I’ve decided that my final guest blog will be a response to some of the posted comments, rather than a new essay.
Thanks to those of you who took the time to read some of my posts, and to keep a cordial tone in your comments. I learned something from reading them, especially about home brewing!
A few of you were interested in knowing more, or having a cite to back up a factual claim. Of course I hope you will consult my book, Paying the Tab. It is quite thorough in presenting the arguments and evidence.
Several comments suggested that I didn’t know the difference between correlation and causation. Actually I believe that I do know the difference. My technical contributions to the alcohol literature have focused on taking advantage of natural experiments to learn the effects of changes in policy. The book explains this matter is detail. It also discusses the evidence on minimum drinking age, discussing two of the issues raised by bloggers -- the state border effects and the effect on the older age group.
I was baffled by comments to the effect that I believe all drinking is bad. My friends would be amused, and it’s surely not what I said in my blogs. Like every other commodity, alcohol has benefits and also costs. The difference for alcoholic beverages is that the costs are not fully reflected in the price. A higher excise tax would help with that problem.
(One great virtue of the price system in a private enterprise economy is that the prices signal relative scarcity and provide an incentive to economize appropriately. But when there are externalities – when property rights are incomplete – the price system does not do those jobs very well without some intervention.)
There were many comments to the effect that taxes designed to change behavior in particular ways are fascist or at least represent an unacceptable imposition on freedom and are certainly no business of government. For what it’s worth, I see alcohol excise taxes as less of an imposition on personal freedom than many other types of alcohol regulations that are intended to limit abuse, including the high minimum age.
A number of comments appeared to take me seriously when I listed some of the options for regulating adverse consequences of drinking – including penalizing DUI more severely etc. The purpose of that paragraph in my third post was not to advocate any of those changes (far from it) but rather to point out that a much-touted alternative strategy to tax increases – penalizing the consequences of abuse – can be costly and oppressive.
Several comments noted that there is evidence that moderate drinking promotes health. So there is. But the main epidemiological evidence is correlational, and very flimsy. Similar evidence has been profoundly misleading in other medical areas, such as hormone replacement therapy.
We’ll probably never do a randomized controlled trial with drinking, and without that it will be very difficult to sort out the causal effects of drinking. Incidentally, just as it is true that moderate drinkers live longer than abstainers, it is also true that moderate drinkers are paid more than abstainers. One speculation (by one of my former students) is that that association is causal, the result of social capital. That’s an interesting idea, but I don’t believe it.
The statistical-inference problem is that people who abstain are different in all kinds of ways (some not readily observable) from those who drink. Differences in longevity and earnings may be the result of those other characteristics, rather than the drinking per se. My advice: Don’t start drinking just because you want a raise in salary or cleaner arteries.
(I’ve read that pipe smokers live longer than nonsmokers on average…)
Cheers!
Sure, it's "difficult to sort out the causal effects of drinking". But that doesn't mean it's difficult to sort out the positive effects and easy to sort out the negative ones.
Was interested in a question that didn't actually seem to come up:
the nature of some addictive and problematic behaviors tends to change over time, with one of the factors being improved delivery systems or improvements in the source "product." For example, marijuana today is much more potent than in the 70's; same with some cigarettes. Yet alcohol, with its thousands of years of development, doesn't seem to have the same trajectory.
It would seem to be a factor in these considerations, as, for example, a more potent delivery system might negate higher taxes or other methods of "control." Did you consider that concept and incorporate it into some of your thoughts? If so, perhaps some of your slower readers, like me, missed it.
(Oh, and I quoted your "free lunch" explanation to my wife, who liked it. Thanks for the free brownie points!)
I guess we could save tax money by drinking Everclear, Bacardi 151, or Stroh 80% (or even 60%).
Reminds me: In Prof. Cook's comparison to the liquor taxes of the 50s, he neglected to mention that the standard bottle of booze was 86 proof back then, not today's 80 proof. (Note that US proof is double the alcohol percentage.)
Of course, if 50's tax rates were reasonable, we could always bring back the IRS schedule of the time. The postwar boom coincided with 90% marginal tax rates, so obviously higher taxes are better for the country.
I believe the answer is: "No, it just seems longer."
I wasn't saying that you didn't know the difference between correlation and causation. I just said I wanted to see proper analysis before I would accept the causative link. I.e. simply saying "states which raise the taxes showed reduced consumption and consumption indicators" doesn't tell me enough to know whether this was specifically the result of the tax increase. For example, such might have a strong effect in a place like Utah, and a weak effect in a place like Washington because of the availability of home-brew supplies. Moreover, the lack of home brew supplies in Utah may say something about the culture as regards alcohol which may make it susceptible to a price increase.
A second area that would be interesting to see would be clear analysis of policy regarding moonshining. My own thinking is that there are some legitimate public health and safety interests in regulating the practice, but that we may be better off with state licenses which are taxed by the government than we are with a full prohibition.
The legitimate interests are still materials contamination (usually lead), fire hazards (making a liquid with the flamability of gasoline around heating elements), etc. There is also a minor question of methanol control but this has been substantially overblown and is as bad or worse with wine-making which has regulatory exceptions for home wine-making.
Is the cost of the death caused by Zack an externality of the purchase of alcohol? If so, why alcohol specifically? Why not the purchase of a car? Why is it not the purchase of gasoline which resulted in the death? Why not the purchase of a driver license? Why not the purchase of the shoes he used to walk around in the bar in while drinking and to push the gas pedal? Why not tax shoes?
There are no negative externalities whosoever when Amy and Bob purchase alcohol. Any tax placed on Amy’s or Bob’s purchase could only have the effect of reduced economic efficiency. Agreed?
The efficient price of alcohol for Amy and Bob is the unadjusted market price. For a drunk driver like Zack, the efficient price may be something like $50,000 per drink.
Pigovian taxes have serious problems even in the “best” cases, when harm is caused a little bit by everyone consuming, like pollution. In this context they are useless. What Professor Cook advocates is setting prices significantly higher (50%? 150%?) than the efficient price for the vast majority of purchases in order to move the price 0.01% in the direction efficiency for a small group of customers.
One way to account for this is to tax on an absolute alcohol basis, rather than imposing a flat tax amount per bottle or can.
It seems unfair to tax me for that failing of politicians.
I'll admit I've not read all of Mr. Cook's posts, but the above quote from today's post gives me great caution. Simply put, the above reflects an opinion on the research about the health effects of moderate alcohol consumption so deviated from what is generally accepted in the medical field that it makes me question how we can trust in Mr. Cook's intellectual integrity or his actual knowledge.
There is massive body of peer reviewed literature about the effects of alcohol and health, stretching back over a century of investigation, with decade long studies involving hundreds of thousands of participants, and the molecular mechanisms that may underlie this phenomenon are being elucidated. The cardiovascular benefits of moderate alcohol consumption, especial in men, can scarcely be described as "very flimsy".
The proponents of moderate alcohol consumption's benefits are unequivocal about the fact that alcohol consumption carries risks that must be weighed and that even moderate alcohol consumption is innapropriate for some individuals.
So, given all that, what are we to make of Mr. Cook's claim that the, with regards to the health effects of moderate drinking, "the main epidemiological evidence is correlational, and very flimsy"? Either,
1- Mr. Cook is engaged in tremendious obfuscation, because he does not want to admit the validity of a contrary point, in which case we cannot trust his intellectual integrity, or
2- Mr. Cook is simply unaware of the relevent information, or unable to properly digest it, in a field of medical research closely related to his field of inquiry, in which case his ability to honestly and accuratly be an advocate for his cause is suspect.
3- Mr. Cook is correct and the edifice of the medical establishment is incorrect; statistical studies misleading, evidence from autopsies misinterpreted, and the molecular mechanisms simply wrong.
Given his eloquence and position, item 2 seems rather unlikely. Without, shall we say, substantial additional information, item 3 cannot be accepted without engaging in a parania about the medical establishment that simply a given.
Which leaves us with item 1; that Mr. Cook is engaged in obfuscation with regards to a point he finds inconveneient. Further supporting this is his mention that "Similar evidence has been profoundly misleading in other medical areas, such as hormone replacement therapy." What does he mean by this? HRT has risks, but also has well established benefits. The issue with HRT is comparing the risks, which were not always appreciated, with the benefits. Using this as an example of the "very flimsy" state of "correlational" evidence is profoundly misleading, given that the issue with HRT was a lack of understanding of the risks, whereas, as I have noted above, the dangers of alcohol consumption are manifestly clear and repeatedly emphasized by the advocates of the health benefits.
Can all of Mr. Cooks points be so disposed of? I don't know, but judging by his treatment of a discrete point of medical information can scarcely make us sanguine for the validity of his more rarefied points.
I don;t think all of the points can be disposed of. For example, consider the effect of raising the federal excise tax on spirits by $10/fifth. WHo would bear the brunt of this?
Consider: 1 bottle of Maker's Mark Burbon goes for $27
1 bottle of Laphroaig 12yr single malt scotch goes for $40
Scotch drinkers can switch to burbon with no net expense change, and the increase on a bottle of Laphroaig is about 25%, while the increase in price on the Maker's Mark is a bit under 50%. If you get into American Blended Whiskey, the price increase would be quite a bit more significant, and the price increase of MacCallan 55yr would be less than 1%. Hence the actual increase of expensive liquor is marginal, but the increase of cheap liquor (affecting those who are more likely to abuse it) is more substantial.
However the bigger issue happens with wine. I frequently buy inexpensive wines for cooking purposes. If a $3 bottle of wine goes to $13, I am more likely to only buy more expensive wines when I want to cook with them because the percentage increase isn't that big and then I can drink it with dinner. I would probably cook things like buffalo wings less often, but when I would, the quality of the wine I would use would go up. (The difference between a $3 bottle and a $10 bottle is much greater than the difference between a $13 bottle and a $20 bottle.)
So there might be some merit to at least increasing the tax on spirits. I am not sure I go as far as to advocate the same tax increase on wine as a rule. However, if one were to increase it to, say, $3/bottle (750ml), that wouldn't be too bad, I guess.
I still think this is better left to state governments, which tend to be more accountable to their constituents. However, I am willing to see a discussion of the policy on the state level (though not the federal level) as useful.
I'm not taking your book out of the libarary, let alone buying it, if you can't suggest out how harmless alcohol use shares the externalities of alcohol MISuse (as road2Serfdom elaborated upon above).
Why? Because the samples used in those studies (and multiple studies have confirmed each other) are so large, and because the magnitude of the impact (approaching 50% reductions in heart disease which is cumulative and not instead of the benefits of an asprin regime) is so great. A low single digit percentage benefit could be a statistical fluke. A 50% reduction in cardiovascular disease is extremely unlikely to be a fluke.
There has also been some lab work show chemically how the casual connection arises.
The cardio-vascular diseases impacted by moderate alcohol use are so high on the top causes of death list that the raw numbers in the studies are high enough not to be statistical flukes, something not the case in therapies that impact very rare diseases.
Finally, the "natural experiment" methodology that you use yourself is what prompted the research into the benefits of moderate drinking in the first place. The "French paradox" whereby the French have immensely high consumption of high fat foods, but don't suffer the predicted rate of cardiovascular disease has largely been explained by widespread regular alcohol consumption in France.
Moderate alcohol consumption has also been a consistent thread in almost every study to examine the defining characteristics of people with great longevity.
The significant health benefits of moderate alcohol consumption is one of the heaviest items in alcohol consumption's favor on the econometric scales.
I also don't see that he adequately answers the point about selective discouragement of excessive consumption. If I read him right, he thinks it's too costly to enforce. I don't think that's the real problem -- rather, it's that such enforcement may conflict with American sensibilities about individual liberties. In Sweden, a real bastion of social control, DUI laws are enforced with almost extreme vigor, and punished very severely by the courts with fines, license suspension, and jail. At the cost of massive forced stopping of cars on the highways and breath analyzing of all drivers randomly, they've driven DUI down to very low levels -- at the cost of creating a police state atmosphere every evening. We could do the same for severe alcoholics who destroy family life and can't hold down jobs by legalizing their institutionalization due to social incapacitation. That may actually improve collective welfare, but at the cost of setting unacceptable precedents for other social engineering interventions. As with everything, we must choose our poisons.
The bottom line? The proper weighing of all these issues is ultimately a political problem, and much too complex for selective and simple cost-benefit analysis, especially when there are serious public choice issues that prevent collective preferences to be properly represented in our Congressional system -- Arrow showed us decades ago that social welfare is a meaningless concept.
Still, I don't sense that the polity in America have a great dissatisfaction with the status quo. So let's leave well enough alone. Perhaps we have a reasonable balance of the various market externalities as it is; in fact, that's what I believe.
I agree with AN Kahn that this is a deeply dishonest presentation and deserves the highest level of skepticism.
I'm not sure that Cook's or the government's view of the benefits of alcohol are relevant. Since the information is public, all of that benefit will be reflected in the price that people are willing to pay. The negative health problems wouldn't matter if health care was paid for privately. Finally, taxing all consumers of alcohol based on the drinking related accidents caused by a few is like a car tax for everyone to offset the cost of speeding by a few.
I think cost/benefit is the barrier to entry where an item is not worth a great deal of continued discussion. If someone can't make a compelling argument that we would be better off if we go a particular direction, why should we even consider the proposal?
However, at the same time, I agree with the above statement. We do need more discussion and thought about a lot of this area, and ultimately, the only governments that can and should be addressing the issue are those of the states. The federal government, IMO, better serves us by staying out of this debate.
No, that analogy only applies if the tax incents bourbon drinkers to switch to scotch.
I said I supported a dialog on the subject of state excise taxes, but that I felt that it was wrong to put it on the federal level.
I still think it is the wrong way to go, but I am open to being wrong and think dialog at the state level is where it should be addressed.
My main point was that the more expensive the drink, the less the tax margin. As it turns out a bottle of McCallan 55yr would probably increase about 0.1% while Calvert Extra American Blended Whiskey may increase around 100%.
If people argue that this is not a tax on alcoholism because we aren't disproportionately taxing low-end products, they are wrong.
Also I will concede that it might cause bourbon drinkers to switch to Scotch, because the price difference percentage-wise would not be as high. The difference between a $40 and a $50 bottle of whiskey is psychologically less than the difference between a $30 and $40 bottle.
I don't think I had read that before. Thank you!
As someone who brought up the positive health impacts of alcohol in earlier threads, I was waiting to see if Mr. Cook was brave enough to tackle the issue. He was not. A total cop out. You can tell when someone has no argument when they deflect. Whatever medical results there are about HRT, it has nothing to do with the subject. Why is it even mentioned?
As noted excellently by AN Khan, the evidence is massive and one sided that alcohol has significant health benefits. That the savings to Medicare alone argues for subsidizing alcohol. More deaths are prevented by alcohol consumption than are lost.
This is simply untrue. The medical literature is, in fact, very good. To dismiss it out of hand reveals a political agenda that ignores medical fact. The medical literature is no longer discussing *whether* there are benefits to moderate drinking -- the J curve is well established -- but instead is now focusing on determining the exact mechanism.
That there are benefits to moderate drinking is no longer a matter of debate in the medical community.
Because Prof. Cooks analysis ultimately depends on ignoring the actual medical facts when weighing the risks and benefits of his social engineering policies, it falls flat on its face.
I believe alcohol beverages should be regulated, but the difficulty of arranging and enforcing that regulation is intensified by the poor quality of debate and by fading memories of the public policy purposes underlying many elements of a regulatory system that dates to the mid-1930s.
The poor quality of debate appears to derive in large part by the tendency of many to disregard alcohol-related evidence that contradicts their preferred conclusions. The alcohol industry too often ignores the problems associated with alcoholism, consumption by minors, drunken driving, and similar alcohol-related conduct. It engages in irresponsible behavior in formulating and marketing its products. It chafes against reasonable regulation. Those who oppose alcohol -- for religious reason or because they consider alcohol sinful in a secular sense -- are no better. They ignore the scientific evidence of benefits associated with most consumers' enjoyment of alcohol beverages. They take it on faith that alcohol is evil, and expect others to accept the teetotaler's 'demon rum' beliefs on the same faith.
After a couple of decades of observing a poorly informed legal and regulatory system wrestle weakly with issues that deserve better treatment, some of my thoughts can be distilled:
(1) Alcohol beverages are healthful and enjoyable products for most people who consume them. Anyone who would deny others that enjoyment and benefit doesn't care much about freedom, science or public health.
(2) Alcohol beverages require (but currently do not have) well-formed and -enforced regulation. Those who would treat them like soft drinks or milkshakes are dopes.
(3) Drunken driving should be discouraged vigorously and punished severely. Drinking and driving should be disassociated entirely for those under 21 (or perhaps 25), by strict rules and severe punishment.
(4) The drinking age should be reduced, perhaps to 18 or to the date on which one begins (or would customarily begin) full-time undergraduate studies. Parent-supervised consumption could begin earlier.
(5) Most states' regulatory systems should be overhauled.
(6) The federal government, whose long-fading oversight of the alcohol industry nearly disappeared in the 'national security' fever of recent years (with a hand from an anti-regulation dogma that couldn't distinguish pro-business from anti-regulation), should recreate a substantial role in regulating alcohol.
(7) The lack of academic interest in alcohol beverage law has puzzled me for many years, as has the poor quality of the meager scholarship I have encountered in this area.
So at the moment, it is reasonably established that, for most people:
1) Moderate drinking reduces key forms of death
2) Moderate drinking increases life expectancy
3) We are developing some understanding of how these effects come about.
I think that at the moment, anyone who says that moderate drinking benefits are not well established is not paying attention to the studies in the area.
This doesn't mean that there are no negative consequences to alcohol. There are plenty. But I still find Professor Cook's claim rather remarkable.
Perhaps alcohol is the worst example, so it's odd that he'd suggest starting there. Alcohol almost certainly has benefits in small quantities and for the vast majority of consumers, there are no negative externalities.
It would be much smarter to start with candy, hamburgers, skiis, motorcycles, and televisions.
For more than a decade, I have reviewed literature concerning the effects of alcohol consumption on health. I have discussed the issues with physicians and researchers. I have debated the issues with those who disliked alcohol consumption to the point at which they believed it necessary to warn children, restrain adults and enlist the American Bar Association in those efforts.
The physicians and scientific literature I encountered tend to disagree with the professor's admonition, to the point at which my wife, after reading some of the research reports I left on the kitchen counter 10 or 15 years ago, drinks a beer most nights in the manner one would take a vitamin pill.
Some vitamins are toxic in doses that can be mistakenly arranged, yet I rarely hear anyone recommend teetotalling or advise against healthful doses of Vitamin D. Few counsel against starting to consume carbohydrates, despite the vivid demonstration of the devastating effects of overconsumption.
Why does alcohol provoke such repressive instincts? Why does the United States, which self-congratulates ostentatiously concerning its devotion to freedom, have so many "dry" municipalities in which an adult can not lawfully purchase a bottle or wine or beer? Why is alcohol classified with tobacco and crystal meth in some school lessons (with any mention of healthful attributes of alcohol banished entirely)? Why has our government fought advertisements and labels bearing truthful health-related information concerning alcohol beverages?
I suspect the answers have little foundation in reason or evidence. What is the motivation?
If you have a comment about spelling, typos, or format errors, please e-mail the poster directly rather than posting a comment.
Comment Policy: We reserve the right to edit or delete comments, and in extreme cases to ban commenters, at our discretion. Comments must be relevant and civil (and, especially, free of name-calling). We think of comment threads like dinner parties at our homes. If you make the party unpleasant for us or for others, we'd rather you went elsewhere. We're happy to see a wide range of viewpoints, but we want all of them to be expressed as politely as possible.
We realize that such a comment policy can never be evenly enforced, because we can't possibly monitor every comment equally well. Hundreds of comments are posted every day here, and we don't read them all. Those we read, we read with different degrees of attention, and in different moods. We try to be fair, but we make no promises.
And remember, it's a big Internet. If you think we were mistaken in removing your post (or, in extreme cases, in removing you) -- or if you prefer a more free-for-all approach -- there are surely plenty of ways you can still get your views out.