Countries all over the world are fighting COVID-19 by reducing social and economic interactions, measures that have proven effective in parts of the world but are extremely costly. We are in the midst of the deepest recession since the great depression, and World GDP is expected to fall by 4.9% in 2020, with two-digit declines in many countries (IMF 2020). A series of excellent papers suggest that the huge costs in terms of livelihoods are justified by the gains in terms of lives saved from the pandemic (e.g. Acemoglu et al. 2020, Alvarez et al 2020, Eichenbaum et al. 2020, Farboodi et al. 2020, Hall et al. 2020).
This focus on a trade-off between lives and livelihoods assumes that households are able to bear the economic losses arising from a lockdown and the global recession. But in parts of society that are at or close to subsistence – already struggling to make a living in normal times – deepening the recession via a lockdown may be extremely costly, forcing more households to suffer from deprivation. In other words, the central trade-off for optimal policy is not one between lives and livelihoods, but between lives and lives. This is not just a theoretical possibility: Global hunger is projected to double this year (WFP 2020), and in June 2020, even in the US 14 million children did not have enough to eat (Bauer 2020). At a more general level, a deeper recession may add to deprivation and to ‘deaths of despair’ (Case and Deaton 2020).
Fiscal policy can cushion the blow and help mitigate the excruciating trade-offs involved in fighting a pandemic. But this requires enormous fiscal space: The world is projected to run a fiscal deficit of 13.9% of GDP this year, and the US deficit may reach 23.8% (IMF 2020). Such policy measures are important in industrialised countries. They would be even more valuable in developing countries, where a larger fraction of the population lives in poverty. The problem is that these countries often had limited fiscal space to begin with, only to see it collapse beneath the global impact of COVID-19 on commodity prices, tourism, remittances, and capital flows (Hausmann 2020, Hevia and Neumeyer 2020).
What does this imply for policy? How does the optimal lockdown depend on accompanying policy measures to alleviate the economic shock? How do such measures impact welfare and compliance with lockdowns? What is the cost of limited fiscal space in a pandemic, and how does it affect policy? Are lockdowns a luxury good, desirable only for countries rich enough to be able to bear the consequences?
A theoretical analysis of horrible trade-offs and policy implications
To address these questions, we introduce a subsistence level of consumption into a tractable general equilibrium framework with heterogeneous agents (Hausmann and Schetter 2020). We then use this framework to analyse jointly optimal policies of a lockdown and transfer payments during a pandemic. In our model, the pandemic causes deaths from the disease and a temporary loss in total factor productivity (TFP)—the latter can also be attributed to global economic shocks. The government can decide to fight the pandemic via a lockdown and use lump-sum transfers financed via costly international borrowing to cushion the blow. The lockdown requires households to reduce their labour supply, thereby lowering disease transmission. A more stringent lockdown also deepens the recession. On the one hand, this gives rise to the much-debated trade-off between lives and livelihoods. On the other hand, it means that poorer parts of society will find it harder to subsist, will be less able to comply with the lockdown, and the poorest may even suffer from deprivation, giving rise to an excruciating trade-off between lives and lives (Levy Yeyati and Sartorio 2020).
We show that this trade-off fundamentally changes the main margin for optimal policy, and that a threat of deprivation to vulnerable parts of society imposes strong bounds on the optimal lockdown. If government borrowing is unconstrained, this is reflected in the fact that optimal policy saves most or all households from deprivation. Because the transfers needed to achieve this are costly, it further implies that the optimal lockdown is less strict in poorer countries where a larger fraction of the population is close to subsistence, i.e. countries more concerned by deprivation vis-à-vis the pandemic. In that sense, a lockdown may be seen as a ‘luxury good’. As a consequence, the aggregate death burden and the welfare loss from the pandemic are higher in poorer countries.
This is true with unlimited fiscal space. It can get much worse if government borrowing is constrained. We show that countries with limited fiscal space see themselves forced to fight less the pandemic itself than people dying from the disease — either directly, because the optimal lockdown is less strict, or indirectly, because compliance with the lockdown is lower. Still, this may not be enough to save all parts of society from deprivation.1
Figure 1 illustrates these insights by means of a numerical example of our model. It presents comparative statics results with respect to aggregate TFP for the cases with (crossed-out lines) and without a binding government borrowing constraint. Further details are provided in the figure’s notes.
Figure 1 Comparative statics with respect to aggregate TFP
d) Death burden
Source: Hausmann and Schetter (2020).
Notes: This figure shows comparative statics with respect to aggregate TFP for a numerical example of our model. Panel (a) shows the optimal lockdown, θ, expressed as the (average) mandated reduction in hours worked over a one-year period. Panel (b) shows the optimal lump-sum transfers, t, as a percentage of pre-pandemic GDP. Panel (c) shows aggregate welfare relative to a no-pandemic case, W. Panel (d) shows the aggregate death burden during the pandemic, D, along with a split up in deaths from the disease (green line) and deaths from deprivation (red line). The normal (crossed-out) lines refer to the case without (with) a binding borrowing constraint. For the cases with a binding borrowing constraint, this constraint has been set at 3% of pre-pandemic GDP, which corresponds to ~50% of the unconstrained optimal transfers for an economy with TFP=1. See Hausmann and Schetter (2020) for further details.
Distributional consequences within countries
The previous discussions highlight distributional consequences across countries. Yet, the pandemic and policy also have important distributional consequences within countries: On the one hand, the lockdown benefits the least—and may even hurt—the poorest households in a society. These households may not be able to afford full compliance with the lockdown and, hence, face a higher probability of dying from the disease (see e.g. Blundell et al. 2020). In extreme cases, they may not be able to live through the recession and may die from deprivation. On the other hand, and for the same reasons, poor households benefit most from transfers. We analyse the arising distributional conflict in two ways: We first show how a ‘median-voter’ framework can be applied to our set-up with a two-dimensional policy space. This analysis suggests that, ceteris paribus, we should expect to observe stricter lockdowns and fewer transfer payments in countries that are governed by elites, and vice versa in countries with populist governments. Our work may thus help explain some of the differences in policy outcomes that we observe. Second, we ask whether supporting the poor may be in the self-interest of the rich. We show that the answer is affirmative if the externality of working is sufficiently large; that is, if compliance by the poor with the lockdown is sufficiently important for the pandemic and the risk to the rich of getting infected. Hence, providing transfers to vulnerable parts of society in times of a pandemic is not only the right thing to do, but may also be to the benefit of society at large.
COVID-19 is the biggest threat to global health and economic well-being in decades. In Hausmann and Schetter (2020), we contribute to the surging literature on COVID-19 by analysing horrible trade-offs in times of a pandemic.
We show how a subsistence level of consumption can shift the main margin for optimal policy from the much-debated trade-off between lives and livelihoods to a trade-off between lives and lives. Transfers are critical for mitigating this trade-off, and they also help fight the pandemic by increasing compliance with lockdowns.2 But many developing countries lack the fiscal space needed for such transfers, with devastating consequences: We show that these countries are forced to fight less the pandemic than people dying from the disease, and still part of society may suffer from deprivation. The dire forecasts of the World Food Programme and others show that such humanitarian tragedies are not just theoretical possibilities. The international community has the means to avoid them by providing additional financial support to developing countries (Bolton et al. 2020, Hausmann 2020b).
Acemoglu, D, V Chernozhukov, I Werning and M D Whinston (2020), “A multi-risk SIR model with optimally targeted lockdown”, Working Paper 27102, NBER.
Alvarez, F E, D Argente and F Lippi (2020), “A simple planning problem for COVID-19 lockdown”, Working Paper 26981, NBER.
1 There are many reasons why optimal lockdowns should depend on the stage of a country’s economic development, encompassing differences in the demographic structure, differences in urbanisation rates, and differences in the quality of the healthcare system. Incorporating such differences might affect the overall comparative statics across countries, but would not fundamentally change our main insights regarding the importance of a subsistence level of consumption and ample fiscal space for optimal policy and welfare.
2 These interactions also highlight the need to jointly analyze optimal containment and fiscal policies.