Sure, but the point remains - they say that know it's wrong because they see a representative sample of working age adults? No, we actually don't: There is str

Sure, but the point remains - they say that know it's wrong because they see a representative sample of working age adults? No, we actually don't: There is strong selection bias by definition of the difference between "of working age" and actually "working". The ones unable to work are less publicly visible.