Towards a Stable Nash Equilibrium of the Game of Global Covid Lockdown
The research report attempts to develop a game theory based model for the payoffs of the lockdown policy in the viewpoints of the government and the general public, and also an economic model containing the demand curve for the vaccine along the timeline. Below discussion will focus on the former, since it’s more course-related. The model is built on various prerequisites, with the key ones including both sides (the government and the public) have two strategies – to adopt the lockdown or reject it – and the government will choose the strategy which result to maximum the public utility and the individual will choose the strategy to maximum the private utility. An quite interesting setting here is instead of quantifying the payoffs, which is actually impossible to calculate, the report chooses to use the qualitative response results such as “individual safe and no spread” and “low marginal social cost”. This lowers the accuracy of the model but does solve the problem of measuring the outcomes towards different strategies. The report suggests that a Nash Equilibrium will be formed when the government withdraws the lockdown but the public still accept and follow the lockdown rules.
Realistically speaking, such an ideal situation is hard to form in the beginning of the pandemic, as people may not follow the lockdown rules without the government’s control. The report suggests the existence of a Nash Equilibrium under some unrealistic assumptions. In the textbook, we have three underlying assumptions for the reasoning about behavior in a game – everything a player cares about is summarized in the player’s payoff, each player knows everything about the structure of the game, and the rationality of the behavior. The model in the report more or less violates all three assumptions, as not all outcomes of lockdown or not can be summarized, not everyone knows about the outcomes, and not everyone performs rationally under the pressure of the pandemic, not to mention that the strategies taken by two sides could have a direct cause-effect relationship, where the enforcement of lockdown by the government could lead to the public sticking to it.
With that saying, the report acts well as a conceptual application of the game theory to discuss the ‘game’ of the lockdown, but it also contains numerous limitations and presumptions when using the game theory based model, as it states. Some of the optimizations that could be made are, to consider the cause-effect relationship between strategies, and to maybe have more players representing the public because it’s impossible for everyone as the general public to make an absolute and the same decision to adopt or reject lockdown. Still, personally, I think for this situation, a Nash Equilibrium model can only be used as a conceptual reference but not an informative simulation. A more data-driven machine learning model might get close to the reality.
Sources: http://hdl.handle.net/10419/227776