No causal relationship between alcohol consumption and high earnings: No impact on hours worked

A joint research group comprising Professor Daiji Kawaguchi of the Graduate School of Economics, The University of Tokyo, Professor Jungmin Lee of the Department of Economics, Seoul National University, Professor Ming‐Jen Lin of the Department of Economics, National Taiwan University, and Professor Izumi Yokoyama of the Graduate School of Economics, Hitotsubashi University, examined the impact of drinking on earnings and hours worked and found that those with alcohol tolerance drank more frequently and heavily than those without, but they did not necessarily have higher incomes. The result was obtained using data from adult males in Japan, Korea, and Taiwan, and was published in Health Economics.

Studies have indicated that alcohol consumption may facilitate business communication and raise income, but critics have suggested that the relationship between alcohol consumption and earnings may be a pseudo‐correlation, since those who drink alcohol and those who do not have different personalities and different careers. In this study, the group estimated the impact of genetically determined alcohol tolerance on earnings and working hours.

The research group applied a biomarker test, called the Alcohol Patch Test, which measures the genetic tolerance of subjects to alcohol, to approximately 2,000, 1,000, and 500 working men in Japan, Taiwan, and Korea, respectively. The data showed that about 50‐60% of the subjects were of the alcohol‐tolerant type, whereas the remaining 40‐50% were of the non‐tolerant type. This distribution is very close to that reported in a meta‐analysis of medical studies based on genomic analysis.

The data analysis revealed that men who were alcohol tolerant drank more frequently and more per occasion than men who were not. The results are consistent across the three countries, confirming that they are robust to controlling for individual characteristics. The result reaffirms the correlation reported in previous studies that those who are alcohol tolerant indeed drink.

Next, the group compared the earnings and hours worked of men who were alcohol tolerant and those who were not and found no statistically significant differences. Not only was the difference not statistically significant, but it was also negligibly small. Overall, the findings indicate that alcohol tolerance has no impact on labor market outcomes.