Rebased and revised data -- their effects on econometric estimates