Sweden exception to general intervention effects

[jscire]

Congratulations for the awesome work on this project.

As confirmed by email by Seth Flaxman, the effect of the ban on public events on R(t) in Sweden is much higher than it is for other countries. The reason for this does not seem to be documented.
How does this work in your method?
Are there other exceptions (besides country-specific lockdown effects) to the rule that interventions have an identical impact in all the countries studied?

Here the R(t) timeline with the minor effect of the ban on public events in Switzerland

And here the major effect (similar to a lockdown) in Sweden