Best practices for conflicting EFFDATEs within a PID in the Verisk dataset

Hi,

I am working with the Consumer Address History data from Verisk. There are many (non-trivial amount) observations with the following pattern:

same PID (same individual)
exact same EFFDATE
different Address/ADDRID

Why would that happen and are there research best practices for dealing with this? Concretely, what should I do if a person is first seen at two different addresses at the same time? This seems to happen way more often than we'd expect if this was only due to people with multiple homes.

Thank you in advance!