Maybe BP has some numbers on that? Your supposition would hold if output increased less than the percentage of "new workers" entering the workplace.