Over at New Zealand's first ever economics blog 'The Visible Hand in Economics' Matt Nolan discusses externality taxes - in particular taxes on cigarettes.
An Otago university study that was sponsored by anti-smoking groups found that cigarette taxes should be increased. We know that an externality tax is a good thing, however 70% of the price of cigarettes is made of of taxes already. The question then is, do we need more cigarette taxes to set the social cost of smoking equal to the social benefit, are we at the social optimum, or have we already gone too far. Where the price is relative to the social optimum is an important question. If the price of cigarettes is already at or above the socially optimal level, further cigarette taxes will be inefficient.Now I have no idea where we are in terms of social cost and social benefit. Ultimately, if the money from cigarette taxes can cover all the additional health expenditure from smoking, then the tax is sufficient.
People know they are killing themselves with cigarettes, so if that is what they want to do we should let them. The problem is that they negatively influence other peoples health and put a drain on the health system by getting sicker than people who do not smoke. If the tax on cigarettes already covers all this, then I don’t want them to lift taxes anymore. The goal of the cigarette tax should be to cover the externalities of smoking, not trying to stop consumption completely.
As someone who thinks that John Stuart Mill's famous dictum ("[T]the only purpose for which power can be rightfully exercised over any member of a civilized community, against his will, is to prevent harm to others. His own good, either physical or moral, is not sufficient warrant") is a pretty good (if not perfect) rule of thumb for state intervention, I'm sympathetic to Matt's view. Instinctively, I am inclined to agree; yet when I think about it more I'm not so sure.
Firstly, even working within Mill's framework, and also limiting ourselves to justifications of taxation in terms of externalities, we can potentially make a case for taxation above and beyond that which covers costs to the health system associated with active and passive smoking. This is because the harm associated with smoking is not just the physical harm of cancer and emphysema, but also the financial, time-related, and emotional harms inflicted on family members, care-givers and friends. Of course, putting a price tag on such harms is incredibly difficult, but this doesn't mean that they aren't real. And while, quite possibly, difficulties in placing a price on these additional externalities provides a practical reason for not pricing them at all, we do need to be explicit about this and the limitations it places on any fairness claims we make.
Secondly, buried in Matt's point that, "[p]eople know they are killing themselves with cigarettes, so if that is what they want to do we should let them" are some tricky questions about choice and will. Questions which start to expose - I think - the limitations of a liberal framework based on Mill's dictum.
To see what I mean let's re-phrase Matt's statement to get the key word 'choice' in there: "People know the risks of smoking. If they choose to smoke knowing these risks, once we have accounted for the costs imposed on the rest of us, then they are entitled to their choice."The first problem with this is the a question of choice: to what extent to people actually - individually and of their own accord - choose to smoke? We know that people have the potential to act rationally in their own best interest. But we also know that there are a near infinite number of influences which may stop them from doing so. In the case of smoking we have: learned behaviour (in the children of smokers); peer pressure (I don't know where Matt went to school , but in the backwaters of the Lower Hutt where I was educated, smokers were cool); advertising; and - of course - the addictive nature of tobacco. With all these influences how accurately can we claim that anyone actually individually chooses to smoke. And, if individual choice doesn't exactly exist, is it really wrong for us collectively to try and establish some countervailing influences. After all, they may actually get us closer to the ideal of choice.
The second problem - and this, I have to confess, is something that I haven't full got my head round yet - is choice across time. Every moment of our lives we make choices; some we can undo, some - thanks to time's arrow - we can't, no matter how much we later regret them. Later in life we may have completely different preferences than we had when we were younger. We may choose to get that tattoo removed. We may choose to stop smoking. We may not, however, be able to reverse the damage done to our lungs. If only it were possible, we might choose to travel back in time and change choices we made. If a juvenile choice (to start smoking) and a mature choice (to live to see our grandchildren grow) are at odds with each other, hypothetically speaking, which set of choices should be given preference. Almost certainly, our mature choices will be made in the possession of more information (until we start forgetting it all), so perhaps they are better choices?
As an example consider the following:
Milton and I are friends at high school. Despite being suspicious of my communitarian leanings, Milton respects my opinions and is often influenced by them. One day, I discover that Milton has started smoking. I seriously consider talking him out of it, but decide to respect his individualist leanings.
Many years later Milton is dying of lung cancer. It is causing him immense anguish - particularly the knowledge that his wife and kids will not be provided for. At one point Milton exclaims "oh Terence, I wish you had talked me out of becoming a smoker at high school".
Mercifully, thanks to an unfortunate accident in an Econ 415 class, in the intervening years I was been given the ability to time travel and can undo Milton's juvenile choice, respecting his mature reasoning.
Would I be wrong to then travel back in time to physically prevent Milton's from smoking?
We can't time travel, of course, (not even those of us taught by Geoff Bertram) but we can make predictions about the future and, if we have every reason to believe that choices made in the present may be regretted by people's future selves, are we really wrong to want to influence them?
Choice, it seems, is an awfully tricky beast to pin down once you start thinking about it.
There is, however, a strong, and simple, counter-argument to my points above though: states stand on slippery slopes. And the more excuses we give them to intervene in our individual lives the more we increase the scope for abuse and illegitimate coercion.
For this reason, perhaps, even acknowledging, the limitations of Mill's dictum and conventional views of choice, we might still continue to use them as rules of thumb because the moment we discard the primacy of individual choice (either because we think we know better or through an over-zealous desire to eliminate negative externalities) we open a door through which abuses of power can be ushered in. Everyone regrets something they've done when they are drunk, therefore we should ban drinking. Spiky green hair upsets me, therefore we should tax it as a negative externality. On a larger scale this is the exact sort of reasoning that was used to justify communism (false consciousness) and fascism (the individual is subsumed to the needs of the fatherland) .
This is a very strong counter argument and it is the main reason why, despite its limitations, I think that Mill's dictum is a worthwhile rule of thumb in many instances. But, at the same time, I'm not totally convinced. After all, in line with my arguments above, we break Mill's dictum all the time (suicide is illegal, we have responsible drinking advertisements, we probably do tax cigarettes in a behaviour modifying manner not just to internalise externalities) and, despite what some conservative blogs might have you think, there are no gulags in New Zealand; nor gas chambers.
Just when and to what extent we should intervene collectively in the name of others' welfare isn't something that is easily decided on, but functioning democracies (and, heck, even moderately dysfunctional ones such as our own) actually do a reasonable job of it most of the time.
This surprises me somewhat, but hey, it happens. And, because of this, I'm willing to consider some welfare interventions at least.
4 comments:
Even by standard economic reasoning, there is a good reason to continue to tax past the point where the revenue gained equals the expenditure on smoking-related externalities. That reason is that the government needs revenue to pay for the rest of its general spending - all those things (at least in lovely happy land) which are public goods and hence can't be made to fully pay for themselves by those who benefit from their provision.
So if you need to tax SOMETHING to raise revenue, you have a choice between taxing cigarettes (beyond the point that is efficient, in the sense of internalising all externalities) or taxing other things.
Some of those other taxable activities: working, consuming, drinking alcohol, have externalities which should influence how much they are taxed. But even abstracting from that, there is a deadweight loss (aka excess burden) (*) from taxing anything, so an efficient level of tax on cigarettes should go past the point where revenue=externality costs right to the point where the deadweight loss per dollar revenue ("marginal excess burden") equals the deadweight loss per dollar revenue from taxing the 'other stuff'.
(*the positive difference between the welfare that would have been gained from the amount of the activity foregone and the amount of revenue raised by the tax)
That's all standard, albeit reasonably advanced (albeit2: intuitively obvious), economics. Unless I am mistaken.
Next, I'm sure that one can mount a plausible argument that there are positive externalities from working; hence high tax rates on labour vis-a-vis cigarettes are inefficient. Even if you argue that people, especially high-income high-stress workers, are OVER-working (and thus the externalities are of the negative "daddy's not coming home tonight...again" type), there surely are some income levels and worker types that are being taxed highly despite their jobs having positive externalities (role models, better home life etc). Anyway, daddy's not coming home tonight because his work culture is sick, not because tax rates are so low that he's rationally working more to achieve his preferred combination of consumption and leisure.
So even when the 'excise' portion of the tax on cigarettes has been met there's a pretty good argument that taxing them further beats (in a total welfare sense) taxing activities like (a) other goods and services consumption, (b) working.
Admittedly many remaining smokers are addicted so aren't likely to change their behaviour further, and are on low incomes, so suffer a large welfare loss when taxed highly on a low-demand-elasticity product. Those factors also seem like a good reason to lower their tax rates, or raise their benefits (distributed in-kind in terms of healthcare, education, etc).
Finally, just because I (and every other sole econ major) don't know enough to talk with confidence about concepts like liberty, regret, and the nature of growing government power and control, doesn't mean I don't think those concerns are important. I wish the "visible hand in econ" would at least acknowledge that "welfare" is only part of arguments about what "should" be done in policy.
Tim
Oh sorry, in the 2nd to last paragraph I meant "lower their INCOME tax rates" not "lower their tax rates"
hey there Tim,
taxing bads not goods to raise income . definitely an interesting point. I guess the best counter argument is the one you make about the (potentially) regressive nature of such a tax and the addictive nature of the drug.
also, richard layard, in his book Happiness, does make precisely that point about the negative externalities of working long hours and using tax as a means of changing this behaviour.
looking forwards to debating this over dinner.
Hi Terence and Tim,
Interesting points.
Terence you bring up the following points I believe:
1) There are other externalities
2) Peoples 'choice' is the result of outside pressure and so might not be 'optimal'
3) People are inconsistent with their choices across time
Now I completely agree that there are other externalities. However, now there may be other quantifiable externalities that we could tack on, but in the case of hardship on the extended family I have two problems with using a cigarette tax.
1) The smoker will care about their family, and so should internalise most of the cost of their smoking habit
2) It is subjective, and as a result we would need to figure out a way of quantifying it that people could agree on.
Now the second point is not much of an issue, but I feel that the first point is. Ultimately, I would agree that the tax should cover all externalities that are not internalised.
Your second point is institutional, there is strategic complementarity between my choice to smoke and others choice to smoke, and as a result if my friends smoke I might feel forced into doing this. This is an extremely interesting point, and I cover a similar issue here:
http://tvhe.wordpress.com/2007/11/01/halloween-an-inefficient-holiday/
Ultimately, it depends on whether the strategic complementarily exists because other people doing it increases the benefit of doing it, or increases the cost of not doing it. If you smoke with your friends because you feel it makes you cool, then that is still socially optimal (as long as the externality tax is in). If you smoke because you feel you might lose your friends then there is a problem. In this case you could definitely defend higher taxes, however I would like to see education with this.
Your third point that people are inconsistent with choices over time really just boils down to an information asymmetry. People do not ‘know’ what the optimal choice will be in the future, and so may make decisions now that turn out to be sub-optimal. The fact that someone regrets the decision in the future does not make it sub-optimal, as most of the benefit may have been in the first 10 years, while most of the cost was in the last 10 years. Once that benefit is sunk, you’ll feel a bit miffed that you are paying for something and not getting any future benefit.
If that asymmetry is always on the side where people under-estimate the future impact on them, then a tax may be useful but this is definitely a case where government funded education on smoking is the most useful tool.
Tim, I believe point is that
1) In order to provide public goods you need some funds, and the optimal tax mix may involve some extra smoking tax
Sure I can buy that. However, I’m not sure that this sort of idea is fair. A public good should be paid for equally by everyone, a cigarette tax is only paid for by smokers. In a sense we would getting people who smoke cigarettes to subsidise our public goods. Currently, we all pay a proportion of our income to fund public goods because that is what we believe is fair, it is definitely not what is efficient.
Your final point is that welfare is only part of the story. Well its not only part of the story, welfare is some arbitrary construction that is supposed to measure the wellbeing of society. Liberty, control etc all influence welfare in the same way that they influence an individuals wellbeing. I discuss more on how I view this here:
http://tvhe.wordpress.com/2007/10/18/utility-of-the-means/
If you feel differentially, then you should definitely make some comments here or on my blog about it, as I’m always interested in learning.
One final point. My post on cigarette taxes was not my whole divided opinion on them. I saw a story and then knocked out a little 5 minute discussion on what came to mind. I’m glad to see you guys are discussing it in more detail. I did microeconomics at university, so I don’t have any strong opinions.
I’d love it if you guys would come and comment on some stuff over there, I love talking about things.
Post a Comment