Application of an Epi-Econ-Model to analyze COVID-19 lockdown policies in the Netherlands: Lessons and limitations

Epidemiological and economic (Epi-econ) models account for endogenous interactions between the epidemic and the economy. We explore the applicability of an Epi-econ model to isolate the effects of
lockdown policies during coronavirus disease 2019 in the Netherlands. To this aim, we recalibrate the seminal Epi-econ model of Eichenbaum and colleagues with updated parameters specific to the Dutch
context. We find that the model performs poorly in replicating observed Epi-econ trends under baseline assumptions. Next, we explore possibilities to improve model fit by relaxing policy and transmission parameters, and by incorporating observed “random noise” in infectivity parameters. This approach spectacularly improves model performance in replicating observed trends. Finally, we test the performance of the model in simulating alternative policy scenarios. We use the Containment and Health Index from the Blavatnik School of Government to replace Dutch policy parameters with exemplary countries on opposite sides of the stringency spectrum. We find that a more stringent lockdown policy would reduce peak prevalence, while aggravating peak economic contraction, but with little effect on overall trends. Conversely, a more lenient lockdown policy was estimated to increase the peak and overall prevalence, with little effect on economic outcomes. We conclude that while rigorous adjustments to existing models were required, a combined Epi-econ model could be informative to policymakers in assessing alternative lockdown policy options.