How are mobility levels correlated with Covid-19 transmission rates?
The authors use a very simple (and natural) approach as a form of first attempt to see how mobility "affects" the spread of the contagion. They run a cross-country panel regression of data on mobility provided by Google and Apple and spread of Covid-19 (using country and time fixed effects). They estimate that a 10% decrease in mobility is correlated with a reduction of the spread (R in epidemiology models) of around 0.05. Naturally one has to interpret the results with caution, but if the correlation is evidence of a causal link, then reduced mobility can play an role in getting an R-value just above 1 to instead be just below (essential in any policy of suppression). The quantitative results of the paper is a key contribution for anyone interested in quantifying epidemic-behavioral models (e.g. for cost-benefit analysis).
Link to paper here.
Posted by: Daniel Spiro
Dept. of Economics