The Economic Cost of Social Distancing

Social distancing is proving to be the most – if not the only - effective tool we have to limit the cost of the COVID-19 outbreak in terms of human lives, but it comes with a difficult-to-estimate price tag in terms of GDP decline.

Two Bocconi Professors, Basile Grassi and Julien Sauvagnat, and one from HEC Paris, Jean-Noël Barrot, estimated the economic consequences of the 6-week lockdown imposed by the French authorities in a 5.6% annual GDP decline using a standard model of production networks that takes into account the sectorial interdependencies of the French economy. The same measures would have different costs in countries with different sectorial composition, they argue, and estimated that the loss of GPD – had other governments taken the same French measures – would have been 6.6% in Italy, 5.7% in Germany, 5.5% in the United Kingdom and 6.7% in Spain. The European country most prone to social distancing economic costs turns out to be Bulgaria (with a decline of 9.2%), the least prone Denmark (-4.3%).

The authors start by estimating as 52% the decline in active workforce due to administrative closings (restaurants, hotels, etc.), school closings (that affect not only the school employees, but also those with dependent children) and confinement rules (that allow only those able to telework to continue). This translates into different production cuts, depending on the features of each sector and on their interdependencies. The model they use takes into account both the elasticity of substitution among goods (i.e. the response of household consumption) and the elasticity of substitution in intermediate consumption (i.e. the response of industrial sectors).

If 5.6% is the total GDP growth decline, sectors are differently affected. Besides the sectors directly interested by the closures (tourism, restaurants), those that record the worst decline are the upstream sectors, more distant from end user demand.

With their exercise, the three scholars intend to provide policy makers with a useful, if imperfect, tool that can complete their dashboard for making decisions about social distancing measures. «Our model», they concede, «doesn't consider international trade, automatic stabilizers, and the measures taken by the government to support the economy. Anyway, we are working on the latter two aspects».

Policy makers could also derive from the study important suggestions for a deconfinement strategy, as the authors run a simulation of what could happen if the measures were lifted in a differentiated way after 4 weeks instead of 6. Deconfining some age groups, regions or sectors before others would make a difference. In particular, the strongest economic effect would come from deconfining the construction sector first.