Thread by @jeuasommenulle

This chart is really intriguing.
It looks like:
1) Around mid-September the rising trend reversed in Spain
2) Around mid-October, Madrid diverged from the rest of the country.
Can we explain this with mobility data?

Time at work indeed suggests a different trend for Madrid and the rest of Spain since mid-august (presumably more office work in Madrid that can be done from home?)
But the change is not October

Slight difference in time spent at home, especially as of mid September ; but hard to see the explanation there

Transit looks like the best explanation : end of September (which is consistent with the 10 day kind of lag expected for measures to kick in) one can see a divergence between Madrid & rest of Spain - which suggests this is more important

But to be honest, none of this looks like a clear explanation for such a difference - which means (& I think that's what is important here) that targeted measures can work, not only massive "stay at home" orders.

I forgot retail, where you can't see any real difference apart from the holiday season & subsequent catchup!