With the benefits of long-term moderate drinking to middle-aged people becoming more apparent, official are beginning to wonder whether an alcohol excise tax would decrease this healthy drinking. In Are Alcohol Excise Taxes Good for Us? Short and Long-term Effects on Mortality Rates, Philip J. Cook, Jan Ostermann and Frank A. Sloan examine the net effect of an alcohol excise tax on mortality rates. They find that the tax has a negative correlation with mortality rates in both the short-term and the long-term.
Already knowing that an alcohol tax increases prices and reduces per capita consumption, Cook et. al first looked at the short term effects of the decreased consumption. Using mortality as a function of short-term variations in excise tax rates, the authors found the tax was significantly negatively correlated to mortality. In the short term, increased alcohol tax lowers all-cause mortality rates.
However, the long-term is more complicate. For youths, any drinking is bad, so net would be to reduce mortality, but it is different for middle-age people. Research shows that moderate drinking for the middle-aged can help prevent heart disease and stroke and acts as an anti-cholesterol drug. The relation between drinking and mortality follows a U curve, with moderate drinking lowering mortality to a certain extent and heavy drinking increasing mortality. The worry with the excise tax is that the lowering of mortality rate from reducing heavy drinking doesn’t outweigh the increase of mortality rate from reduced moderate drinking. Using a regression created from state per capita sales of alcohol, the authors calculated the response to a survey of drinking frequency of different type of drinkers. Because moderate drinkers’ responses are closely related to sales of the state and heavy drinkers’ responses are not, one would worry that state level taxes only affect the margin and do little to change the shape of the drinking distribution.
To calculate the effect of the tax specifically on middle-age people, Cook et. al calculated actual distributions of people across drinking categories and then created a hypothetical world that calculated the mortality rate increase or decrease for a 1% decrease of per capita consumption under three different scenarios.
Scenario 1 and 2 resulted in an increase in mortality rates, while scenario 3 had a decrease. However, the results in terms of number of lives or deaths are so small they are trivial in comparison to the 700,000 deaths of middle-age people per year. The authors thus concluded the possibility of people dying from drinking too little is unfound.
The authors conclude that an alcohol excise tax reduces mortality rates in both the short and long term. The fear that the middle-aged would drink to little is incorrect.
This paper is lacking in information. The authors include almost nothing. There are no equations to explain the regressions, no explanation on the nonexistent equations, no instructions on how to understand the elusive equations. As a reader, I had a difficult time understanding the mathematical work that went into coming up with the results and what exactly those results mean. When trying to decipher the tables for the equations in the appendix, I was lost again. The authors didn’t clarify enough of the details.
While the content was lacking, I though the conclusion was correct. The J curve relationship only has a small part of beneficial drinking and infinity of detrimental drinking. In addition, I don’t believe moderate drinking is a complete preventative measure for heart disease or stroke, or those not drinking at all would increase one’s chances of heart disease to 100%. Moderate drinking is probably one of many things people do to reduce the risk of heart disease, so taking away drinking might not have a huge effect. I would have liked the authors to included more information on the correlation between moderate drinking and heart disease, perhaps also some probability on drinking and abstinence on heart disease. This information might strengthen the authors’ argument.