Faculty of law blogs / UNIVERSITY OF OXFORD

Business Shutdowns in Light of Substitution Effects: Why Some Covid-19 Mitigation Strategies Fail


Jan-Philip Elm
Visiting Research Fellow at the Hebrew University of Jerusalem’s Faculty of Law; Doctoral Student at the Institute of Law & Economics, University of Hamburg
Roee Sarel
Professor at the Institute of Law and Economics, University of Hamburg


Time to read

3 Minutes

As the Covid-19 pandemic continues to disrupt the global economy, regulators are struggling to find cost-effective mitigation strategies. The goal of such strategies should be simple: reduce the spread of the virus, while causing the least amount of damage to people’s everyday lives, including economic activity. Yet, the diverse measures taken by governments, such as shutdowns of non-essential business activity, full lockdowns, restrictions on gatherings, and closure of borders, only seem to have limited success in achieving the said goal. Given that current estimates of the economic damage indicate a potential cost in the range of a double-digit number of trillions (in USD), figuring out why some Covid-19 mitigation strategies still fail should be a top priority.

Indeed, many studies try to follow this path by examining individual mitigation strategies in a specific industry or a geographical area in isolation. Such studies investigate, for instance, whether specific lockdowns are correlated with the number of Covid-19 cases, hospitalizations, or deaths, while abstracting away from some potentially important aspects. In particular, the study of isolated mitigation strategies only reflects what economists refer to as a ‘partial equilibrium’—how a shock (ie, a given mitigation strategy) affects the outcome in one market, while neglecting spillover effects to (or from) other markets. Respectively, this sort of analysis overlooks two important economic effects: substitution and complementarity.

To illustrate, suppose that the owner of a restaurant is ordered to shut down (under the threat of criminal sanctions) but there are no other restrictions in place. What would happen next? Ideally, the owner of the restaurant will switch to what we refer to as a ‘benign’ activity (eg, stay home and avoid meeting others), but whether this happens depends on available alternatives: the restaurant owner could just switch to catering service and continue to serve people at their private residence—thereby rendering the shutdown ineffective (or even counter-productive), for instance, if the private residence is insufficiently ventilated. This reflects a substitution effect, as the mitigation strategy merely incentivizes people to switch to another activity (without determining which one). Along similar lines, suppose instead that one adopts a strategy that prohibits the sale of alcohol (both retail and wholesale), which prevents the restaurant owner from providing an entertaining service to their customers. This alone may discourage the owner from operating any kind of business activity, causing them to stay at home. This then reflects a complementarity effect: by banning the sale of alcohol, the benefit of its complement—sharing food in a social setting—decreases.

When one considers substitutes and complements, one gets closer to estimating what occurs in a ‘general equilibrium’: how the eventual incentives for different behaviors, after accounting for the complete chain of events, translate into the measured metrics (Covid-19 cases and deaths).  In our recent paper we use this line of argument to explain why some Covid-19 mitigation strategies still fail. Consider, for example, Germany’s decision to enact a ‘Lockdown Light’ for the month of November 2020, which failed to achieve its goal—escalating thereafter into much harsher back-to-back lockdowns. One problem with the initial lockdown was that while restaurants and bars were ordered to close, it was still allowed to gather in groups of up to 10 people in public. A substitution effect could then lead people to have joint picnics or set up outdoor food stands (possibly with less social-distancing compared to a restaurant, especially in times of cold, rainy weather). Respectively, Germany’s later decision to ban the consumption of alcohol in public (in several areas) may act similarly to the example described above, discouraging social gatherings as a complement of drinking alcoholic beverages. Our reasoning, in turn, can help explain this chain of events.

However,  the efficacy of the measures may depend on the combination of mitigation strategies: some legal interventions are ‘strategic complements’ (synergistically reducing Covid-19 infections, eg by prohibiting two harmful behaviors that are otherwise close substitutes) whereas others are ‘strategic substitutes’ (leading to both unnecessary spending and only minor reductions in Covid-19 infections). The degree to which substitution and complementarity effects influence behavior depends on various factors, including the degree to which people share similar utility functions and whether they behave as predicted by standard models of decision-making.

Yet, one thing is clear: crafting mitigation strategies without consideration for spillover effects inevitably keeps policymakers from effectively countering the spread of Covid-19.

This post is based on a recent paper, which is available here.

Jan-Philip Elm is a visiting research fellow at the Hebrew University of Jerusalem’s Faculty of Law and a doctoral student at the Institute of Law & Economics, University of Hamburg.

Dr. Roee Sarel is a post-doctoral research associate at the Institute of Law & Economics, University of Hamburg.


With the support of