> Sometimes you are wrong, and everyone else is wrong,

Happens all the time.

> and only an empirical test can separate the wheat from the chaff.

Not for the vast majority of political issues and indeed for most of Social Sciences. In these cases, empirical evidence is just an accessory, it's still evidence but it's never conclusive, you need reasoning to sort out the complexity.