An ongoing debate in legislatures and academic papers is what should be the excise tax on vapor products. It turns out that answering this is really quite easy once the goal is specified. However, almost all the debate seems to be assertions of personal opinions, devoid of calculations.
The first lesson in this science lesson is, well, that this is a matter of science. Many people who would not make the mistake of asserting mere personal opinion about whether an exposure causes cancer or the age of the Earth do make that mistake when it comes to economics. This is partially explained by the same words being used for both assertions of political opinion (“it is morally fair to impose high taxes on the rich”) and hypotheses can be tested scientifically (“imposing high taxes on the rich will slow economic growth”). This is made worse because tobacco policy discussions seldom separate any scientific analysis from personal opinion.
Once we specify an objective for taxing vapor products, which is a matter of opinion, the optimal tax level can be calculated. That is not a matter of opinion.
Most assertions about what would be the best tax rate fail to specify what “best” means, suggesting that the author has no idea what he is even claiming. A common assertion is something along the lines of, “cigarettes are very harmful and have a high tax rate, so since vapor products are only a little harmful they should have a low tax rate.” But even if we assume vapor products are harmful on net (which is far from clear), this is nonsense. There is no objective specified, making the “should” meaningless. It turns out that setting the tax rate proportional to the health risk is not optimal for any credible objective.
(The basis for that conclusion and the math behind what follows can be found in my paper. The math is really quite easy. Any decent first-year economics student could replicate it on an exam).
When the goal is stated or implied, it is almost always about health outcomes. To flesh it out, the goal usually offered is “set a tax level for vapor products that manipulates consumer choices to minimize total health risk, adding up across consumers of all products.” It is quite easy to calculate that, for this goal, the optimal tax level for vapor products is zero.
A higher tax rate discourages people from vaping. Some would-be vapers who are discouraged from vaping will keep smoking or start smoking, dramatically increasing their health risk. Some will not use any product, slightly reducing their health risk (if we assume vaping causes some health risk). A few would switch to another low-risk product, but outside of Scandinavia – where smokeless tobacco is not a cultural niche product – this effect would be trivial.
Exactly how many consumers would change their choice if the tax rate increased – the question of the “price elasticities” to use the economics jargon – is an empirical question. Unfortunately, we do not have very good estimates to populate the calculation. Fortunately, we do not need them.
For any remotely plausible estimates of the elasticities, and of the health risk posed by vaping, having zero tax on vapor products results in lower population health risks than any positive tax rate. The added risk from encouraging more smoking is far higher than any risk reduction from causing some would-be vapers to quit entirely. Only if the population of smokers who might switch to vaping drops to a tiny fraction of what it is now, and the prevalence of vaping increases similarly, would this change.
Indeed, the optimal tax rate is really negative. That is, if the goal is to minimize population health risk, then e-cigarettes should be subsidized. Calculating the exact optimal subsidy level would require generating much better estimates than we have to populate the calculation. This is not worth the effort, however, since there is no chance of a subsidy being offered. We can ignore this theoretical option and just observe that zero is better than anything larger than zero.
There are, of course, other possible objectives for setting the tax rate. Sometimes the apparent goal is to extract as much revenue as possible from vapers, using issues of health as a mere rationalization. There appears to be no ethical justification for such regressive and inequitable taxation. Even legislators who are clearly trying to do exactly this will not admit to this goal, let alone attempt to defend it as ethical.
A goal that is widely admitted by tobacco controllers is reducing the number of people using any of these products, with no concern about health risks. This is oddly ignored in most discussions of tax rates, perhaps because it is so unsavory. It turns out that if we look at actual policies, they seem to come closest to reflecting this goal.
Discouraging the use of all tobacco products, without concern for their comparative risks, would call for higher taxes on vapor products than on cigarettes. Smokers’ costs include both the purchase price and the major health risk. For vapers and smokeless tobacco users, though, it is basically just the purchase price. Thus that price has to be much higher to give them as much reason to quit as smokers have. Bans, which are basically equivalent to a very high tax, exist in many jurisdictions for one or both of the low-risk product categories, while cigarettes are legal basically everywhere. Moreover, many jurisdictions without bans tax the low-risk products at a higher effective rate. This is what we would expect if the goal is not about health, but is just about the number of product users.
Finally, it is worth mentioning the goal that is usually considered the obvious proper objective in economic analysis: maximizing people’s welfare. People do not care only about health. They certainly do not care about tobacco control’s headcount of how many people are using products. They make decisions based on their personal tradeoffs among the benefits of consuming a product, the purchase price, health effects, and other considerations that are not particularly relevant in this case. There is no apparent way that a consumer’s welfare can be improved by raising the purchase price of vapor products. Thus if the goal is making people’s lives better, the optimal tax rate is again zero for vapor products, and cigarettes too. (That is, zero for the optimal excise tax added on for purposes of manipulating people’s behavior. Obviously, we still need broad-based taxes of some sort to pay for government services, and some of these taxes will be paid by consumers of any product).
For either of the objectives that is morally defensible and that policymakers claim to pursue, there should be no extra tax on vapor products. Not merely less than than the tax on cigarettes. Not proportional to the risk. Zero. Anyone who suggests otherwise is either pursuing some other goal or does not understand what they are talking about.
