact on savings has been much more varied across households. Middle and higher-income households have in general accumulated savings through the pandemic, as their spending was likely to have fallen by more than their income (Charts 3 and 4). This is likely to be driven by cuts to discretionary spending due to social distancing. But lower income households – who are likely to have a higher share of essential spending – were less likely to have built up savings during the pandemic.

Chart 2: The impact of Covid has been large and uneven across households, varying based on income and job characteristics

Net balance of households reporting a change in income across income quintiles and employment status, reported in April and August/September 2020 and March 2021 (a) (b)

Footnotes

• Sources: NMG Consulting survey and Bank calculations.
• (a) In April and August/September 2020 and March 2021, respondents were asked: `How has your household income changed relative to usual because of coronavirus?’. Respondents could report different levels of increase or decrease, or report no change in income.
• (b) Net percentage balances are calculated as the percentage of households that reported an increase in income minus the percentage of households that reported a decrease in income. A negative net balance implies that more people are reporting falls in income than rises.

Chart 3: Middle and higher-income households were more likely to cut back spending

Net balance of households reporting a change in spending across income quintiles, reported in April and August/September 2020 and March 2021 (a) (b)

Footnotes

• Sources: NMG Consulting survey and Bank calculations.
• (a) In April 2020 and August/September 2020 and March 2021, respondents were asked: `Thinking about the past month, how has your total household spending differed from what you would have usually spent before the coronavirus pandemic?’. Respondents could report different levels of increase or decrease, or report no change in spending.
• (b) Net percentage balances are calculated as the percentage of households that reported an increase in spending minus the percentage of households that reported a decrease in spending. A negative net balance implies that more people are reporting falls in spending than rises.

Chart 4: Middle to higher-income households were more likely to increase their savings

Net balance of households reporting a change in savings across income quintiles, reported in August/September 2020 and March 2021 (a) (b) (c)

Footnotes

• Sources: NMG Consulting survey and Bank calculations.
• (a) In August/September 2020 and March 2021, respondents were asked: `As a result of changes in income or spending due to the coronavirus pandemic, would you say that your household’s savings have increased, decreased, or stayed the same?’. Respondents could report different levels of increase or decrease, or report no change in savings/no pre-existing savings.
• (b) Net percentage balances are calculated as the percentage of households that reported an increase in savings minus the percentage of households that reported a decrease in savings. A negative net balance implies that more people are reporting falls in savings than rises.
• (c) This question was not asked in the April 2020 NMG survey.

The impact of Covid on household debt burdens

Households with unsecured credit have generally been more exposed to the Covid crisis.
As set out above, the impact of Covid has varied a lot across households. In general, wealthier households have tended to accumulate savings and may have been able to pay down debt should they have wished to. Lower to middle-income households have seen their income fall more persistently and may have needed to take on additional debt to manage spending commitments.

By looking at the characteristics of those most affected by the crisis, and how they compare to the characteristics of households with different types of debt,^{footnote [10]} we can identify where risks to financial stability might develop. In general, mortgage borrowers tend to have higher incomes compared to households with only unsecured debt, and are more likely to be in employment (Chart 5).

Households with unsecured debt, also known as consumer credit, are more likely to have experienced a drop in income. This is because hits to income have been more concentrated among those with lower incomes or in less secure jobs. This difference was larger earlier in the pandemic, with the gap narrowing through 2020 as economic activity started to recover (Chart 6).

Chart 5: Mortgage borrowers are more likely to have higher household incomes and be employed, compared to those with only consumer credit

Distribution of households by income and employment status pre-Covid (2019 H2) (a) (b) (c)

Footnotes

• Sources: NMG Consulting survey and Bank calculations.
• (a) ‘Inactive’ includes students and retirees. ‘Other’ includes those not in work due to a long term illness, disability or other reason.
• (b) Income defined as total pre-tax household income.
• (c) Employment defined by the working status of the individual providing the response to the survey.

Chart 6: More consumer credit borrowers faced a fall in income compared to mortgage borrowers

Net balance of households reporting a change in income across households with different types of debt, reported in April and August/September 2020 and March 2021 (a) (b)

Footnotes

• Sources: NMG Consulting survey and Bank calculations.
• (a) In April and August/September 2020 and March 2021, respondents were asked: `How has your household income changed relative to usual because of coronavirus?’. Respondents could report different levels of increase or decrease, or report no change in income.
• (b) Net percentage balances are calculated as the percentage of households that reported an increase in income minus the percentage of households that reported a decrease in income. A negative net balance implies that more people are reporting falls in income than rises.

So far, we have not seen a large increase in the number of households with a high debt burden.
Despite the fall in incomes, survey evidence suggests that, so far, the pandemic has had a limited impact on the number of households with a high debt-servicing ratio (DSR), measured as debt repayments as a percentage of income. The Government’s policy response to support incomes and the payment deferrals offered by lenders have reduced the pressure on household finances. In turn, this has reduced the risk of households falling into arrears or needing to sharply cut their spending.

The share of high DSR households with mortgage debt is likely to have picked up at the start of the crisis, but has remained fairly stable since and remained well below historic highs (Chart 7).^{footnote [11]}

And a similar pattern is seen with unsecured debt, despite these households being more exposed to the crisis. Although the share of households with high consumer credit debt burdens did increase slightly in 2021 H1, this only represents a return to levels last seen in 2019. Reduced consumer spending and a decline in credit card balances due to lockdown restrictions will have helped offset the drop in income. And many households may also have used their accumulated savings to pay down balances on credit cards.

Chart 7: Mortgage DSRs peaked in April/May 2020 driven by a fall in mortgage borrower income, but have remained stable since

Percentage of households with mortgage DSRs at or above 40% (a) (b) (c) (d)

Footnotes

• Sources: British Household Panel Survey (BHPS)/Understanding Society (US), NMG Consulting survey and Bank calculations.
• (a) Percentage of households with mortgage DSR at or above 40% calculated using BHPS (1991–2009), US (2009–19), and the online waves of NMG Consulting survey (2011–21). Mortgage DSR calculated as total mortgage payments as a percentage of total household pre-tax income.
• (b) US is built on the BHPS, which ran from 1991–2009 and included around 10,000 households. US includes around 8,000 of the original BHPS households. From April 2020, participants from the main survey were asked to complete the more frequent Covid survey to capture experiences during the pandemic. See footnote six and footnote eight for full citation.
• (c) The US special Covid survey did not include questions on mortgage repayments or total household pre-tax income, therefore estimates are calculated by adjusting mortgage DSRs reported in Wave 9 of the main survey for the change in total household after tax earnings since January/February 2020, reported in Waves 1–7 of the US Covid survey.
• (d) A new household income question was introduced in the NMG survey in 2015. Adjustments have been made to data from previous waves to produce a consistent time series.

How households responded to the shock

There are a number of ways that households with debt might respond to increased pressure on their finances. This includes cutting back on spending, drawing down savings, or falling into arrears on debt or other household bills. Further cuts in spending, particularly if severe, could amplify the downturn or drag on the recovery and present risks to financial stability.

So far, there is no evidence that higher levels of mortgage debt have amplified spending cuts during the crisis.
Our survey evidence suggests that, in response to a fall in income, those with unsecured debt were more likely to use savings through the crisis or take on additional debt, whereas those with mortgages were more likely to cut their spending (Chart 8). This may reflect the fact that mortgage borrowers are concentrated in higher income brackets, and are therefore likely to have a higher share of discretionary spending, which was more restricted by lockdown measures.

Chart 8: Consumer credit borrowers were more likely to report having used savings in response to an income fall, compared to those with mortgage debt

How borrowers dealt with income loss since January/February 2020, by debt type (a) (b) (c)

Footnotes

• Sources: Understanding Society survey and Bank calculations.
• (a) Based on responses to Wave 4 of the Understanding Society Covid survey, conducted in July 2020. See footnote six for full citation.
• (b) Question: ‘If your household is now earning less than in January/February 2020, have you done any of the following to deal with this?’. Respondents were able to select more than one response, including other responses in addition to those shown in the chart. These included ‘accessed pension or reduced pension contribution’; ‘found new work/increased hours’; ‘another member of my household found new work or increased hours’; ‘new or increased welfare benefits’; ‘dealt with earnings loss in another way’; and ‘none of these/does not apply’.
• (c) Calculations are based on all responses including those not shown on the chart other than ‘none of these/does not apply’. Shares sum to over 100% for each debt type as respondents were able to select more than one response.

In past crises, cuts in spending by indebted households were a cause for concern. Evidence from the financial crisis shows that UK households with high pre-crisis mortgage debt relative to their incomes were around three times more likely to cut back on consumption compared to those with low debt relative to their income.

But so far there is little evidence that households with high mortgage debt relative to their income were more likely to cut back on spending compared to those with lower debt to income (Chart 9). In part, this is likely to reflect that lockdown measures have restricted spending regardless of pre-existing levels of debt, making it quite distinct from the experience of the global financial crisis. And payment deferral schemes mean that some households with debt haven’t faced the same dilemma of missing repayments or cutting their spending.

Chart 9: Mortgage borrowers with high and low loan to income ratios reported a reduction in spending

Net balance of households reporting a change in spending due to Covid, reported in April and August/September 2020 and March 2021 (a) (b) (c)

Footnotes

• Sources: NMG Survey and Bank calculations.
• (a) In April and August/September 2020 and March 2021, respondents were asked: `Thinking about the past month, how has your total household spending differed from what you would have usually spent before the coronavirus pandemic?’. Respondents could report different levels of increase or decrease, or report no change in spending.
• (b) Net percentage balances are calculated as the percentage of households that reported an increase in spending minus the percentage of households that reported a decrease in spending. A negative net balance implies that more people are reporting falls in spending than rises.
• (c) Mortgage loan to income (LTI) calculated as ratio of outstanding mortgage debt to pre-tax annual income. For example, an LTI of three, means a household reported an outstanding mortgage debt of three times their current pre-tax income.

The option to defer regular loan repayments has also helped households with mortgages avoid sharp spending cuts.
Payment deferral schemes, also known as ‘payment holidays’ offered by lenders following FCA guidance, have provided significant support to many borrowers. Since March 2020, the schemes have allowed borrowers to temporarily freeze mortgage and unsecured loan repayments, and so have provided a cushion against any fall in income.

Though payment deferrals were available to borrowers with either mortgage debt or unsecured debt, they appear to have had more of an impact in supporting those with mortgages. UK Finance figures suggest that around one in six UK mortgages was on a payment deferral around their peak in June 2020. In contrast, survey evidence and supervisory intelligence suggest that a smaller proportion of unsecured borrowers opted to defer payments. Interest accumulates over a payment deferral, and since unsecured debt typically has a much higher interest rate, this may make payment deferrals a less attractive option for this type of debt.

Payment deferrals have helped to reduce pressure on household finances. Mortgage borrowers with payment deferrals were less likely to report a cut in spending, despite facing a slightly larger fall in income (Chart 11). The widespread use of mortgage deferrals in particular, has helped borrowers to manage temporary changes in their income through the crisis.

However, survey evidence also suggests that many payment deferrals at the onset of the crisis may have been taken for precautionary reasons, as about a third of households who benefited did not experience a fall in income (Chart 10). Consistent with this, the vast majority of those that took out a payment deferral have since resumed full or partial repayments.

Chart 10: Many payment deferrals appear to have been taken out on a precautionary basis

Percentage of households taking out payment deferrals, by income change in April or May 2020 (a) (b)

Footnotes

• Sources: British Household Panel Survey (BHPS)/Understanding Society (US) and Bank calculations.
• (a) Based on Wave 1 and Wave 2 of the US Covid survey, conducted in April and May 2020 respectively. See footnote six and footnote eight for full citation.
• (b) Change in any income is calculated as the change in total household after tax earnings since January/February 2020 and the month in which the payment deferral was granted (April/May), ie relative to pre-Covid.

Chart 11: Mortgage borrowers with payment deferrals were less likely to cut spending

Net balance of households reporting a change in income and spending due to Covid, reported in April and August/September 2020 (a) (b) (c) (d)

Footnotes

• Sources: NMG Consulting survey and Bank calculations.
• (a) In April and August/September 2020 and March 2021, respondents were asked: `How has your household income changed relative to usual because of coronavirus?’. Respondents could report different levels of increase or decrease, or report no change in income.
• (b) In April and August/September 2020 and March 2021, respondents were asked: `Thinking about the past month, how has your total household spending differed from what you would have usually spent before the coronavirus pandemic?’. Respondents could report different levels of increase or decrease, or report no change in spending.
• (c) Net percentage balances are calculated as the percentage of households that reported an increase in income/spending minus the percentage of households that reported a decrease in income/spending. A negative net balance implies that more people are reporting falls in income/spending than rises.
• (d) In April and August/September 2020, respondents were asked whether they had taken out a mortgage payment holiday. Questions varied slightly across 2020 H1 and H2, though both try to capture the number of respondents that were currently or previously on a payment holiday, or had taken a payment holiday extension. We include all respondents that confirmed any of these options.

But many households still face financial difficulty, especially those with unsecured debt.
Despite the various support measures, more households are now reporting being in financial difficulty than at the start of the crisis. Around one in five households have reported experiencing some kind of financial difficulty as a result of the Covid crisis, and around one in ten households with some kind of debt report being very concerned about the current level of this debt. And while incomes have partially recovered since the initial fall in the first half of 2020, subjective vulnerability metrics^{footnote [12]} have started to increase, albeit only moderately.

Those with unsecured debt have been significantly more likely to report being in financial difficulty relative to those without, and this has increased over time (Chart 12). This is consistent with households with unsecured debt facing larger hits to incomes, and being less likely to accumulate savings through the crisis.

Survey evidence suggests that the share of borrowers falling behind on debt repayments has increased since the pandemic, but remains relatively low as government schemes continue to support households. And this is reflected in the fact that the proportion of mortgage balances in arrears has remained relatively stable since the start of the pandemic.

Chart 12: Those with consumer credit debt were more likely to report being in financial difficulty as a result of the Covid crisis

Share of respondents reporting financial difficulty as a result of Covid, by debt type (a) (b) (c) (d)

Footnotes

• Sources: NMG Consulting survey and Bank calculations.
• (a) In April and August/September 2020 and March 2021 respondents were asked: `Have you found yourself in any type of financial difficulty as a result of the outbreak of coronavirus?’. With possible responses including different types of financial difficulty.
• (b) Across those with mortgage debt, borrowers that also had consumer credit debt were more likely to report being in financial difficulty compared to those that only had mortgage debt.
• (c) Excludes respondents that indicated `Don’t know` or `Prefer not to say`.
• (d) There are a roughly similar number of respondents in each of the debt type categories with around a third of respondents each.

Conclusion

Covid and the measures to address it have had a large impact on households’ incomes and spending. And some households have been particularly badly hit. But we have also seen unprecedented levels of policy support through this crisis, including measures to support incomes and the option to defer repayments on loans. And households entered this crisis in a much stronger financial position than before the global financial crisis.

As a result there is currently little evidence that household debt has amplified the Covid recession. Many households have been able to weather the worst of the Covid crisis reasonably well. But the share of households struggling with their finances has increased, particularly among those with unsecured loans – who tend to have lower incomes and are less likely to be in employment. This group is likely to be one of the more vulnerable in the next phase of this crisis.

Looking ahead, the extent to which household debt may yet play a bigger role in the Covid crisis will in part depend on the evolution of the pandemic and the resulting pace of economic recovery, which remain uncertain. It will also depend on how governments, households, businesses and financial markets respond to these developments.