There seems to be a pretty heavy correlation v. causation problem here because your two primary examples are situations where the shut downs occurred in part because everything else was already falling apart a fair bit.