Ugh, here we go again. Three researchers at the University of Florida, led by epidemiologist Alexander C. Wagenaar, have just released a new study which they claim shows that raising alcohol taxes — in fact doubling them — will reduce consumption and cure society’s problems.
The study, Effects of Alcohol Tax and Price Policies on Morbidity and Mortality: A Systematic Review, is to be published in the November issue of the American Journal of Public Health, but was released online last week, as is common for academic journals.
As I don’t have the resources to buy a subscription to every related academic journal, I have to make do with the abstract and what other news outlets write about it. Here’s the abstract:
Objectives. We systematically reviewed the effects of alcohol taxes and prices on alcohol-related morbidity and mortality to assess their public health impact.
Methods. We searched 12 databases, along with articles’ reference lists, for studies providing estimates of the relationship between alcohol taxes and prices and measures of risky behavior or morbidity and mortality, then coded for effect sizes and numerous population and study characteristics. We combined independent estimates in random-effects models to obtain aggregate effect estimates.
Results. We identified 50 articles, containing 340 estimates. Meta-estimates were r=–0.347 for alcohol-related disease and injury outcomes, –0.022 for violence, –0.048 for suicide, –0.112 for traffic crash outcomes, –0.055 for sexually transmitted diseases, –0.022 for other drug use, and –0.014 for crime and other misbehavior measures. All except suicide were statistically significant.
Conclusions. Public policies affecting the price of alcoholic beverages have significant effects on alcohol-related disease and injury rates. Our results suggest that doubling the alcohol tax would reduce alcohol-related mortality by an average of 35%, traffic crash deaths by 11%, sexually transmitted disease by 6%, violence by 2%, and crime by 1.4%.
Those are some pretty specific promises and some pretty specific recommendations, something most academic papers assiduously avoid. To me that’s a red flag about the intentions of this study.
Science Daily covered the study in an article today (thanks to Richard S. for sending me the link) entitled Increasing Taxes on Alcoholic Beverages Reduces Disease, Injury, Crime and Death Rates, Study Finds. Obviously, I’m as predisposed to question such a study as the average anti-alcohol wingnut is to swallow it unquestioningly. And I confess something doesn’t smell right with it. My alky sense is tingling.
Having not seen the full article, I’m left wondering exactly what the “50 published research papers containing 340 estimates” means. What is being “estimated?” It reads like it’s the supposed harm that’s being estimated, because I can’t for the life of me understand how you could ever say there’s definitive causation for such a complex relationship as the price of something to “other misbehaviors,” or indeed any of the laundry list of issues the researchers believe are caused by people drinking alcohol. In my experience at looking at these studies, any event in which there was alcohol present is usually sufficient to consider the incident alcohol-related, but that’s nowhere near the same as having been caused by the alcohol. And so these statistics tend to be inflated and, consequently, misused.
But the key insight into the study came in the very last paragraph of Science Daily’s coverage of the study, where they reveal that the funding for the study came from the notorious Robert Wood Johnson Foundation (RWJF), the godfather of neo-prohibitionist groups. The RWJF funds many other neo-prohibitionist groups, and also sets the national agenda in the anti-alcohol community. That they funded this, and other similar studies, suggests that the answer preceded the study, that is it was designed to support their agenda, its conclusions a fait accompli.
To me this also explains professor Wagenaar’s statement. “Results are surprisingly consistent.” Of course, they would be if you’re looking for a correlation. The same team did a similar study in 2007, Raising Alcohol Taxes Reduces Deaths, Study Finds where they examined alcohol-related deaths in Alaska after beer taxes were raised in the state. That study was also funded by the RWJF. Predictably they found the correlation they were looking for, but this is playing with statistics for incredibly complex relationships. Their simple conclusions seem absurd. They ignore any underlying causes for alcohol abuse or suicide or anything else, for that matter. As almost every study like this I’ve ever seen, “alcohol-related” is a thinly veiled attempt to paint any alcohol use, however responsible or moderate, as dangerous and life-threatening. Beer is not a syringe of heroin, despite these same groups’ attempts to portray it that way.
Mark my words, we’re going to see this study used by groups all over the country in renewed efforts to raise beer taxes in state after state. But the only thing I remember happening when the federal excise tax on beer was doubled in 1990 was a loss of jobs and long term economic harm visited on the brewing industry. I don’t recall seeing any victory parties by the anti-alcohol groups once that doubling cured all the problems they previously ascribed to alcohol. They went right on complaining about all the supposed damage caused by the industry. That’s a real world example of what they want to do having none of the outcomes this new study claims would occur under the exact same conditions.
It would help homebrew shops probably though, but I don’t think it’s worth the cost.
Like I said, Jay, you write a much more coherent rant than I ever could. Thanks.
This study echoes the studies that came out in the late 90s and early 00s about smoking. CDC kept repeating that raising excise taxes on cigarettes would result in a 10% decrease in consumption. The State of Texas bought into it, and in 2001 raised excise taxes on cigarettes to fund education, and revenues fell as people quit smoking. The Prohibitionists know revenue won’t increase, but they keep that , that’s just the hook they use to make it attractive to politicians. Texas schools are still in trouble, and children are still uninsured after Obama raised taxes to fund children’s health insurance. These taxes do nothing but reduce consumption.
Could we please not bring up non-beer topics. It’s one thing if it serves a purpose with the topic, but if it only serves personal agenda, it doesn’t do any justice to advancing the discussion. At worst, it just opens up rants unrelated to what Jay is discussing and pits craft beer drinkers against each other.
And this: http://lewbryson.blogspot.com/2009/06/eerie-rise-in-college-age-drinking.html
It’s all about the estimates, baby. This kind of piece is called a meta-analysis, a combining of studies to overcome small sample sizes, and meta-analysis has real problems when the studies involved are done by different people, under different rules…like this one. I hate to send you to Wikipedia, but take a look at the article on “Meta-analysis”, paying particular attention to the very last paragraph of text (http://en.wikipedia.org/wiki/Meta-analysis):
“If a meta-analysis is conducted by an individual or organization with a bias or predetermined desired outcome, it should be treated as highly suspect or having a high likelihood of being “junk science”. From an integrity perspective, researchers with a bias should avoid meta-analysis and use a less abuse-prone (or independent) form of research.”
Uh-huh. Good one, brother!
Beware, too, of studies that use percentages rather than giving hard numbers. “Doubling the alcohol tax would reduce alcohol-related mortality by an average of 35%.” Right – but what’s the current rate of alcohol-related mortality, and how many actual lives would be extended by doubling taxes? And for how many years? And in a true cost-benefit analysis, you’d also have to take into account the lost pleasure from all that reduced drinking as well. (But, of course, the idea that people actually drink for pleasure is something none of these studies ever takes into account.)