Minor Vapists

The Yale School of Medicine reports on some of the benefits of vaping and some of the costs of bans:

More than 40 states have banned the sale of electronic cigarettes to minors, but a new study out of the Yale School of Public Health indicates that these measures have an unintended and dangerous consequence: increasing adolescents’ use of conventional cigarettes.

Using data from the National Survey on Drug Use and Health, the research finds that state bans on e-cigarette sales to minors yield a 0.9 percentage point increase in rates of recent conventional cigarette use by 12 to 17 year olds, relative to states without these bans.

“Conventional cigarette use has been falling somewhat steadily among this age group since the start of the 21st century. This paper shows that bans on e-cigarette sales to minors appear to have slowed this decline by about 70 percent in the states that implemented them,” said Abigail Friedman, assistant professor of public health and the study’s author. “In other words, as a result of these bans, more teenagers are using conventional cigarettes than otherwise would have done so.”

The paper by Abigail Friedman is forthcoming in the Journal of Health Economics (working paper version).

Hat tip: Mitch Berkson.


But this is only an issue if the total number of adolescents who consume a tobacco product of any kind increases as a result of the ban. Is that what the findings indicate?

Why is it only an issue in that case?

Why would we care whether adolescents use cigarettes instead of vaping? I mean, if we accept the regulators' position that both activities are dangerous, then the issue is not which tobacco product adolescents use, but rather whether they use any tobacco products at all.

I'd rather we not use bans and regulations, but I don't think we'd ever convince regulators to change their tune based on the notion that there is a substitution effect between two forms of contraband.

Except that vaping is NOT a tobacco product!

Oh, my bad. It's inhaled nicotine and other chemicals designed to simulate the effect of consuming a tobacco product. Thanks for keeping me honest.


It is a distinction worth making though, both are unhealthy so both must be banned even though one is several orders of magnitude more unhealthy?

Jay, two points:

First, everything I've read about e-cigarettes suggests that there is no firm evidence one way or another about their relative risk. Perhaps you have read something different, and if so, that is fine. I would appreciate a link to sources, but I already acknowledge that the question could go either way. Should it be determined that e-cigs are safer, then yes it is a relevant point. (The effect of the ban, not the question of whether it is a "tobacco product.")

Second, we are talking about "regulator logic" here. If the regulators think e-cigs are dangerous, then showing substitution between two forms of contraband will not impact their decision to impose bans. Showing that e-cigs are safer might do it, though. That's all I'm saying.

No tar in vape.

This is not close.


When people say there's "no firm evidence" with regard to vaping, the implication is that it could be better or worse than cigarettes with equal likelihood. In reality, we are virtually certain that vaping is far healthier than cigarette smoking. It's true that vaping has only been around for a decade or so, so we don't have long-term subjects. However cigarette smoking is one of the least healthy activities known to man. Some argue that since vaping is a cigarette replacement that Bayesian prior should be the relative risk of smoking. However vaping shares *none* of the properties that are known to make smoking dangerous. Smoking is unhealthy because of the inhalation of smoldering organic compounds. Vaping does not involve that risk factor at all.

The characteristic the both do share is nicotine consumption. However the scientific evidence is absolutely firm on this. Nicotine is at most minority harmful (certainly making up less than 5% of the mortality contribution of cigarette smoking), and may in fact even be health-promoting. Again it's difficult because we don't have long-term large sample of nicotine consuming non-smokers, but the indicates that are Bayesian priors should be somewhere around the risk factors of caffeine consumption.

The other potential risk factor associated with vaping (which has absolutely zero to do with smoking) is the inhalation of vegetal and ethyl glycol. Both are additives that are extremely safe, extremely well-studied and extensively used in food products. It is possible that inhalation rather than ingestion may pose an unknown risk, but there's no a prior reason to believe this. There are many cases of workers who are exposed to large quantities of glycol aerosols in the course of their occupation, and the worse identified complication is minor respiratory irritation.

If vaping is less unhealthy than cigarettes (or vice versa) then we should care about the proportion using each

Yes, if.

I don't think it is "if" and its not even close.

Inhaling thousands of known toxins versus inhaling water vapor laced with nicotine is easy to compare. But, anti-smoking campaigns were never *really* about health.

Haha, what were they about? I'm guessing buying democratic votes, but let me know.

Regulators are not hereditary nobles.

Elections can change laws.

I think that the idea that "both activities are dangerous" does not mean that both activities are equally dangerous. I don't know that the science is settled re: health risks of vaping versus smoking, but it certainly is with respect to chewing tobacco versus smoking. Both activities are dangerous, both are "contraband" for minors, yet from a public health perspective it is far preferable that people use chewing tobacco over cigarettes...You may be right that regulators do not care, but they should.

The regulators' position is that both vaping and smoking are dangerous, but there is no credible evidence that the risks of vaping are anywhere near as severe as the risks of smoking. Regulators like to lead people to believe that they are equally dangerous, but they can't say so explicitly because it's clearly not true. Smokeless tobacco is similar. If you look carefully, you'll see that the more reputable anti-tobacco activists never explicitly say that vaping or smokeless tobacco are "as dangerous" as smoking. However, they often respond to claims that vaping or smokeless tobacco are safer than smoking by saying that those other products are "also dangerous" (and leaving the reader to assume that they are not actually much safer).

If smokers all gave up smoking and took up vaping, those specific regulators would be out of a job.

Its in their interest to consider vaping "the same" even when its just a cosmetic appearance.

Water vapor vs. burning vegetable matter, hmmm, which sounds safer?

Our city already advertises that having a nice fire in your fireplace is bad because of the smoke, but they aren't running ads against humidifiers.

I also suspect Nanny's are worried because smoking was their greatest tool in getting the public to accept banning things for health reasons.

Isn't it ethylene glycol or something?

"First, everything I’ve read about e-cigarettes suggests that there is no firm evidence one way or another about their relative risk"

Cigarettes are unquestionably more dangerous than e-cigarettes.

Relative risks of nicotine delivery systems

Thanks, Mark. I appreciate your offering a citation on this.

This certainly makes AT's point a lot better, but it still would have been stronger had he made it clear in his post that vaping was unequivocally less harmful than cigarettes. This may be common knowledge to some, but I don't consider myself totally ignorant on the topic, and what I'd read was mixed.

But really, my point wasn't that vaping is or isn't as harmful as cigarettes, it was that an argument that substitution between banned substances occurs is unlikely to persuade a regulator against a ban. Maybe i'm just cynical.

Nicotine is an optional additive to vaping products, and tobacco flavors are a subset of available flavor profiles. Your attachment to tobacco health language does more harm than good. Rather than indulge in myopia, the discussion should be about vaping specific health problems. For example, acetoin and diacetyl products are still employed in some juices. And there is limited data on the relationship between wattage and toxicity of common ingredients.

Innocent until proven guilty?

"Vaping" and "Smoking" are both processes, and the risk might be more strongly associated with the substances used. We have heard (surprising to me) that similar amounts of marijuana produce less cancer when smoked than tobacco (really I guess a chemically treated tobacco product).

Vaping might be less dangerous than those, depending on the substances and temperatures involved, but regulators have to guess about a moving target as substances and temperatures rapidly change in an innovative environment.

I personally would not be an beta tester for new vaping products, YMMV.

To document a little experimentation in the target market:

About 27% of high school students who have used both marijuana and e-cigarettes reported using the devices to vaporize cannabis. Those most likely to vaporize pot with e-cigarettes included males and younger students.

There is no great stagnation!

> We have heard (surprising to me) that similar amounts of marijuana produce less cancer when smoked than tobacco (really I guess a chemically treated tobacco product).

You're misinterpreting the research. Smoking the same amount of *any* dried plant material is nearly equally unhealthy. This is true regardless of whether it's tobacco, marijuana, pine needles or chamomile. The health effects come from the inhalation of smoldering organic compounds.

That being said, typical marijuana use results in *far* less raw weight smoked than tobacco use. A pack of cigarettes contains 20 grams of dried tobacco. A typical user may inhale 140 grams a week of smoke. In contrast even a very heavy marijuana user would find it difficult to exceed 7 grams a week of consumption. This is doubly true with modern high-THC content strains, which require less plant material to achieve the same level of intoxication.

"The heaviest marijuana users in the study had smoked more than 22,000 joints, while moderately heavy smokers had smoked between 11,000 and 22,000 joints.

While two-pack-a-day or more cigarette smokers were found to have a 20-fold increase in lung cancer risk, no elevation in risk was seen for even the very heaviest marijuana smokers."

No elevation in risk is a pretty big claim

22,000 joints in a lifetime is nowhere even near comparable, even to light smoking levels. Over a 30-year span, that averages to 14 joints a week. At 0.4 grams for a median sized joint, thats 5.7 grams of marijuana smoked a week. To put that in perspective that's equivalent to smoking less than one cigarette a day. Not one pack, one cigarette. There aren't even studies of people who consistently smoke tobacco at such a low level because they don't exist. And on top of that you have several more factors. People inhale a higher proportion of tobacco smoke than marijuana smoke. Light tobacco users almost certainly under-report consumption and heavy marijuana users most likely over-report (or high-joint smokers tend to roll smaller joints than the median).

Either the effect is likely to be so small and the sample size of ultra-heavy smokers so small that there's insufficient statistical power. Or like many carcinogens, smoke exposure has a super-linear impact on health. I.e. doubling consumption increases relative risk by a greater factor than two. Someone consuming 1 gram of smoke a day (very heavy marijuana smoker) vs 40 grams a day (two-pack a day cigarette smoker, would have only 2.5% chance the risk in a linear model and 0.0625% chance in a quadratic model.

Forgot to mention another point. People share joints much more frequently than cigarettes. Someone who colloquially smokes a few joints in a day has probably personally consumed only a fraction of the marijuana rolled.

If you do not have evidence against a behavior should you preemptively ban it?

Yawn. Boring. The small increase in adolescent tobacco usage is a bore. More significant is that adults quit smoking faster; it's not even close in terms of tradeoffs.

Here in the Philippines you see street urchins, age 12 or so, smoking all the time. Yawn.

Now to spice up this article and really underscore a problem, change the "v" to "r" in the title...now that's a real problem, since most minor committing crimes only go to prison until they're 21 (including for murder).

You mean that bureaucracies get onto moral crusades that have no connection with science?

Does not address of cost of how lame people look vaping; disregard.

Seriously though, (in addition to what's mentioned above in regards to measuring lifetime health costs of additional vaping against lifetime health benefits of decreased smoking uptake), before we leap to policy conclusions, the study needs to be combined with models of likelihood to begin smoking over lifetime if vaping, among other considerations. While banning minor-vaping may have undesirable secondary effects, not banning minor-vaping may have more undesirable secondary effects, and it's irresponsible to suggest that the measurement of a single secondary effect is enough to drive policy.

Our desperation to empirically validate policy leads to a stupid extrapolation of isolated conclusions. This is a nice data-point, but smart qualitative observations hold superior until much more work is done.

> [T]he study needs to be combined with models of likelihood to begin smoking over lifetime if vaping, among other considerations.

Extremely unlikely. In fact nicotine on its own is not actually a very addictive substance. Probably less so than caffeine. Tobacco is very addicting because of the combination of MAOIs which changes the neuro-biology of nicotine (making it more of an anti-depressant than pure stimulant). Its unlikely that someone who starts vaping before smoking would even develop a vaping addiction, let alone move onto tobacco.

Nicotine dependency is only one potential road to smoking. Other obvious roads could include socialization, habituation, de-stigmatization. There are also nearly infinite other secondary effects, such as marijuana use, other drug use, decreased disposable income, social stigma (anecdotal evidence: I would personally be much less likely to hire a vape user, due to conclusions I would draw on judgment and perceived associated qualities) - your imagination is the only limit to these hypotheses.

My point is that it's specious to conclude, as much as this comment board has, that we should roll back minor-vaping bans based on a single study like this: I'm much more swayed by the qualitative assumption that we don't want masses of children vaping, because it's prima facie a socially unproductive behavior, and this is not the preponderance of data necessary to overturn that conclusion.

Based on the available evidence it's pretty difficult to make a coherent case that minors should be banned from smokeless nicotine consumption, but not banned from caffeine consumption. If children can't vape because its "socially unproductive" than there's also no real compelling reason why they should be allowed to drink coffee or energy drinks.

In fact, I'll go one further. Nicotine has substantially more medical benefits than caffeine. Its highly effective at treating ADHD, schizophrenia, depression, anxiety, and obesity. Allowing minors to self-medicate with nicotine to treat these conditions can substitute away from commonly prescribed drugs with much more serious side effects. Millions of minors are prescribed high-doses of amphetamines for ADHD, which are long-term neuro-degenerative instead of neuro-protective like nicotine. Even outside this, nicotine has proven highly neuro-protective with regards to the permanent brain damage we now know is associated with sleep deprivation. Since the majority of adolescents are chronically sleep-deprived, encouraging smokeless nicotine usage among this group may prevent thousands of cases of Alzheimers and dementia in the future


"Highly" effective? Really? Do you have some non-gated links?

Could be you're right, could be you're wrong, but it has absolutely nothing to do with how we should regulate vaping.

Surely it has SOMETHING to do with it?

Fallacy of relative privation. Reaching consensus on the net effects of all things potentially put into minor body, then ranking them, and then only allowing regulation in descending order after complete consensus has been reached on purportedly high items is perhaps the silliest logical argument I've heard today, and that's what's being implied here.

That is quite a burdensome standard, as it requires the public to justify freedom from government intrusion, rather than forcing the government to justify intrusion.

Agreed in the case of adults, but think caution has to run in the other direction in the case of minors when risk-assessment abilities are undeveloped. The existing preponderance of evidence says primary effects are highly negative. All positive secondary effects may outweigh that negative, but a single data point indicating a single positive secondary effect is not enough to forgo caution in the regulation of interaction between children and products with massive negative externalities.

This doesn't make any sense. If I wasn't a smoker in the first place, why would I all of a sudden pick up smoking just because I don't have the option to vape?

Children who would otherwise pick up vaping as they got older pick up smoking instead

I think this is a reasonable theory. It's not all or even most of the kids, but even a small share makes a difference.

The study isn't saying kids who would otherwise be vaping are smoking cigarettes instead. It's saying kids who wouldn't be smoking anything at all are smoking cigarettes instead because they can't vape.

What? I don't see how the report doesn't say that at all. My reading of it is that vaping bans increase cigarette usage. That's consistent with vaping acting as a substitute to traditional cigarettes. Where did they identify the sub-group of "kids who wouldn't be smoking anything at all." How would you even identify a hypothetical group like that?

No, it doesn't say that. It suggests that vaping crowds out smoking by a small amount,

Seems like a good time to just forget about Jevons paradox

Why doesn't the government increase the market price of conventional cigarettes? As a result, the quantity demanded of conventional cigarettes would fall (a decrease in adolescents' use of conventional cigarettes), right?

They have.

In fact, this is widely thought to be one of the leading drivers of the decline in smoking rates.

An informal, unscientific study made by me indicates that the major reasons teen-age girls have part-time jobs is to pay for magic phones and cigarettes, The weeds are purchased for them by parents, older siblings and friends.

We wouldn't want teen-age boys to smoke but it's OK for them to play football.

Or ride motorcycles!

Nicotine is a cheap stimulant and anti-depressant. Big Pharma hates nicotine.

The vaporizer that exploded in Florida is (to me) a strangely under-reported story. http://www.usatoday.com/story/news/nation-now/2015/10/29/fla-man-hospitalized-e-cigarette-explodes-face/74791722/

Also, like somebody said above, don't these devices use ethylene glycol, i.e., antifreeze? How is that not harmful?

I think we actually come out ahead re: cigarettes due to earlier deaths but I don't know. And cigarette use goes down, alcohol and drug use and obesity go up. I don't have answers.

The most common bases are vegetable glycerin and propylene glycol.

Comments for this post are closed