David R. Henderson  

Jonathan Gruber on Sin Taxes

PRINT
Statist policies in China... ISIS and Reproach...

I found an interesting article on sin taxes written by MIT economist Jonathan Gruber. It's titled "Taxing Sin to Modify Behavior and Raise Revenue." When I want an article for my class on Cost/Benefit Analysis that is written by someone on the "other side," that is, someone who trusts government to impose the right taxes, I'll consider using this piece; there's a lot packed into a short space. In that lot are both evidence and economic analysis. It's also a good example of someone who wants the government to save us from ourselves, that is, make decisions for those of us who would otherwise make bad decisions. So it has reasoning both about externalities and about coercive paternalism.

The article is short--only 2 pages--and so I won't quote from it.

I will, however, raise a question. In one paragraph, under the heading "The Case of Cigarettes," Gruber writes:

Most studies conclude that the social costs of smoking are relatively small and likely less than existing taxes on cigarettes, although controversy remains over the social costs of second-hand smoke. The relatively low social costs can be attributed to the "death benefit": smokers live long enough to contribute to Social Security and Medicare but not long enough to collect benefits.

Later in the piece, Gruber writes:
Within 50 years, obesity is expected to shorten the average life span by at least two to five years. Thus, a large government role in addressing obesity could be justified on the grounds of reducing societal costs and mitigating self harm.

He then writes:
Ultimately, what may be needed to address the obesity problem are direct taxes on body weight.

Question for Econlog readers: Do you notice anything inconsistent? If you choose to answer, then, as my co-blogger Bryan Caplan says, "Show your work."




COMMENTS (30 to date)
Bill writes:

Why no consideration of a "death benefit" associated with obesity?

Mike Hammock writes:

It is strange to treat the early death of smokers as a benefit (because it saves the Social Security system money) and then treat the early death of the obese as a social ill.

An economist should know that dying, early or late, is a cost, and that the pleasure of smoking or eating unhealthy (but tasty) food is a benefit.

Brandon Berg writes:

It's worth noting that Obamacare's community rating mandate allows insurers to charge smokers extra, but not obese people. In a free market, the costs of obesity would be internalized by charging obese people higher rates on their insurance, but the government has prohibited them from doing so. Ironically, Gruber is now proposing a government intervention to fix a market distortion created by an earlier government intervention that he helped to design.

Bill writes:

Arnold Kling has a related post:

http://www.arnoldkling.com/blog/obesophobia/

R Richard Schweitzer writes:

The purpose of "taxation" is to provide revenues for the functions of governments.

What "functions" of governments are being discussed here?

How have these matters become functions of governments?

Were these made governmental functions with the objective of providing a rationale for the power of taxation?

Can these objectives be "unraveled?"

Dan writes:

It's easy to find inconsistencies with selective quoting. Just quoting the sentence before and after will help clear any 'inconsistencies'.

David R. Henderson writes:

@Brandon Berg,
You might want to read the short Gruber article that I link to, and then reconsider your statement.

Justin Ross writes:

Would a body weight tax be a direct tax? I think so, and if so, then a body weight tax would need to be levied among the states with apportionment (i.e. tax revenue per capita would have to be equal across states).

The implication would be that a state where average body weight was below the national mean would have to charge a higher tax per lbs in order to have the same per capita revenue as an above national mean state. I don't think that is what they have in mind.

Hard to say if it really would be considered direct, but it looks more like a direct tax than income (which required a constitutional amendment to circumvent apportionment). Of course the U.S. Supreme Court just does whatever it wants, so perhaps paying attention to these kinds of details is not very important.

David R. Henderson writes:

@Dan,
Just quoting the sentence before and after will help clear any 'inconsistencies'.
Not true in this case. But go ahead and make your case.

Andrew_FL writes:

"The relatively low social costs can be attributed to the "death benefit": smokers live long enough to contribute to Social Security and Medicare but not long enough to collect benefits."

One begins to see where economists get their reputation for being heartless money counters. People die and all Gruber can think of is how it benefits or harms the State. To call this callous wouldn't do it justice.

Jon Gunnarsson writes:

I don't see any inconsistency, at least not on a somewhat charitable reading. In the first snippet Gruber is talking about externalities, in the second he is (presumably) talking about internalities. Hence there is nothing inconsistent in ignoring the personal utility of longevity in the first, but not the second case.

Ray Lopez writes:

I'm multi-tasking, surfing the net, listening to pop music, and playing chess, but I'll answer: it's the fact that smokers die early yet pay taxes, so it's better if they smoke? 120 IQ baby! I still got game. I shake it off shake it off.

David R. Henderson writes:

@Andrew_FL,
One begins to see where economists get their reputation for being heartless money counters. People die and all Gruber can think of is how it benefits or harms the State. To call this callous wouldn't do it justice.
I think you’re right that this kind of thing causes people to call us heartless, but I think they’re wrong to do so and that you’re being unfair to Jonathan. Remember that there were so many people claiming for years that smokers should pay higher taxes to cover their higher health costs. So if that argument is fair game, it’s fair game to point out that by dying early, they’re saving the government money.

Toby writes:

I can think of reasons to threat obesity different from smoking, but not based on these two quotes.

The reason I can think of treating these different is that obesity starts to cost money before a person dies as it leads to chronic conditions (e.g. diabetes 2), whereas smoking hardly does. The cost-benefit calculus is a bit different then.

Christopher Snowdon writes:

Gruber favours higher taxes on tobacco too, on the basis that smokers aren't as smart as him. http://velvetgloveironfist.blogspot.co.uk/2014/12/grubers-downfall.html

Dan writes:

"Not true in this case. But go ahead and make your case."

Sure. Gruber mentions that there are external and internal costs associated with smoking. He then mentions that there may be justification for intervention in the case of internal costs (right after the paragraph that is quoted here) citing his own research.

Wrt obesity, Gruber says that unlike cigarrettes (where internal costs are greater), both external and internal costs are high. This is int he sentence preceding the quote. Then in the quoted sentence he gives examples of external and internal costs, and says intervention is justified.

I see no contradiction at all. Even if you read it differently, it seems unfair to not quote the prior sentence, where he is clearly distinguishing internal and external costs and saying both costs unlike cigarettes are high.


john hare writes:

The inconsistency would seem to be the idea of taxing now something that might be a problem in fifty years. It would also seem to be in that cigarette smokers medical costs to the public are seen as less important than the medical cost of obesity to the public.

It is also in the idea of penalizing people on something that has very little to do with their health or its' cost to the public. A direct tax on body weight will penalize me for being a 200+ pound construction worker, while riding easy on the featherweight that might well be in far worse shape than I am. IOW penalties for natural variation instead of behavior considered undesirable.

Andrew_FL writes:

@David R. Henderson-I guess when you put it that way, it doesn't seem as bad. Still, I think the apparent moral absurdity of considering the death of a person a benefit to the State, exposes the original idea, that smokers should pay the state for their medical expenses, as itself morally absurd.

Somehow I don't think that's what Gruber had in mind, though.

Charlie writes:

The appearance of inconsistency comes from omitting the sentence before where Gruber argues that obesity leads to large public medical costs. Thus, obesity has social costs (to medicare/medicaid, etc) and private costs (early death).

"Estimates suggest that our obesity-related med- ical costs were $147 billion in 2008; these costs are borne by taxpayers for publicly-insured patients and reflected in higher premiums for private health insurance. Within 50 years, obesity is expected to shorten the average life span by at least two to five years.Thus, a large government role in addressing obesity could be justified on the grounds of reducing societal costs and mitigating self harm."

So, there are two grounds and two separate sentences explaining the grounds. I think that is the most sensible reading.

Brandon Berg writes:

@Prof. Henderson: Huh. I stand corrected, and apologize for the error. I wasn't sure of this point, so I checked before making my first comment, and found a couple of sources that said that smoking is the only exception to community rating rules. Maybe that's true of individual insurance but not employer-provided insurance?

Ray Lopez writes:

Here's an older Sep. 2014 Telegraph article that makes the same case as Prof. Henderson makes:

http://blogs.telegraph.co.uk/finance/jeremywarner/100028193/the-truth-about-smoking-it-saves-us-all-a-lot-of-money/

In any case, they bored for America on all the sins of tobacco, including its alleged cost to the public purse in treating the various health conditions associated with smoking. Actually that's not the case, piped up Professor Doll. "Tobacco smokers will on average have reduced life expectancy, and are quite likely to die before they become a burden on the state and the rest of society. What's more they tend to die quite quickly, so their medical costs are not off the scale, compared with keeping, say, an eighty year old alive. The economy will have had the benefit of their productive life, but few of the costs of their inactive ones".

MikeDC writes:

@David

Remember that there were so many people claiming for years that smokers should pay higher taxes to cover their higher health costs. So if that argument is fair game, it’s fair game to point out that by dying early, they’re saving the government money.

It's not unfair to call Gruber callous and somewhat evil at all.

Yes, Gruber is pointing out that society at large is effectively preying on the smoker (by distributing more wealth away from her and to the rest of us than a purely economic accounting would allow).

That's not evil. But pointing out that we're preying on these people and then advocating that we continue to do it certainly can't be called good, can it?

The truly liberal and economically fair position would be to argue that, since they're likely to die sooner, smokers should be taxed at a somewhat lower rate.

Chris writes:

I also noticed that that graph doesn't seem to show any evidence that taxation affected smoking rates. In fact, from looking at other sites, I don't think the taxation line is CPI adjusted, and if so, then that puts a nail in the coffin for that theory. If I mentally CPI adjust that line, I see a decline in taxes until 1982 or so, then probably close to a flat line until about 2000 and then a sharp increase. I couldn't find any data on the internet with a quick search to confirm this, except one site that claimed that inflation adjusted federal taxes have mostly gone down since 1960.

Maybe if one were to break it down by state and compare states with low taxes to states with high taxes one might see a causal effect? But the aggregated data shows no such thing.

As far as the sharp rise in taxes since 2000, the cynical side of me would not attribute that to some altruistic wish to reduce the smoking rate but more that politicians needed to find politically correct ways to raise revenues so they could spend more money.

Shayne Cook writes:

Apologies to all.

When writing a comment, I generally write the draft, hit "Preview", let it sit for a while, edit (several times) and then hit "Submit".

I hit "Submit" well in advance of completing my edits on this comment - and it's a mess as a result (e.g., J. Everett Coop instead of Koop, totality of comment in this form being way too long and barely coherent, etc., etc.).

If the EconLib Editor would be so kind as to remove my previous comment, I'd be grateful.

[okay, I've removed it. Be careful what you wish for!--Econlib Ed.]

micah writes:

We ought to levy a tax on people with heritable medical conditions who have children. Grandparent with a heart condition? $10k per child. That's at least as reasonable as a "direct tax on body weight."

If we can show a correlation between IQ and welfare payouts, we could extend this in that direction too. Evolution through taxation!

Mark V Anderson writes:

Shayne -- Please retract your request to remove your long comment. It was interesting and had discussion in there that added much to the discussion. It wasn't perfect, and it was too long, but none of our comments are perfect. We need your comments here.

Joe McDevitt writes:

Have a drink on me?

J Mann writes:

[Comment removed pending confirmation of email address. Email the webmaster@econlib.org to request restoring this comment and your comment privileges. We'd be happy to publish your comments. A valid email address is required to post comments on EconLog and EconTalk.--Econlib Ed.]

Shayne Cook writes:

To Mark V Anderson:

First, thanks for your comment and request. I'm flattered that you consider my comments of value. I'm afraid I won't be re-posting even a properly edited version of my original comment. But I do owe you an explanation of why I won't.

Dr. Henderson challenged us EconLog readers (of which group I am a faithful and grateful member) to identify the inconsistency in Jonathan Gruber's paper/statements. In fact, I was only able to identify the single consistency in Gruber's paper/statements - increase government revenue (taxes). That is his sole imperative with this piece. And I found almost no end to the inconsistencies in the "evidence" and "economic analysis" he uses and mis-uses to justify that imperative.

The problem with the length of my original comment was that it was not nearly long enough. Identifying and challenging even only the major inconsistencies in Gruber's paper - while "showing one's work" - is the stuff of graduate level research methods and/or economics term papers. Not blog comments. I know that because I did such critical term papers in my own grad courses.

(I'd dearly love to be a Naval Officer right now in Dr. Henderson's class in order to have him task me to do a thorough treatment of the inconsistencies in just this Gruber piece. Of course, I'd also dearly love being 40 years younger in order to even remotely qualify as a Naval Officer in Dr. Henderson's class.)

You are correct that none of our comments are perfect - perhaps least of all mine. Actually, I only aspire to, and look for, reliability in my work and that of others. Not perfection. And misspelling the name of a significant figure, such as C. Everett Koop, was an easily corrected error - that if uncorrected renders my "product" unreliable. I merely hadn't corrected that and some other easily edited errors before I posted. My bad.

(and thanks to EconLib Ed. for honoring my retraction request.)

James writes:

This is not necessarily an inconsistency when people endorse taxes to manage behavior, they seem always to skip an important step and ask if any existing policies are contributing to the problem. This is important because if the problematic behavior is being subsidized and then taxed, there is an avoidable dead weight loss going on. I believe Pindyck & Rubinfeld have an excellent textbook example demonstrating this by the use of income and substitution effects associated with a tax on gas that gets rebated back.

In the case of obseity, the number of policies contributing to the problem are too many to count. Just to name a few, there was the switch from the four food groups / balanced diet to the deliberately unbalanced food pyramid in state run health classes. There are the subsidies to the food industry that lean toward high calorie low nutrient foods. There are trade restrictions on sugar that lead food producers to switch to even more lipogenic ingredients such as high fructose corn syrup. There are unhealthy subsidized meals to low income school children regardless of health status. There is subsidized health care that allows unhealthy people to externalize the costs of unhealthy lifestyles. There are food stamp type subsidies that can be spent only on food. There are regulations which require insurers to cover the costs of predictable recurring medications and more regulations which prohbit writing policies which cut premiums by not covering health problems caused by obesity. Every single one of these policies costs money to maintain and reduces the costs of becoming overweight.

I wonder if Gruber has even compared cancelling all of these pro-obesity policies with a weight tax, both in terms of revenue and behavior. It's hardly a given that a bodyweight tax is the most efficient way to generate the desired outcome.

Of course if the goal is to generate revenue, there is the risk that people might decide to quit being overweight. Every good economist knows the inverse elasticity rule: apply greater taxes to those things which are less likely to be affected by the tax. Maybe Gruber should consider a tax on the drugs used to treat irreversible obesity related health problems such as diabetes. No current diabetic will ever stop being diabetic no matter how high the taxes on metformin or insulin.

Comments for this entry have been closed
Return to top