COVID policy simulator

Latest version (05/09/2020): Introduces a social evaluation taking account of inequalities in QALYs among social groups, and a more concrete description of testing policy.

Previous updates:
(05/01/2020): Distinguishes hospital beds (acute care) and ICU beds (critical care); introduces Belgium and Guinea.
(04/30/2020): Corrects a mistake in the estimation of fatality impact of overwhelmed hospitals.
(04/25/2020): Incorporates  new estimates of R0, fatality and hospitalization rates and shows the importance of testing.

The Excel spreadsheet at the bottom of this page contains a simulator which offers an intuitive way of understanding the dynamics of the COVID-19 pandemic. It simulates the evolution of the pandemic in absence of intervention, as well as under policies reducing contacts between people or implementing wide testing to identify infectious people early. It provides a well-being evaluation of the health and economic impacts on the population, including a special attention to the most disadvantaged. It also includes a cost-benefit assessment of the health benefits and economic costs of policies, using the value of statistical life approach commonly used for the assessment of safety measures.

Everyone is free to use it either by modifying the policy assumptions, the model parameters, or even by changing the equations and introducing new policy instruments. This simulator should NOT be used to make predictions, but only to get a better grasp of some of the mechanisms.

The results spreadsheets provide the possibility to compare a benchmark scenario and a different policy scenario. It is also possible to use these scenarios to assess the impact of parameter assumptions, which can be changed jointly or separately in the two scenarios.

The model is explained in one sheet of the file and the computation of societal well-being is explained in the "SWF" file at the bottom of this page. It is similar to the classical SIR epidemiological approach (Kermack and McKendrick 1927), with some important variations. In particular, infection is based on intuitive parameters of contact frequency and probability of infection during any contact, and mortality is endogenous to hospital capacity and this plays a key role in determining when and for how long to implement social distancing and lock-down policies. The model allows for precautions spontaneously taken by the population when the number of fatalities surge, as well as additional policy interventions. (It also allows for people who have recovered or been vaccinated to be infected again, as is possible with a coronavirus, but the main calibration assumes that this risk is negligible.)

The other sheets present illustrations for a few countries. Given the lack of accurate data on key characteristics of the virus, the model is not calibrated in a precise way. For more details, see the "Model" sheet.

Users can study other countries by copying a country sheet into a new sheet and entering the country data.

Please send comments and requests to covidspreadsheet@gmail.com.

Download the simulator

Tutorial videos: