As most advanced economies have entered into extensive lockdowns of unforeseeable duration to contain the spread of COVID-19 (Baldwin 2020), household incomes are bound to be severely hit. How many households will be able to get through these hard times alone and for how long?
In this column, we use the recently published third wave of the Household Finance and Consumption Survey (HFCS), the harmonised survey of European households’ balance sheets coordinated by the ECB (HFCN 2020), to gather some insights into these pressing questions.
Material deprivation is typically associated with an income level that is below a conventional, and socially acceptable, threshold (Atkinson and Bourguignon 2000, OCSE 2016). For example, within the Europe 2020 strategy the European Commission deems households to be at risk of poverty if their equivalised disposable income falls short of 60% of the national median. Because their incomes are so low, these households are seen as being particularly exposed to standard shocks to the income process. Indeed, most analyses of the effects of lockdown measures on household budgets have essentially focused on the labour market effects, the size and distribution of the income drop and its consequences in terms of expenditures and saving behaviour (Guiso and Terlizzese 2020, Crawford et al. 2020).
However, households can also draw on their wealth to make up for income losses. Indeed, wealth is accumulated also for precautionary motives and to be able to face unexpected circumstances. Hence, a realistic assessment of the potential economic hardship caused by the lockdown and social distancing measures, and the ensuing freeze of large portions of economic activity, must bring household wealth into the picture. Combining income and wealth in a single index of potential deprivation is not easy, however, and several approaches have been proposed (e.g. Brandolini et al. 2010, Mueller and Schmidt 2015).
In light of the current events, we have chosen to assess European households’ resilience to the lockdowns by focusing on ‘wealth poverty’ – i.e. the situation where, absent government intervention or an assistance programme, a household cannot maintain a socially acceptable living standard should its income suddenly evaporate. Operationally, we label a household as ‘wealth poor’ when spending its savings is not enough to keep it at the national at-risk-of-poverty line for a specific period of time. We must therefore choose the duration for which, and the resources with which, a household should be able to go it alone.
As regards duration, epidemiological analyses are still struggling to define a potential exit date, even abstracting from the pace at which economic activities will return to fully operational. However, the Chinese experience may be of help. Early containment measures were introduced on 23 January in the Hubei cities where the first outbreaks were recorded. On 27 January, they were extended to the entire Hubei province. Now that contagion has been nearly halted, these measures are only gradually and partially being removed over the month of April, with the opening of Wuhan on 8 April. Overall, the lockdown will have been in place for roughly three months. It is therefore reasonable that in Europe lockdown will last at least three months. As an example, in Italy – the European country furthest ahead in its epidemiological curve – the first measures were enacted in late February in the Northern towns initially hit. The measures were extended nationwide on 11 March and, at the time of writing, were expected to be only partially lifted on 14 April. Therefore, we consider a conservative period of three months.
As regards the resources households can draw upon, we only select easily expendable liquid financial assets. Specifically, we include deposits, bonds, and listed equities and exclude non listed shares, voluntary pensions and whole life insurances. We thus consider all households whose liquid financial assets are not enough to stay above the at-risk-of-poverty threshold for three months to be ‘financially poor’.
Figure 1 plots, for each EU country in the HFCS, the share of population in at-risk-of-poverty households (those with insufficient income) against the share of financially poor households (those with insufficient financial assets). The two sets only partially overlap: with few exceptions, about 80% of the population at risk of poverty is also financially poor, and between 30% and 40% of the financially poor households are also at risk of poverty. Across countries, the two measures are positively correlated but the incidence of financial poverty is much more heterogeneous, ranging between 20% and 90% of the population compared with between 15% and 30% being at risk of poverty. In Italy and Spain, the two countries most severely hit by the epidemic, and in France and Germany, the other two countries following in their footsteps, about half of the population are in financially poor households. With the notable exceptions of Malta, Austria and the Netherlands, in all other countries the share of the financially poor is at least as large.
Figure 1 At risk of poverty and financially poor
Source: HFCS, UDB 3.0 (for Spain, UDB 2.4)
The sources of income of financially poor households are likely to suffer differently from containment measures (Adams-Prassi et al. 2020, Bell et al. 2020). For example, transfers such as pensions are highly insulated from the consequences of the lockdown. Also, self-employed workers are more exposed than open-ended employees because they do not enjoy employment protection legislation and are also usually not eligible for unemployment insurance schemes. Among employees, those on fixed-term contracts have a harder time in keeping their jobs and in reaching eligibility criteria for generous unemployment insurance. In a similar vein, it is reasonable that, everything else equal, financial pressures are stronger for households renting their main residence than for those who own their residence, for whom delays in mortgage repayments are being enacted or discussed in most countries.
In Figure 2 we decompose the share of population in financially poor households according to these simple indexes of vulnerability. In the left panel, the share of financially poor is split according to the riskiness of the occupational status of the main earner in the household (pensioners and open-ended employees, self-employed, fixed-term employees and jobless); in the right panel the share is split according to home ownership status (that is, renters vs owners).
Figure 2 Main income sources and home ownership status of the financially poor
Source: HFCS, UDB 3.0 (for Spain, UDB 2.4)
In Spain, a quarter of the population are in financially poor households whose main income earner works in activities more exposed to the consequences of containment measures or is jobless; in Italy, France and Germany this number is quite a bit lower (ranging from 17% in Italy to 13% in Germany). However, the different exposures of incomes to containment measures is partly compensated by the vulnerability stemming from home ownership status. In Germany, a third of the population belongs to financially poor households renting their main residence; this is the case for only 13% of the population in Spain and for slightly less than a fifth of the population in Italy.
This coarse picture neglects the composite nature of household incomes. For example, if, in addition to the self-employed main earner’s income, the household can draw on an open-ended employees’ income, the impact of containment measures will presumably be weaker than if all income stemmed from self-employment. Also in this respect, there is great heterogeneity across European countries. In France, Germany and Spain, between 10% and 15% of the total income of financially poor households whose main income source is relatively immune from containment measures stems from activities more exposed to the lockdown; in Italy the share is only 5%. At the same time, in France and Germany, around 85% of the total income of households mostly relying on an exposed main earner stems from similarly exposed activities of other members; in Italy and in Spain the share is much lower at 70% and 68%, respectively (Figure 3).
Figure 3 Share of household income from risky sources
Source: HFCS, UDB 3.0 (for Spain, UDB 2.4)
These coarse gauges are based on sharp assumptions, but they highlight that across European countries there are large (and similar) shares of the population that are likely to suffer from the economic fallout of containment measures – albeit through different channels – and that might not have enough own financial resources to maintain a minimum threshold of wellbeing for a whole quarter. The need for household support measures will most likely be felt throughout Europe.
Authors’ note: The opinions expressed are those of the authors and do not necessarily reflect the views of the Bank of Italy or the European System of Central Banks.
Adams-Prassi, A, T Boneva, M Golin and C Rauh (2020), “The large and unequal impact of COVID-19 on workers“, VoxEU.org, 8 April.
Atkinson, A B and F Bourguignon (2000), “Introduction: Income distribution and economics” in Handbook of income distribution, Vol. 1, pp. 1-58.
Baldwin, R (2020), “COVID, remobilisation and the ‘stringency possibility corridor’: Creating wealth while protecting health", VoxEU.org, 10 April.
Bell, B, N Bloom, J Blundel and L Pistaferri (2020), “Prepare for large wage cuts if you are younger and work in a small firm”, VoxEU.org, 6 April.
Brandolini, A, S Magri and T Smeeding (2010), “Asset-based measurement of poverty”, Journal of Policy Analisys and Management 29(2).
Crawford, R, A Davenport, R Joyce, and P Levell (2020), “Household spending and coronavirus”, IFS Briefing note.
D’Alessio, G and A Neri (2015), "Income and wealth sample estimates consistent with macro aggregates: some experiments," Questioni di Economia e Finanza (Occasional Papers) 272, Bank of Italy.
Guiso, L and D Terlizzese (2020), “How much the COVID-19 weighs on Italian households” (in Italian).
HFCN (2020), “The Household Finance and Consumption Survey – Results from the 2017 wave”, ECB Statistics Paper Series No. 36, March.
Kennickel, A (2019), "The tail that wags: differences in effective right tail coverage and estimates of wealth inequality," The Journal of Economic Inequality 17(4).
Müller, P and T Schmidt, (2015), "Identifying income and wealth-poor households in the euro area", Discussion Paper No 35/2015, Deutsche Bundesbank.
OECD (2016), Poverty rate (indicator).
1 It is known that household surveys are subject to relevant underreporting of household wealth, especially as concerns financial assets (Kennickel 2019). However, this underreporting is particularly concentrated in the top tail of the distribution, where financial assets beyond deposits are typically concentrated (D’Alessio and Neri 2015).
2 These latter shares differ from those reported by Eurostat. This reflects both the different underlying household survey and different definitions of the relevant variables. Differently from the European Survey of Income and Living Conditions (EUSILC) underlying Eurostat estimates, the HFCS collects household income before-tax and does not include imputed rents in this estimate.