Bryan Caplan  

Should I Take to Drink?

The Elasticity of Applications... Accreditation of Colleges...

I'm the kind of patient who tries doctors' patience. A memorable check-up from a few years ago (before Robin Hanson convinced me they were a waste of time):

Doc: Do you smoke?

Me: No.

Doc: Do you drink?

Me: No.

Doc: Not at all?

Me: Not at all.

Doc: Good!

Me: Is it? Isn't there evidence that moderate drinking is good for you health?

Doc: Uh, yes.

Me: Then why did you say "Good"?

Doc: Well, heavy drinking is bad for you.

Me: But you said it was good that I don't drink at all.

Doc: [awkward silence]

Me: Do you ever advise a patient to start drinking?

Doc: [fear of lawsuit in her eyes, channeling Darth Vader at the end of Episode III] Noooooo!!!

Since I turn 35 in a week, I'm once again pondering whether I should take to drink. The last time I thought about it, I looked around for a good survey of the medical literature, but didn't have much success. I do recall reading a skeptic saying that "moderate drinking is good for you" research bundles together lifelong teetotalers with recovering alcoholics. Seemed plausible, but the skeptic didn't really have any data.

But now, just in time for my birthday, there's a new meta-study backing up the skeptic. Lucky me!

Researchers at UCSF pored through more than 30 years of studies that seem to show health benefits from moderate alcohol consumption, and concluded in a report released today that nearly all contained a fundamental error that skewed the results.


The common error was to lump into the group of "abstainers" people who were once drinkers but had quit.


Fillmore and colleagues from the University of Victoria, British Columbia; and Curtin University, in Perth, Australia, analyzed 54 different studies examining the relationship between light to moderate drinking and health. Of these, only seven did not inappropriately mingle former drinkers and abstainers.

All seven of those studies found no significant differences in the health of those who drank -- or previously drank -- and those who never touched the stuff. The remaining 47 studies represent the body of research that has led to a general scientific consensus that moderate drinking has a health benefit.

Fillmore's team of researchers took their initial finding one step further, and introduced the error into the data compiled in the seven studies and, voila, the results changed to show drinkers had better health than abstainers.

So for now, my birthday present to myself is to stick to delicious IBC root beer.

Comments and Sharing

TRACKBACKS (3 to date)
TrackBack URL:
The author at Foolippic in a related article titled I'll Drink to That! writes:
    EconLog, Should I Take to Drink?, Bryan Caplan: Library of Economics and Liberty Researchers at UCSF pored through more than 30 years of studies that seem to show health benefits from moderate alcohol consumption, and concluded in a report released... [Tracked on April 3, 2006 8:58 AM]
COMMENTS (22 to date)
James writes:

"Since I turn 35 in a week, I'm once again pondering whether I should take to drink."

As soon as you are old enough to run for president, you start to consider boozing it up?!?

jaimito writes:

Happy Birthday! Till 120!

Now, you seem to have missed the last paragraph of the article you quote:

In a study published in May 2005 in the American Journal of Preventive Medicine, Naimi and other CDC colleagues found that... non-drinkers tended to be poorer than drinkers, had less access to health care, and had less healthy diets.

Moreover, you are missing all the fun.

Robert Schwartz writes:

Meta-studies, doctors with computers, makes me think about having a drink and I have not even had my coffee yet.

"There are more drunks than doctors, a livin' today".

Funny conversation.

Happy birthday in advance, and IBC rootbeer is good. But I prefer Sprecher root beer.

(But I'm not sure about it's availability outside of the midwest - it's brewed in Wisconsin. I've never seen it available here in the DC area since I came out to GMU this year.)

Zac writes:

Bryan- The sugar in the root beer is probably worse for you than the alcohol in the beer. High fructose corn syrup is linked to wide variety of ailments. So, as a replacement for root beer, regular beer would do you a world of good regardless of the skeptic meta-study. Although there are a number of other drink that might be even better than beer.. but then you're sacrificing euphoria for health.

MjrMjr writes:

Zac speaks the truth. High fructose corn syrup is one of the worst things you can put into your body.

Tyler Cowen's former colleague Art DeVany agrees, favoring regular beer instead:

Zubon writes:
All seven of those studies found no significant differences in the health of those who drank -- or previously drank -- and those who never touched the stuff.
Which would suggest that moderate drinking does not significantly hurt, but that the "quit drinking or die" crowd is in a much worse state.
non-drinkers tended to be poorer than drinkers, had less access to health care, and had less healthy diets.
And this would be my usual complaint: not taking account of wealth effects. Anecdotally, we expect people who have a glass or two of wine with dinner to be of a higher economic class than average.

I should go read the full meta-study, but I know that I am just going to exhibit massive confirmational bias for my own teetotaler status.

Zac, high fructose syrup "in moderation" is probably good for you. Certainly if you have no other source of carbohydrates in your diet. It's strange that the sugar companies haven't capitalised on this simple fact.

Alcohol use is also linked to a wide variety of ailments. "Excessive" use can cause a wide variety of cancers, and also problems with the liver, kidneys and brain.

One thing the alcohol industry has in common with the tobacco industry is that they are completely economically dependent on harmful use. "Moderate" use doesn't cut it, if everyone drank only in moderation the companies would lose money compared to today. That's one good reason not to support them, IMO.

Zac writes:

Harold, there are a plethora of carbohydrate sources other than high fructose corn syrup. All sugars should be consumed sparingly. High fructose corn syrup, in particular, is linked to obesity and adult-onset diabetes.

One thing the soft drink business has in common with the tobacco business is that it is completely dependent on harmful use. If that's not reason enough not to support them, consider that by using HFCS instead of sugar, taxes (by way of corn subsidies and sugar tariffs) increase their profit margin.

Actually, I wouldn't be surprised if the big winner wasn't the soft drink companies but the manufacturers of HFCS. In other countries without excessive agriculture taxes, sugar cane is still preeminent as a sweetener. Not only does it taste better -- its better for you.

I see the Accidental Hedonists has more, today, on the evils of soft drinks.

Robert Speirs writes:

Why is including former drinkers an "error"? That seems to assume the conclusion that drinking or ever having drunk is bad for you. Wasn't that the question that the study was examining? If moderate drinking really is good for you, those who used to drink moderately and then stopped would be healthier than those who never drank.

Zubon writes:

Robert Speirs,
I believe the assumption is that those who quit had not been moderate drinkers, but rather heavy drinkers. Alcoholics seem unlikely to successfully convert from heavy to moderate drinking, so it is all or nothing. People who have an occasional drink probably continue to have an occasional drink.

Your point is well taken. An ideal analysis would compare several groups: heavy drinkers, moderate drinkers, former drinkers who have quite (possibly again divided between heavy and moderate), and teetotalers. Grouping the recovering alcoholics with the moderate drinkers does not make much more sense than grouping them with the lifelong teetotalers. I suppose that they might average over time to moderate drinking, but that is not nearly the same thing.

Timothy writes:

High fructose corn syrup is linked to wide variety of ailments.

Being that table sugar (sucrose) is chemically just fructose and glucose, HFCS is usually 42% or 55% fructose (it is also made at 90% fructose), with the rest of the sugar being glucose. So, roughly, HFCS contains the same amounts of fructose and glucose as refined table sugar.

The one study linked in the wikipedia article on a link between fructose and insulin levels seems interesting, but thinking about the simple chemistry of the matter makes me pretty skeptical.

Plus, if you look at the sugar make up of regular old 100% orange juice you're getting basically 9.31-11.3 g/100ml*, with the breakdown 3.03g fructose, 2.22g glucose, and 5.25g sucrose (which is 50/50 fructose and glucose). That means that OJ's sugar content is, roughly, 50% fructose: about the same as HFCS, depending. OJ has nutrients, soda doesn't, but if the fructose paranoia is correct you could expect have OJ drinkers to end up diabetic, and fat as well. Or it could be that increased caloric intake in general combined with a lower level of physical activity is what's making people fat.


Jonathan writes:

The obvious implication is that once you start drinking, you shouldn't stop, since the former drinkers are lowering the health standards of the abstainers. Once you satr, don't stop... now there's a principle I can live with.

MjrMjr writes:

High fructose corn syrup causes a big insulin surge when you consume it. Your blood sugar level goes up rapidly, then your insulin response kicks in big time to push the blood sugar back down. This up and down rollercoaster for people who consume lots of soft drinks majorly saps energy levels and is a contributor to diabetes in the long run.

HFCS is better than no carbs, sure. Harold, they have capitalised on it. Witness Gatorade, Powerade, and all sorts of other "sports drinks". But these sugary drinks are still relatively inferior to any number of snacks you could have instead if the question is one of a quality energy source.

Zac, I always thought it was the agricultural subsidies the US has which made HFCS so common. Doesn't corn get huge subsidies?

Kevin Postlewaite writes:

Your error is to lump those who drink root beer along with abstainers and assume that root beer is healthier for you than to not drink alcohol, which I do not believe has ever been shown.

L writes:

Zac: You seem to claim that drink for drink sugar soda is worse for you than beer. I am skeptical. My impression is the overall effect of sugar soda is worse because people drink so much more of it.

MjrMjr: I think sugar import quotas are more important than corn subsidies in encouraging the use of corn syrup. Corn subsidies affect the world-wide price of corn syrup, but sugar quotas only affect the US price. Other countries are subject to the corn subsidies, but don't use corn syrup. This isn't conclusive--it only demonstrates that corn subsidies alone aren't enough, but I find it pretty convincing.

Zac writes:

MjrMjr, corn does get huge subsidies. Sugar also has huge tariffs/quotas in the US. I am not sure which influences the prevelance of HFCS in the US more, but both are at work. "Agriculture regulation" would have been better than "agriculture taxes" in this case.

L, well, if HFCS contributes to obesity and diabetes, and there is inconclusive evidence as to whether beer actually extends your life or not, I think its fair to say that sweetened sodas (which contain a ton of Calories too, and cause a severse spike in insulin, if you don't buy the HCFS is evil argument that is enough of a health risk) are worse for you than beer, drink for drink. Even so, the point that people drink far more of them is important: these sodas are not at all filling or thirst quenching. Beer, on the other hand, contains very little sugar and tends to be very filling (so much so that "less filling" is considered a good quality to some).

Its a lot easier to drink beer in moderation than soda.

Interesting discussion. Zac, I do agree that the sugar industry has perverse incentives as well.

But I do wonder about your assessment that it is easier to drink beer in moderation than soda. People drink a lot of beer, and alcohol in general. And as alcohol also contributes to obesity and obesity-related problems (it has an energy value of 29 kJ/g), my bet is that Caplan is better off sticking to his root beer.

Zac writes:

I think its pretty clear that far more calories come from soft drinks than alcohol in the American diet.

If you drink 1 beer a day that's better than 1 root beer a day, in terms of calories (12oz of Guiness has 126 calories. A 12oz can of coke has 155 calories) and in terms of the harmful effects of sugar/HFCS (well documented) versus the effects of moderate drinking of alcohol (inconclusive evidence on actual life extension, safely consumed by humans for thousands of years). Of course, if you only drink 1 root beer a week versus 1 regular beer a week, I doubt there's any difference at all.

Also consider that the average size for a beer may be 12oz, but sodas are typically 16oz or 20oz. These beverages really are poison.

The calorie counts are interesting, I think they make my point rather well - I hadn't thought they were that close.
I knew there are fewer calories in beer than in soda, but I meant that the damaging effects of alcohol more than makes up for it. In "moderation", the health effects are small anyway.

However, the cost of alcohol damage in a society is propotional to total use, not number of heavy users. Moderate drinking is harmful in practice, because alcohol-related accidents ("trauma" in the literature) are common in moderate drinkers as well. Also, it seems that when median consumption is low, high consumption is rarer as well. That's reasonable, when you think about it: The more the people around you drink, the easier it is for you to drink to excess as well. And conversely, if everyone you know is like me and Caplan, you will feel a little weird if you drink a lot.

That's one of the reasons I don't drink at all. I figure it's the best way to reduce harms related to heavy drinking (and moderate, too), especially in those closest to me.

But it would be hard to apply the same thinking to sugar. Alcohol is not really very useful, and easy to do completely without (ask me!), but sugar isn't. Also, part of the preventive effect of alcohol abstention comes from the moral challenge it poses. I believe that most people who use intoxicants feel at some level that they ought not to (whether that feeling is reasonable or not).

There is no similar idea about sugar, at least not very wide-spread. While people might feel that the sugar they eat is bad for them, only a few dieters feel genuinely guilty about it. Why should they? In one form or another your body needs sugar.

If people can be convinced that alcohol is beneficial for their health, however, it will come in the same category. They can more easily convince themselves that they're not just using it as an intoxicant. Consumption will then rise, and damages and overuse as well, for the reasons I've talked about.

So even though sugar is undoubtedly harmful in some similar ways, telling people to switch to beer is a bad idea, in my opinion.

tripp writes:

I'm confused by their results, but maybe it's just because I'm a statistical dunce (or I read something wrong), but, if this is true:
"All seven of those studies found no significant differences in the health of those who drank -- or previously drank -- and those who never touched the stuff."
I condlude: health of drinkers=past-drinkers=lifelong abstainers (maybe not truely equal, but differences were not statistically significant).
If this is true, then how does subsequently lumping ex-drinkers with abstainers suddenly make them significantly less healthy than drinkers, unless both ex-drinkers and abstainers were both less healthy separately (but just not significantly so)?

Daniel writes:

You should have picked up a taste for Stewarts Root Beer when you were at Princeton. Nothing like it.

Comments for this entry have been closed
Return to top