The BLS job creation numbers are not accurate because they are largely a function of the birth-death model, which makes assumptions in the current month that may not be true and get revised later—so employment looks stronger than it actually is, particularly for recent college graduates.