Social distancing and supply disruptions in a pandemic
By the end of March 2020, nearly four months after the first detection of significant infections from the novel coronavirus in China, most advanced economies adopted measures restricting people’s movements and activity on their territory, introduced tough controls at their borders, and mandated norms implementing social distancing. Albeit with some delay, governments have now converged on the idea that some restrictions are required to reduce the human cost of the disease.
After imposing strict lockdowns that, in the absence of more effective health measures, have avoided overwhelming the capacity of healthcare systems, many countries are now relaxing their initial measures. Lockdowns have been criticised as safeguarding lives but destroying livelihoods.
To highlight the importance of a systematic health policy response both for saving lives and in support of economic activity, we find it important to consider a counterfactual world in which no health policy measures were taken. As Gourinchas (2020) notes, a recession would occur anyway, fuelled by the uncertainty of dealing with a pandemic without adequate public health support.
Are health measures to be blamed for the economic downturn?
In recent work (Bodenstein et al. 2020), we also question the view that social distancing measures, rather than COVID-19, should be blamed for the economic costs of the pandemic. Complementing the growing literature that integrates epidemiological and economic models to examine the trade-offs between economics and public health, we build a model to highlight the economic risks stemming from potential supply disruptions that the pandemic may create – if social distancing measures are either ignored or applied inefficiently.
To start with, we show that, if only because of the effect of the disease on labour supply, letting the infection rise unmitigated could have caused a large upfront contraction in economic activity. The key problem is the high number of symptomatic infective individuals who would not have been able to work at the peak of the pandemic. Economic activity could have dropped as much as 30% per month – independently of any additional negative impact from changes in demand patterns or from financial disruptions.
This adverse economic scenario is based on the same epidemiological model that motivated strict lockdowns in view of higher mortality risks expected when the capacity of health care systems is exceeded. That is, the same model could support a lockdown in view of the high economic (supply-side) risks from not taking any health measures.
How labour supply disruptions may propagate through the economy
To gauge the potential economic consequences of an unchecked pandemic, we propose thinking of the production network in the economy as consisting of two sectors – one producing goods and services that are essential inputs to production in the other. While it may be difficult to identify exactly which firms and industries are essential, there are some clear candidates: distribution services, energy production, healthcare services. In Table 1, we show the employment and value-added shares of these essential industries for the US.
Table 1 The core sector: Share of US GDP and employment
Source: Authors’ calculations based on the GDP by Industry from the U.S. Bureau of Economic analysis and based on the Productivity Release of the U.S. Bureau of Labor Statistics.
The economic risks from labour supply shortages in essential sectors can be large. Our view is that the production of essential inputs may be subject to minimum scale requirements: a certain number of doctors and nurses are needed for hospitals to function, a certain number of conductors for a subway system to run. When the labour force goes below this minimum scale, supply disruptions become larger and propagate through the economy as a whole.
We also stress that the risk of supply disruptions remains high when social distancing measures are cast aside before establishing ample contact tracing capacity, testing, and the ability to effectively isolate infected individuals. The key problem is that strict lockdowns delay the acquisition of herd immunity. If and when strict measures are lifted completely, a resurgent spread of the disease could impose another large economic toll, apart from resulting in further loss of life. Again, an important ingredient in our analysis is that a precipitous reduction in the workforce may have disproportional effects on industries that provide goods or services that are essential for the rest of the economy to function. This scenario is shown in Figure 1.
Figure 1 A lockdown gone wrong
Structuring health measures to avoid supply disruptions
Nonetheless, social distancing measures need neither be blanket unstructured measures, nor suspended abruptly. We use our analysis to provide a blueprint for how to structure measures that achieve societally acceptable infection rates while reducing the economic risks and costs for the economy. Once again, our perspective is from the supply side of the economy – there are other important considerations concerning income, consumption and investment demand that are bound to play a role but that we do not fully capture, for the time being, in constructing our scenarios.
We elaborate on the idea that efficient lockdowns should be structured addressing the problem of infection externalities. Using mobility data, Chen et al. (2020) found that spontaneous social distancing preceded formal social distancing across US states. While people may spontaneously take precaution to avoid being infected, they may not completely internalise the risk that their choices imply for others – for instance, by crowding public transportation, stores, or other public spaces.
How does this externality play out in our analysis of supply disruption of the disease? The key point is that workers in essential industries (such as utilities, distribution, agriculture, police and sanitation) tend to be less able to work from home than workers in other industries and sectors. Looking at US data, the essential industries account for about 30 percent of the labour force. Based on the American Time Use Survey conducted by the U.S. Bureau of Labor Statistics, we gauge that only 15% of workers in these industries are able to work from home. In the rest of the economy, 40% of workers are able to work from home (see Barbieri et al. 2020 for an analysis using Italian data).
Given the importance of protecting essential industries, it would make sense to design health measures that recognise the different risks to which essential and non-essential workers are exposed. Our model confirms the validity of the intuitive idea that lockdowns should be skewed towards workers in industries with a greater prevalence of occupations that can be carried out from home – by staying at home non-essential workers who can continue working from home help reduce the infection rate among workers in essential industries.
The infection externality is a strong argument in favour of higher rates of lockdown of workers in the non-essential sector. In the scenario, shown in Figure 2, these structured measures smooth out economic costs over time.
Health measures can be detailed also at the micro level, addressing complex issues concerning how to help firms to ensure enough sanitation/care and reduce the infection risks for their workers. We should note that these measures are bound to raise production costs at a time where demand is dwindling.
Infection externalities and coordination problems
In addition to the infection externality, health measures need to address an additional problem. Individuals may spontaneously adopt sound precautionary behaviour. Yet, they may find it difficult to avoid showing up at the same time in the supermarket, at a public park, or at a bus stop, thus increasing contacts and contagion rates. As another example, Maloney and Taskin (2000) stress the coordination required for schools to close and move to distance learning, a step concerned parents cannot accomplish on their own. Without the coordination of structured lockdowns, outcomes may be erratic and depend on people’s anticipation of what others do. At times, parks and markets will remain empty, at other times they will be dangerously crowded.
By the same token, individual firms may be happy to adopt costly measures to fight the virus, but they would like their efforts not to be wasted by relaxed standards and/or disruptions among their suppliers or distributors. Guidelines, shared protocols and rules should help all firms to reinforce each other’s virtuous behaviour along the production chain. Without attempting to engage in costly micro-management, some health guidelines could go a long way in resolving this coordination problems.
Figure 2 A targeted lockdown
By addressing the infection externality and the coordination problem, health measures could be structured and communicated to reassure people that, under the proposed rules, they can conduct their lives in a reasonably safe way, and, if falling sick, they will receive adequate medical assistance. Under these conditions, the economy can come back to life, reactivating many of the valuable links that exist between firms, workers, and financial institutions that are now threatened by the prolonged economic standstill.
Authors’ note: The views expressed in this article are solely the responsibility of the authors and should not be interpreted as reflecting the views of the Board of Governors of the Federal Reserve System or of any other person associated with the Federal Reserve System.