Tag Archives: incentives

Right level of excise tax on THR products? Zero.

posted by Carl V. Phillips

I have seen the claim come up a few times (this week it was here, coming from an e-cigarette industry group) that the tax on low-risk alternatives to cigarettes should be proportional to the tax on cigarettes, with the proportion determined by the comparative risk.  They seem to be saying that, for example, if the tax on cigarettes is $1 per pack and it were determined that e-cigarettes are 3% as harmful as cigarettes, then the tax on an equivalent quantity of e-cigarettes or nicotine solution would be 3 cents.

The reasoning behind this is completely wrong at several levels.  I cannot really call this an anti-THR lie because it is almost always coming from the pro-THR side (the ANTZ tend to prefer a tax as close as possible to infinity — i.e., a ban — for THR products), and is presumably motivated by trying to appease the ANTZ, an utterly pointless exercise.  Still, it is wrong and it is anti-THR compared to what is right.

The first observation is one you have probably already thought of:  How could we define what is an “equivalent quantity” (answer: it would necessarily be fairly arbitrary) and how can we determine what the comparative risk is with enough precision for this purpose (answer: we cannot, except for the well-studied smokeless tobacco products).

The second simple observation is that using this proportion assumes that the cigarette tax is somehow “right” in the first place, and thus is a proper baseline to use for setting other taxes.  But cigarette taxes vary radically (even across jurisdictions that have roughly the same levels of wealth and attitude toward cigarettes), so there clearly is no accepted correct level.  Moreover, the increasingly popular level of taxation is “as high as possible without tipping too many consumers into the black market”.  I know the following is a technical point and one that most of us do not agree with, but just to note it:  This means that those with the power really think the cigarette tax should be higher, but they are running up against a ceiling (a maximum caused by natural forces).  So the target (“optimal” in their minds) level of tax is higher than the actual level.  So, if it really were sensible to tax e-cigarettes based on the cigarette tax, it would need to use that higher “optimal” target level, not the constrained actual level.

That is moot, however, because the main point is that the proportion just makes no sense.  The excise tax on tobacco products can only be justified based on incentives.[*]  There is a lot of talk about the taxes paying for the burden smokers impose on the public coffers, but this is clearly disingenuous rationalization by those who want to create incentives:  First, we do not tax most things (e.g., cars, rock climbing equipment) based on the health danger they represent, so this is obviously not how we really make policy.  Second, most risky behaviors, including smoking, tend to balance out their own costs, with any increase in medical consumption being offset by a fairly similar decrease in total net consumption (from dying earlier rather than living a long retirement).

[*This is not quite true:  A high tax on a single good can be legitimately justified as a way of raising government revenue while taxing a relatively inelastic market in order to minimize distortions — economist talk for not creating much of an incentive to avoid the taxed activity because consumers are not very sensitive to prices.  That is undoubtedly much of the reason for high cigarettes taxes, but policy makers will not admit to that.]

So, if the motive is incentives, high taxes to discourage smoking, what is the right tax level to discourage the use of low-risk alternatives?  It should be obvious that it is zero.  Actually, it is really negative — i.e., we should be subsidizing these products to create more incentive to switch from smoking — but it is realistic to assume we are stuck with a floor of zero.  (The exact optimal level of the negative tax would depend on our objectives, consumer behavior, and factors such as the incentive effect of lower prices on non-users — I intend to write a paper about that sometime.)

The lower the price of low-risk alternatives is compared to cigarettes, the more incentive there is to stop smoking.  Since the cigarette tax is intended to discourage the highly-risky behavior, smoking, if we are going to set taxes of other goods based on this, they should further that goal of reducing the highly risk behavior.  Any positive level of tax on alternative products actually interferes with this goal.

Of course, a large portion of the ANTZ are anti-tobacco extremists, who do not actually care about improving health outcomes, but really want to torment users and eliminate all tobacco/nicotine products regardless of risk for various reasons.  From their perspective the right tax for low-risk alternatives is the same as it is for cigarettes: “as high as possible without tipping too many consumers into the black market” (that will be different numerically, but is the same conceptually).

So basically there are these possibilities from the above:  If the goal of a tax is to encourage healthier behavior, the tax on low-risk alternatives to cigarettes should be negative (which is to say, realistically, zero).  If the goal is to hurt users and eliminate tobacco product use without regard to risk, it should be as high as possible.  If the goal really were to make vapers or snusers “pay their own way”, it should probably be negative (zero), though that would require some tough calculations to nail down.  Finally (and just ignore this if it is too technical — it is not all that important) if the goal is to maximize government revenue by taxing an inelastic demand, the additional excise beyond regular sales tax should probably also be zero since overall demand for these products (due to the risk of switching to higher-risk alternatives) is probably not much less elastic than that for most categories of consumer goods.

For none of these does it make any sense to peg the tax level to that of smoking, via the comparative risk or any other function.