Several countries and states are leaving or have partially left the lockdown and, with constraints, are recovering economical activity after getting hit by the covid-19 crisis.

WHO has warned that there could be a second wave (likely to occur this fall) of covid-19, coinciding with other seasonal diseases like the common flu. Several scientists are also warning about this fact (see here and here). These personalities are warning that this could occur and that it could be more devastating, but to my knowledge we do not have a probabilistic prediction of the likelihood of this to happen.

I wonder if we have some unified intergovernmental probabilistic prediction for this to happen. Something that encompasses all the scientific predictions made (ensemble prediction)?


Yes, there are.

This study for France says there is a 1.5% chance of a second wave. This article discusses three different papers modelling successive waves of covid (not only two). As usual, conclusions depend on modelling assumptions and different scenarios. For instance, this article states the outcome depends on how R and other factors (whether the virus mutate or not) behave.

A personal reflection, if I may. The modelling answer has two sides. One is the epidemiological dynamics, and the other is the behavioural one. As an economist, I am particularly concerned with the latter, which I think is particularly difficult to predict and the models above do not consider. Covid-19 was a new thing four months ago and uncertainty led to panic (remember toilet paper buying spree?), significant willingness and openness to testing and, in the end, lockdowns. A second wave will for sure be different. Since the virus is evidently less dangerous as originally imagined (particularly for the population not at-risk, the majority), and most normal people do not want lockdowns again (as it can be seen from the reaction to their easing), there could be less willingness to cooperate with public health authorities. For instance, it might well be that individuals with probably mild cases won't report them. Why bother? This behavioural change could have devastating consequences for the development of a second wave. I would expect any reasonable analysis to incorporate this and other behavioural responses in the analysis. As said, the above do not consider them.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.