MORTALITY AND UNEMPLOYMENT: A CRITIQUE OF BRENNER'S TIME-SERIES ANALYSIS