Working Paper: NBER ID: w11902
Authors: Philip J. Cook; Bethany Peters
Abstract: Drinkers earn more than non-drinkers, even after controlling for human capital and local labor market conditions. Several mechanisms by which drinking could increase productivity have been proposed but are unconfirmed; the more obvious mechanisms predict the opposite, that drinking can impair productivity. In this paper we reproduce the positive association between drinking and earnings, using data for adults age 27-34 from the National Longitudinal Survey of Youth (1979). Since drinking is endogenous in this relationship, we then estimate a reduced-form equation, with alcohol prices (proxied by a new index of excise taxes) replacing the drinking variables. We find strong evidence that the prevalence of full-time work increases with alcohol prices - suggesting that a reduction in drinking increases the labor supply. We also demonstrate some evidence of a positive association between alcohol prices and the earnings of full-time workers. We conclude that most likely the positive association between drinking and earnings is the result of the fact that ethanol is a normal commodity, the consumption of which increases with income, rather than an elixer that enhances productivity.
Keywords: alcohol consumption; earnings; productivity; labor market
JEL Codes: I11; J24
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| Higher alcohol prices (H29) | Increased full-time work participation (J29) |
| Increased full-time work participation (J29) | Reduced drinking (I12) |
| Higher earnings (J31) | Increased alcohol consumption (L66) |
| Drinking (L66) | Increased productivity (O49) |
| Drinking (L66) | Quality of labor supply (J24) |
| Drinking (L66) | Quantity of work offered (J29) |
| Binge drinking (I12) | Earnings decline (J31) |