Covid-19 briefing: epi-macro 101

Cristiano Cantore, Federico Di Pace, Riccardo M Masolo, Silvia Miranda-Agrippino and Arthur Turrell

The Covid-19 crisis has led to a swift shift in the emphasis of macroeconomic research. At the centre of this is a new field of inquiry called ‘epi-macro’ that combines epidemiological models with macroeconomic models. In this post, we give a brief introduction to some of the earliest papers in this fast-growing literature.

SIR Model

The starting point for epi-macro is the compartmental SIR model. This workhorse epidemiological model has three different compartments (Susceptible, Infected and Recovered) and models how individuals transition between them. Epi-macro models extend these models in two ways. First, the spread of the virus is no longer assumed to be exogenous but is instead affected by individuals’ behaviour, for instance consumers might not visit restaurants as much if they fear becoming infected while dining. Second, the linking of a simple model of production and consumption model to the progression of the disease facilitates predictions of economic variables. Though simple, these epi-macro models generate qualitative results with clear policy relevance.

Result 1: Voluntary social distancing can generate large and persistent recessions even in the absence of legally enforced measures.

Eichenbaum, Rebelo and Trabandt (2020a) assume that those who are susceptible are more likely to get infected the more they interact with infected consumers and workers. In their model, the epidemic has both aggregate demand and aggregate supply effects, as agents reduce both their labour supply and their consumption to reduce their risk of contracting the disease. The probability of becoming infected depends on the current number of infected individuals. When an individual becomes infected, this creates an externality as the higher stock of infected people raises the risk for others. In a decentralised economy, agents do not fully internalise the impact of their actions on the number of infected and macro aggregates. For this reason, voluntary social distancing is sub-optimal. Optimal mandated containment measures are stricter, saving more lives, but at the cost of generating a deeper economic recession.

In an extension, Eichenbaum, Rebelo and Trabandt (2020b) show that a policy that uses tests to quarantine infected people has large social benefits when people are uncertain about their health status. In their framework, agents differ only in their health status but are otherwise homogeneous. Conditional on their health status, agents have the same consumption, productivity, and working conditions. The authors show that a policy that uses tests to quarantine infected people yields large social benefits relative to broad-based containment policies like lockdowns. For the US this implies a reduction of total infections and deaths by half, and a decline in consumption and hours worked by one third. However, the recent national and international experiences demonstrate that the economic consequences of a pandemic depend crucially on the composition of the workforce. Sectors that rely on social consumption, e.g. hospitality, transport and retail, are more likely to be adversely affected than others. Equally, not all occupations are flexible enough to enable workers to work from home.

Result 2: Social distancing policies have distributional effects, with first round costs falling more on poorer households.

Kaplan, Moll and Violante (2020) look at the distributional effects of the pandemic by combining a Heterogeneous Agent New Keynesian (HANK) framework with an epidemiological model. Their model presents the same two-way feedback between the infection risk and economic behaviour as in Eichenbaum, Rebelo and Trabandt (2020a) but allows for different occupations among workers and different sectors in the economy. An important dimension they consider is the heterogeneity in labour income and amount of liquid wealth across workers. The authors find that the first round costs of the epidemic and lockdown policies hit poor households the most in the US. The most exposed households also have very little liquidity and may not be able to survive for long without financial help. The authors also quantify that the recession initially induced by the lockdown is not much worse than in the counterfactual scenario in which the government does nothing. In their model, the duration of the lockdown is crucial. Only a long lockdown (of 500 days) can achieve a substantial reduction in deaths because a short one produces a second wave of infections. 

Result 3: Retired agents benefit the most from the implementation of lockdown policies and young workers in ‘non-essential’ sectors bear more of the costs.

One dimension Kaplan, Moll and Violante (2020) abstract from is that the benefits and costs of the lockdown can vary enormously by age. Heathcote et al (2020) introduce two age groups in an Epi-Macro model which accounts for those above 65 year of age being 25 times more likely to die from Covid-19 than those below 65. Therefore, the lockdown produces enormous benefits for the old. Younger agents pay most of the costs of the lockdown, especially those working in ‘non-essential’ sectors that are forced to close. Heathcote et al (2020) quantify the welfare effects of the lockdown by focusing on two government instruments: the fraction of the economy to be shut down (extent of mitigation) and a tax on workers used to make transfers to retired, unemployed and infected agents (redistribution). They show that different age/working groups prefer different policies. While old agents want an extensive and prolonged lockdown, the young working for ‘non-essential’ sectors prefer a modest and short lockdown. A government maximising utilitarian social welfare compromises, picking an intermediate path for mitigation. This rationalises current disagreements among the population on when to end the lockdown.

Result 4: Social distancing has a positive ‘infection externality’ that helps to isolate workers in the core sector of the economy.

Bodenstein, Corsetti and Guerrieri (2020) challenge the standard view of a trade-off between the health and social/economic costs of the pandemic by building a model that combines a two-sector structure with a standard SIR environment. The macroeconomic part of the model features a core sector that produces intermediate inputs that are not easily replaceable by the non-core sector. The study finds that, absent social distancing, the high peak of an infection may cause very large upfront economic costs in terms of output, consumption, and investment. Therefore, social distancing measures that encourage workers in non-core occupations to work from home are effective in reducing the economic costs of the pandemic: this policy helps to dampen the surge in infections by reducing contacts with workers in the core sector. Thus, social distancing has a positive ‘infection externality’ that helps isolate workers in the core sector from succumbing to the disease. This helps to slow down the spread of the disease.

Final thoughts

These models can be used to quantify the costs and benefits of alternative policies and their distributional consequences (subject to the uncertainty on both epidemiological parameters and model specifications). This early work is already being built on by more sophisticated models.

While these early efforts to capture the complex interplay of epidemiology and macroeconomics focus on the short run, there are many more important questions to answer — the longer-term effects of the pandemic on productivity, unemployment dynamics, and monetary and fiscal policy being chief among them.

Cristiano Cantore works in the Bank’s Research Hub, Federico Di Pace, Riccardo M Masolo and Silvia Miranda-Agrippino work in the Bank’s Monetary Policy Outlook Division and Arthur Turrell works in the Bank’s Advanced Analytics Division.

If you want to get in touch, please email us at or leave a comment below.

Comments will only appear once approved by a moderator, and are only published where a full name is supplied. Bank Underground is a blog for Bank of England staff to share views that challenge – or support – prevailing policy orthodoxies. The views expressed here are those of the authors, and are not necessarily those of the Bank of England, or its policy committees.