A comment of the Ashenfelter analysis of labor-market fluctuations