The last months have seen an explosion of research papers on the pandemic (to which I have done my small contribution, with another one coming up shortly). Herewith three recent ones I liked especially, two of them
by Thiemo Fetzer (who has achieved certain fame among political economist by linking austerity policies in the UK to the Brexit vote):
In Subsidizing the Spread of COVID-19: Evidence from the UK’s Eat-Out-to-Help-Out Scheme, he shows that government subsidies for the hospitality
sector in England (50% of food and non-alcoholic beverages on Monday through Wednesday in August, up to 10 pounds per client and order – and yes, yours truly also benefitted from it) are responsible for an increase in new COVID-19 infection clusters.
He does so by linking the number of participating restaurants in an area to mobility and restaurant booking data and COVID infection data, comparing Monday-Wednesday in August to the same weekdays in July and September. To control for any simultaneity
bias, he uses rainfall data (during lunch and dinner hours) to show that rainfall dampened restaurant bookings and COVID infections during Monday-Wednesday in August but not the months before and after and not Thursday-Sunday in August (weekdays with no subsidies).
His estimates suggest that the subsidy scheme may have been responsible for around 8 to 17% of all new detected COVID19 clusters emerging during August and into early September in the UK. If this does not sound too dangerous, remember that due to the
exponential growth of infection, this economic support programme might very well have contributed to the second wave (on 10 August, a week after the programme started there were 826 new cases in the UK, while there were 2,948 on 7 September, a week after the
programme ended). While some might argue that this shows the trade-off between economy and public health, that would certainly be penny-wide and pound-fool, given the second lock-down we are currently in, plus the enormous economic (not to speak of human)
costs of COVID patients.