Low levels of savings have made it hard for households in developing countries to adapt to reduced incomes caused by Covid-19. Emergency income support could help to alleviate hardship, but emerging economies struggle to finance such schemes.
Around the world, Covid-19 has exposed the fragility of households’ economic and financial wellbeing. Policies implemented to contain the pandemic have restricted vast numbers of individuals from being able to earn, and this has threatened households’ economic security. Policy-makers have had to balance the need to limit the spread of the virus with economic considerations. These include how long the average household can cope with lockdown-induced disruptions to income and what fraction of households can draw on savings before they need to access government support.
Our evidence suggests that a large part of the world’s population have insufficient savings to weather an income shock. Indeed, financing an income shortfall for a period of three months would lead to a complete depletion of liquid financial assets for a third of households in our global sample. To determine effective policy, understanding the resilience of households to severe income shocks remains essential for any future economic disruptions, such as those resulting from the looming climate crisis.
One important policy instrument during such emergencies is the direct financial support to households. But the extent to which countries can provide this type of support varies, not least because of the differences in their capacity to take on debt to finance such a scheme.
While advanced economies usually have the ability to access global capital markets quickly and at a reasonable cost, emerging economies are seldom permitted this luxury and face higher rates on loans. For example, while Germany, Switzerland and the UK can borrow virtually at zero cost, the rates on government debt reach 6% per year for India, 7% for Mexico and 9% for Brazil (Bloomberg, 2021). This ‘credit constraint’ can particularly limit the space for government action in emerging economies (Kose et al, 2017).
But even if governments can access capital quickly and at scale, without a clear understanding of households’ financial circumstances and resources, direct support mechanisms are blunt instruments, often unable to target with accuracy those most in need.
What does the evidence tell us?
Since early 2020, a number of studies have explored households’ ability to cope with the economic costs of the pandemic. Typically, these use high-frequency data for a non-representative sample of households within specific countries to learn about how spending patterns have changed and how income opportunities have dwindled in different sectors of the economy.
For example, research shows that in the UK, roughly a third of households have trouble keeping up with bill payments, 42% have cut other expenditure to prioritise housing costs and over half have dipped into their savings (Brewer and Handscomb, 2020). Other work uses high-frequency commercial data from financial technology (‘fintech’) companies to document rising inequality during the pandemic, with the most economically vulnerable groups experiencing the largest declines in spending and the slowest recovery (Baker et al, 2020; Hacioglu et al, 2020).
In terms of suggested solutions, studies indicate that traditional macroeconomic tools used to stimulate demand across the whole economy have little power to restore employment when consumer spending is limited due to lockdowns (Chetty et al, 2020). For example, encouraging spending has little effect when shops, restaurants and entertainment venues are closed. Instead, these studies suggest that direct government support for households is more effective for reducing economic hardship.
Measures designed to support households helped to mitigate income inequality in the early stage of the crisis (O’Donoghue et al, 2020). This is chiefly because of their effects on vulnerable sections of the population (such as those who are self-employed and cannot rely on social networks). Households are prepared to sacrifice, on average, up to 10% of their regular income to be guaranteed similar support during future shocks.
It is also important to have a clear understanding of household expenditure across the income and wealth distributions. Without such information, the identification of financially vulnerable households can end up being imperfect. For example, one study shows that 60% of postponements to mortgage or loan payments in the United States was provided to individuals with above the median income, since higher income individuals tend to have greater debt balances (Cherry et al, 2021).
In emerging economies, policy targeting issues are an additional problem to contend with, as support schemes need to be implemented with little existing infrastructure in place, and the social protection systems that financially vulnerable households can access, such as unemployment benefits and insurance, are generally weak (Gerard et al, 2020).
What do we mean by financial vulnerability?
Two important features of a household’s financial position determine its vulnerability. The first is the minimum level of spending required for food, essential utilities, housing or rental expenses and any monthly debt repayments.
The second is the level of liquid financial assets at the household’s disposal at the point of a shock (such as cash savings, deposit accounts or other liquid financial assets). Without state support, households will need to draw immediately on their liquid financial assets and/or reduce consumption to weather the storm.
Using data comparing 12 developed countries and 12 emerging/middle-income countries, Figure 1 shows the fraction of households in each category that hold different types of asset classes, debt and consumption.
Most households around the world own positive amounts of liquid financial assets (mainly in the form of cash or deposit accounts), as well as physical assets such as cars and property. A lower fraction of households in all countries hold illiquid financial assets, such as pensions and retirement savings accounts. Overall, roughly 50% of households in the developed world have illiquid financial assets, while less than 25% do so in emerging economies.
Figure 1: Participation rates in assets and debt markets
Source: Authors’ calculations based on international household survey data in the euro area, South Africa, Thailand, India, China, the UK and the United States.
In terms of debt, one in two households in developed countries have outstanding debt, but less than a quarter of households in emerging economies do so. Secured debt (debt that is collateralised by assets of some form) is found mainly in developed economy households, while the fractions of households holding unsecured debt are similar in developed and emerging economies.
In terms of monthly consumption expenditures, households in developed countries are more likely to be renters and to repay debt. But they are also more likely to have excess savings – that is, enough excess income to engage in non-essential purchases and to accumulate savings.
Figure 2 takes a ‘balance sheet’ view of household assets and liabilities, adding up all households into a single ‘representative household’ and showing the shares of different assets and liabilities for this grouped entity. Across all countries, physical assets account for the largest share of total assets, and households hold only relatively modest levels of liquid financial assets.
Figure 2: Balance sheet of the representative household
Source: Authors’ calculations
On the liabilities side of the household balance sheet, unsecured borrowing accounts for a very small share of outstanding debt. In emerging markets especially, this reflects the limited ability of households to access credit products, as well as the often-prohibitive costs of both formal and informal borrowing. This may also reflect unrecorded informal debt obligations, or obligations to family and friends.
Finally, the representative household in emerging economies spends a larger share of income on food, utilities and rent than their counterpart in a developed economy, an almost negligible amount is allocated to debt repayment, and a significantly lower fraction of income remains to cover non-essential consumption and savings – around 40% compared with 63% for a typical household in a developed economy.
How can we quantify and compare financial vulnerability?
Liquid financial wealth on household balance sheets is an essential buffer in the event of an income shock. A household’s resilience or vulnerability can be computed as its total level of liquid financial wealth divided by its total monthly consumption expenditure.
This measure of resilience tells us how many months a given household can maintain its level of consumption after an income shock by solely relying on its financial wealth. In a given country’s population, the level of financial wealth across households and the patterns of consumption across this distribution jointly determine the level of resilience of the country’s households.
Figure 3 reports the fraction of households in each country in the sample (of 12 developed and 12 emerging economies) that have a resilience level below three months, six months and one year.
There is significant variation in resilience across these 24 countries. While the fraction of households that cannotself-sustain using financial wealth for three months is higher on average for emerging markets (around 50%) than developed markets (around 40%), it is noteworthy that by the one-year mark, the vulnerable fraction of the population is not very different across emerging and developed economies, at around 70%.
Figure 3: Financial vulnerability in the face of income shocks
Source: Authors’ calculations.
Note: A small number of households in each country consume amounts that exceed their incomes – for example, because they receive emergency relief or other forms of short-term income support that are not reported in the surveys. Such households are excluded from the computation of net measure of financial vulnerability.
How can households cope following an income shock? Figure 4 displays the impact of different coping mechanisms available to households, including the effect of the introduction of policy measures such as debt relief.
The left panel of the figure shows the cumulative share of vulnerable households for various months after the initial income shock. The lower the curve on the vertical axis, the lower the fraction of vulnerable households; conversely, the higher the curve, the higher the fraction of resilient households.
The figure also shows that if all household debt repayments were completely stopped through policy, the vulnerability/resilience curve shifts only very marginally for developed markets, and it hardly moves at all for emerging markets. This illustrates the relatively low power of temporary debt relief programmes, especially in emerging economies.
Figure 4: Margins of adjustment and coping mechanisms
Source: Authors’ calculations.
Instead, if households were to liquidate their illiquidfinancial assets fully, there is a much more substantial positive effect on resilience. But roughly 40% of developed market households continue to be vulnerable six months after the initial shock. In contrast, in emerging markets, illiquid financial assets shift the curve down, but only marginally, reflecting relatively low levels of savings in the form of retirement savings or pensions held by most households in such economies.
These alternative coping strategies highlight the important role of wealth inequality in amplifying vulnerability, and demonstrate the ineffectiveness of particular policy options during crises such as the pandemic. For example, debt relief as a policy tool is likely to be ineffective as those it benefits are already resilient to an income shock, and those who would benefit most from accessing illiquid savings during difficult circumstances have not been able to accumulate such assets, given their relatively low disposable income in good times.
In both emerging and developed markets, a policy instrument that dramatically changes the shape of the vulnerability curve is income support. A 50% income support policy brings the total fraction of vulnerable households to near zero for the first six months following the shock in both sets of economies.
But such policies are enormously expensive to implement and pose major administrative challenges for governments. This is especially true in emerging markets where fiscal imprudence is unsustainable. Finally, efficiently implementing direct transfers is often infeasible for sheer lack of information about the population and limited infrastructure to administer payments.
Faced with enormous income uncertainty in the spring and summer of 2020, households around the world were faced with two main options if the worst materialised: either to cut consumption deeply and defer non-essential purchases; or to draw down accumulated savings.
Unfortunately, for a large part of the world’s population, the second option is just not viable. Financing an income shortfall for a period of three months – not an unusual circumstance for self-employed entrepreneurs and wage earners in many industries – results in complete depletion of liquid financial assets for roughly every third household in the sample of countries that we have analysed.
In the absence of substantial government support, over a longer horizon of a year, income shortfalls given these low levels of observed financial resilience can push two-thirds of households into significant hardship. This generates enormous reliance on government support, but most emerging markets possess little wiggle room to provide support at such a massive scale. What’s more, when shocks are long lasting, neither advanced nor emerging economies can sustain such interventions for prolonged periods of time.
Where can I find out more?
- Cushioning the poor from the COVID-19 shock: The Abdul Latif Jameel Poverty Action Lab.
- The Economic Impacts of COVID-19: Evidence from a New Public Database Built Using Private Sector Data: Opportunity Insights.
- COVID-19 Government Response Tracker: Blavatnik School of Government.
- Policy Responses to COVID19: International Monetary Fund.
- Heterogeneous effects of Covid-19 on households' financial situation and consumption: VOX, CEPR Policy Portal.
- Inequality: the Institute for Fiscal Studies Deaton Review.
- COVID-19 Household Survey: UChicago Urban Labs.
Who are experts on this question?
- Cristian Badarinza, National University of Singapore
- Vimal Balasubramaniam, Queen Mary University of London
- Louiza Bartzoka, Imperial College London
- Marianne Bertrand, University of Chicago
- Richard Blundell, University College London
- Karen Dynan, Harvard University
- Rema Hanna, Harvard University
- Jonathan Parker, MIT
- Tarun Ramadorai, Imperial College London
- Silvana Tenreyro, LSE and Bank of England