Alcohol Abuse, Alcoholism, and Labor Market Outcomes: Looking for the Missing Link

Abstract

There is puzzling evidence that alcohol abuse and alcoholism reduce labor earnings but have no effect on either hours worked or the hourly wage. This study revisits the link between problem drinking and earnings using data from the 1989 and 1994 waves of the National Longitudinal Survey of Youth. Questions about problem drinking were keyed to a table of symptoms for alcohol abuse and alcoholism in the Diagnostic and Statistical Manual of Mental Disorders. The author finds no effects associated with alcohol abuse. In OLS regressions, alcoholism appears to have had negative effects on both labor market outcomes. In the lag variable and in the first difference regressions, alcoholism's negative effect on wages disappears, but its negative effect on hours of work remains, suggesting that the negative effect of alcoholism on earnings operates through reduced work hours. These results of the two-stage least squares are inconclusive.