Dr Monscheuer, how did the idea for the discussion paper come about?

There is already a large scientific and economic literature on factors that investigate the spread and containment of SARS-CoV-2. Building on this field of research, my co-author and I asked ourselves how demonstrations like the one in Leipzig in November impact the occurrence of infection. We thought about how the effects of such demonstrations could be measured, since there is, of course, no information about the direct occurrence of infection at such events. During our research, we became aware of the “Honk for Hope” website and the departure points that can be booked there. By means of the regional distribution of the bus connections and the time of the demonstrations, we wanted to estimate the effect these demonstrations had on the occurrence of infection in the regions of origin of the demonstrators using established micro-econometric methods and based on clearly described assumptions.

What did you find out in your study?

The results show a clear and statistically significant increase in the seven-day incidence following the demonstrations in administrative districts in which bus services to demonstrations are offered. The effect is stronger in districts where there are bus connections even in small towns. This supports the causal interpretation of the results, since such bus connections testify to a high demand for journeys. Our results are not driven by different developments in the number of infections in districts with or without bus connections, as we observed parallel trends before the demonstrations. Furthermore, the results are not dependent upon the occurrence of infection in individual federal states, such as Saxony, Thuringia or Bavaria. We also show that they are very robust even in various specification tests. That is to say, we can rule out many alternative explanations for the increase in these districts.

The paper has provoked some vehement reactions. Did that surprise you?

Our paper primarily contributes to international research on this topic. It is embedded in a very active economic literature that deals with comparable questions using similar econometric methods. These studies are predominantly based on events in the USA, and so we are adding new insights to the literature with our results (e.g., Dave et al., 2020).

It is common in our discipline for a discussion paper to be published and made available for scientific discourse before the paper enters into the peer-review process. Since scientific work doesn’t always get that much attention, I was, indeed, surprised by the numerous reactions.

Interview: Frank Aischmann

Further information

Link to the study (PDF)