> if encrypted information is actually published at all

Given the arguments surrounding searches of "private papers", I'd analogize 'publishing' an encrypted SSN to be somewhat like putting a document in the public square... in a safe.