FDA and e-cigarettes: the dangers of radical policy action

by Carl V Phillips

If U.S. FDA had succeeded in banning e-cigarettes in 2009 as they did a decade before that, it would have been a bad policy action for many reasons. It would have served no legitimate purpose, violated the mission and spirit (and, according to Judge Leon, the letter) of the FD&C act, violated the ethical norms of free society, dramatically lowered the welfare of many people while making pretty much no one better off, and hurt the public’s health. But at least it would have been a fairly conservative incremental policy. Relatively few people were using the products and had not been doing so for very long. A few small businesses depended on it. It turns out that enormous benefits would have been lost. But at the time, the action would have been a relatively modest departure from the status quo.

This contrasts sharply with the present FDA plan to ban e-cigarettes (for that is what it is) which has all those other negative characteristics and additionally suffers from being a radical departure from a status quo in which millions of Americans use e-cigarettes at least occasionally, and probably hundreds of thousands rank them in importance below only basic necessities, friends, and family. A multifaceted industry and an entire culture have formed around them. The changes in the world that would be brought about by the ban are staggering. 

Incremental changes are almost always better public policy. Sometimes radical policy action is warranted, of course — desperate times call for desperate measures. But radical policies are properly reserved for desperate times because they dramatically increase the chance of what are typically called “unintended consequences”. (This is really a misnomer, since all costs of an action are unintended consequences — the lightening of your wallet resulting from buying coffee is not something you intended to bring about, even though it is a cost you knew you must accept. “Unforeseen consequences” is closer to what is meant, though it is not quite right either because insightful observers might indeed foresee them. “Secondary effects” is more general, though a little vague.)

In a complex system — such as the public policy arena, the human body, or even a computer operating system — any change will trigger some secondary effects. If the policy change is small then there is a good chance that the secondary effects will also be small. But secondary effects can also “tip”, such that a small change in conditions results in a cascade of events that result in a major change of outcome. The more radical the policy, the greater the chance that such a major change will occur and also that it will not be possible to reverse it. The increase in the risk of a major tipping tends to increase more than proportionally to the magnitude of the policy change. Small changes to a reasonably stable system tend to result in very small secondary effects, and we may even have evidence suggesting this will be the case. Radical interventions move into uncharted territory that is difficult to fully imagine.

Examples: Raise cigarette taxes by one cent and black market sales will increase, but probably not enough to notice; raise them by a dollar and that might tip 20% of consumers into a black market which then becomes more efficient and accessible and so attracts another 20%. The Affordable Care Act, for all the rhetoric about its radicalism, was really a handful of incremental changes, the minimum possible intervention to achieve the goal; not surprisingly, the only unpredicted outcomes have also been incremental (mostly in the form of it working better than expected). However, one feature of the ACA — the online exchange — was a de novo creation of a reasonably complicated computer system, and thus a radical departure from the status quo; not terribly unexpectedly, it had substantial problems. The iPhone has a history of incremental improvements while Windows, of late, is a good demonstration of radical changes, and everyone knows how much fun that has been. Hit a planet with an asteroid and, boom, over the course of a few million years, furry creatures hunt epoch-spanning lizards to extinction instead of the other way around, and then put up little boxes in their yards to house dinosaurs. You get the idea.

An e-cigarette ban in 2009 would have had predictable consequences (at least until such a time that e-cigarettes became so popular in the UK or somewhere else that they spilled over). There would be very little new smoking cessation caused by e-cigarettes and limited technological advances. Some then-current vapers would have gone back to smoking and some would have quit tobacco products entirely. Some vapers would have continued to sneak in products from China, but given the numbers it would be a niche trickle, not a pipeline. Obviously I did not say these outcomes of a ban would have been good, but they would have been more-or-less exactly what was intended. Not today. Widespread e-cigarette use, sales, and popularity are the status quo now, and thus a ban is radical.

E-cigarettes turned out to have become a much more significant phenomenon than the evidence in 2009 suggested they would be. Now some might be inclined to say this means the emergence of e-cigarettes was a radical action, or at least transformation, and so was no different from radical policy action. (Some who get all their information from academic journals and corporate media are still saying it could become a radical transformation.) But this is fundamentally wrong. The emergence of e-cigarettes resulted from billions of small decisions and actions, each of which seemed like a good choice for the actor and was continued only if this proved to be right. (Tobacco controllers reading this might want to look up the concept of “free markets” at this point.)

Despite the breathless talk of transformative technologies and whatnot, a billion individual decisions in a free market represents a careful way to change the world. Yes, someone will always object to the choices or outcomes. We may have collective action problems and individual choices can even immediately create an outcome that is dispreferred even by the actors (though there is no reason to believe either of these is the case in the present context). But free-market evolution rarely really screws things up like radical public policy decisions can because each action is incremental and each change is, in effect, tested for goodness before continuing on.

A generic implication of all this is that it is outlandish that FDA has made no attempt to assess what the unintended/secondary consequences of the action might be. It is bad policy making to gloss over those even for a conservative policy change, but it is unconscionable for radical actions. Of course the nature of unforeseen consequences is that they are hard to predict with much precision. After all, who could have foreseen that trying to establish a puppet empire in Asia would not work out well. Oh, wait, that would be everyone other than those who just believed the government or New York Times pronouncements. Specifically predicting the rise of Islamic State in 2003 would have been quite a feat, of course, but the fact that we would ultimately lose the war in Iraq — to say nothing of Afghanistan — was an easy call.

Similarly, it is fairly easy to predict that the e-cigarette ban will result in the creation of massive alternative supply chains. FDA has no excuse for failing to acknowledge and analyze the implications of this. FDA and others seem remarkably — perhaps genuinely — unaware of just how obvious this is if you are at all familiar with the vaper culture. “Public health” people live in an echo chamber populated by the small minority of the population that are already doing what public health tells them to do. FDA talks only to “public health” people and businesses (with a preference for large businesses), and the latter are not likely to give this issue quite as much attention as it deserves.

The obvious truth is simple: This rule is intended to eliminate the widespread use of all e-cigarette products that FDA does not choose to approve for market. The rule will actually just create other sources of a wide range of unapproved products.

Now if you want to drill down to exactly what the new supply will look like, that is much more difficult (like predicting exactly what post-war Iraq would look like). We are back to a billion little choices, a few major decisions, and accidents of fate. Crux points include whether FDA tries to regulate mod hardware or zero-nicotine liquid (and whether courts allow it if they try), whether FDA or someone prohibits the sale of pure nicotine to consumers, and the legal status of personal importation. The more the efficient options are banned (with enforcement), the more complete and organized the full-on black market is likely to be.

Then there are countless individual decisions. Will hundreds of thousands of enthusiasts find that do-it-yourself liquid mixing is an enjoyable hobby that entertains one’s friends (like brewing beer) or a workaday mechanical task (like making coffee), or will most who try it eventually find it too much of a pain to deal with (like brewing beer)? Will more than a few actually try to grow tobacco and extract nicotine themselves, or are widespread asserted plans to do so just cheap talk? It would not actually be too hard to do some research that could support predictions. Someone should have done it before proposing a ban. But these are details that do not affect the overall prediction that a lot of consumption will continue and the average quality of the products (including in terms of health risk) will decrease.

FDA cannot get past thinking regulating e-cigarettes is like regulating pharmaceuticals. If FDA bans sale of a pharmaceutical, it will no longer be used, except perhaps to a very tiny population who swear by it or are desperate to try it and thus obtain it (which they can, of course). On the other hand, if the drug is a formulation of OxyContin or the like, lots of interested consumers find a way to get it or hack it — or they just switch to heroin. It is not as if FDA has not seen this play out before.