by Carl V Phillips
I continue to be appalled by what passes for ethical analysis in the realm of THR. This is clearly a symptom of the ethical failings of public health in general. Of course it is somewhat better to see someone actually trying to analyze ethics as compared to the normal “public health” approach of simply making a declarations about what should be done without any mention of what ethical goal they are basing that upon, let alone defending the legitimacy of that goal. The latter is a level of political discourse comparable to the average social media or comments section “debate”. But the attempts at analysis seem only to rise to the level of a freshman term paper.
The latest such analysis of ethics in this space is “Ethical considerations of e-cigarette use for tobacco harm reduction” by Caroline Franck, Kristian B. Filion, Jonathan Kimmelman, Roland Grad and Mark J. Eisenberg of McGill University. It appeared in Respiratory Research, which suggests a good rule of thumb: Do not get your advice on how to practice respiratory medicine from papers published in forums devoted to the study of ethics, and apply the same skepticism to an analysis of ethics in a respiratory medicine journal.
I have chosen to write a case study of this particular paper because (a) the authors seem to be genuinely trying to do what they claim, so the problems represent legitimate faults in the understanding of ethics in public health, not the intentional misdirection you see from tobacco controllers, and (b) these guys are not my readers’ heroes. But keep in mind that every one of the other “ethical analyses” of e-cigarettes I have seen in a health journal, whether written by detractors of e-cigarettes and/or THR or by vaper idols, is comparably flawed. At least these authors get credit for citing the author who I am pretty sure has written more serious analysis on the topic of THR ethics than anyone (ahem), which many fail to do, though they apparently did not read much of what I have written.
The paper also includes a simplistic summary of some of the science about e-cigarettes, as well as a simplistic analysis of the politics. Much of this is wrong, but I will just gloss over that in order to focus on the ethical analysis. Probably the best lesson from those bits is that it is always a mistake to try to summarize a complicated body of science en passant in a paper that is not about summarizing that science.
In the abstract and again in the introduction, the authors assert:
Several ethical issues have been identified pertaining to their use both as recreational products and harm reduction devices, including their potential appeal to non-smokers, their potential to act as a gateway to cigarette smoking, and their potential to renormalize a public smoking culture.
Notice any problem here? The three items on that list are not actually ethical issues. This is a list of material features of the world. A material fact (or uncertainty about it) is not an ethical issue; you can declare it to be fortunate or unfortunate, but it is neither ethical nor unethical. These authors, like most public health people writing about practical ethics, skipped over two crucial steps: They need to identify what particular action (presumably, policy) they are subjecting to ethical analysis, and they need to identify their basis for assessing its ethical goodness. Only with those two in place is it possible to say something like “the ethics of this action according to ethical rule X hinges on e-cigarettes’ potential appeal to non-smokers.”
You might think that they covered those two steps and were just sloppy in their summary here. Or perhaps they were approaching the analysis a different legitimate way, noting a particular worldly phenomenon and attempting to catalog the many policies that could affect it and analyzing their ethics, ideally based on various ethical rules. But of course they are not attempting the second in this little paper. While they try to explain what ethical bases they are using, they fail to identify what actions they are analyzing. It is possible to analyze the ethics of something only if it is an act carried out by someone with moral agency, which basically means an adult human or collection thereof.
The (not at all unusual) conflation of states of the world with ethical questions is a symptom of the ethical blindness in the health sciences. In spite of setting for themselves the goal of analyzing ethics, the authors cannot escape assuming that any fact about the world that piques them or their colleagues across the hall is, ipso facto, an ethical issue. They gloss over the question of what action they are analyzing because they assume that they can implement any change in the world by fiat, and are apparently oblivious to the additional major ethical questions that exercising such power would introduce.
They go on in the abstract to say:
Specifically, our framework draws upon tensions between utilitarianism and liberalism in public health ethics.
As with most every paper by public health people that makes this pronouncement (there have been several about e-cigarettes), the authors apparently do not understand what the words in this sentence mean. Utilitarianism is a precisely defined (in theory) concept which they do not actually invoke.[*] Liberalism (libertarianism) is a broad family of possible ethical rules, not a well-defined rule in itself.
[*Utilitarianism is a welfarist ethical rule. Welfarist rules are those consequentialist rules (i.e., the ethics of an action or policy is judged based upon its consequences, as opposed to upon the action itself) in which the consequences are measured in terms of the welfare (happiness, well-being) of affected people. Utilitarianism imposes a particular rule for aggregating the overall social measure across people, by adding up each individual’s net change in welfare across everyone. My parenthetical “in theory” comment was a recognition that the difficulties of somehow measuring and adding up people’s welfare make this rule well-defined as a concept, but not in practice.]
A bigger problem is that public health people who casually assert that there is such a dichotomy seem oblivious to the fact that most liberal ethics and utilitarianism usually lead to almost exactly the same ethical conclusions regarding policies about individual consumption choices and commerce. This is a corollary to economics’ invisible hand theorems, which show that under certain assumptions, and taking the distribution of wealth as given, welfare is maximized (at least in the short-run) by allowing freedom of consumption choices and commerce. Put more simply, people generally choose to do what makes them happiest, given their options, and the free market facilitates maximizing that happiness. Of course there are many tensions between these two bases for ethical judgment. Some of these are irrelevant to the present discussion (e.g., redistribution of wealth will usually result in a gain by utilitarian measures, but violates some versions of libertarian rules), but others might be (e.g., if someone’s preferred choice hurts others’ welfare, or if people are acting based on incomplete information). But these public health attempts to analyze ethics do not tend to understand such nuances, in part because they seldom understand what “utilitarian” means, or even what welfare is.
Consider one of only two places in they actually claim to compare the implications of utilitarian and libertarian ethics. Their entire analysis of the ethics appears in this one sentence (what came before it is simple background information):
From both utilitarian and liberal perspectives, misinformation through the provision of inaccurate comparative risk is fundamentally unethical for its failure to allow consumers to make informed choices, and for effectively conveying the message that smokers may as well continue to smoke .
Reference 59 is to me. Right author, wrong paper. I have written quite a lot about both the libertarian and welfarist arguments against disinformation in this context, but not in that paper. I thank the authors for the citation and the paraphrase of the title of that paper (though why they chose an inferior paraphrase over the iconic original, I do not know). But that was not the right one of my writings to cite, because it does not analyze either utilitarian or liberal ethics.
In that paper, I give welfarism only a passing mention, and do not address libertarian ethical rules at all. That paper’s ethical analysis is brief (for it is an empirical study, not an analysis of ethics) and focuses on much simpler rules of conduct. My claim there, which is neither libertarian nor welfarist, is that the lying is per se unethical, and that it furthermore violates the primary rule of health ethics, that people must be allowed to make informed choices about their health. The latter bears some family resemblance to a libertarian argument because it is about the ethics of actions (rather than their consequences) and invokes rights, but it is actually a much stronger condition than a mere negative right of liberty. The informed autonomy ethic affirmatively obligates health professionals to actively inform people in order to facilitate their informed free choice. Thus it constitutes a positive right (to be given the information) not merely a negative right (to be free of someone stopping you from making your choice).
It is certainly true that lying about THR harms welfare, and thus is unethical from a welfarist perspective. I have argued that extensively, but not in the cited paper. However, not all libertarian rules of ethics condemn lying and some libertarian ethics do not allow for there to be any positive rights, which necessarily impose affirmative obligations on others. Thus my position there was flatly contrary to some libertarian views. For example, there are libertarian rules (which are endorsed by some people) that allow caveat emptor. That is, you have the liberty to make your own choice, but someone trying to sell you on a choice has the liberty to freely speak and act to try to manipulate that choice, so long as they do not force you, and it is up to you to deal with that. Most of us agree that this would be a terrible rule to live by, but is is one libertarian ethical rule, and thus it is wrong to make a blanket assertion about what libertarian rules imply on this issue.
At this point, it should be apparent that we need to figure out what specific rule these authors had in mind when they invoked the broad category of “liberal perspectives”. Similarly, we have to figure out what they mean by “utilitarianism”, since it is apparent that they keep using that word, but it does not mean what they think it means. These problems might represent mere gaffes in a paper that made a passing reference to ethical analysis; they are fatal flaws in a paper that focuses on ethical analysis. In fairness to these authors, such fatal flaws can be found in almost every such paper in public health.
So, scrolling back,
Specifically, our framework draws upon tensions between utilitarianism and liberalism in public health ethics, the former aiming to produce the largest public health gains through the greatest reduction in the burden of disease, and the latter holding paramount individuals’ right to self-determination in health.
Ok, now we know how they are misusing “utilitarianism”. Credit to them for actually defining their terms, though a major penalty for using the word incorrectly. The goal they describe is what I have dubbed the “public health pseudo-ethic”, and it can be described as a special-interest goal, but not as an ethic. I refer to it as a pseudo-ethic because it gets discussed as if it were an ethical rule, but it is completely indefensible as such. No one ever defends that goal on the basis of any established ethical system, and indeed they could not because there is no ethical theory that supports it. It is not welfarist, because the measure is not welfare, but merely one arbitrarily-chosen component of welfare, reduction of disease. No one ever tries to defend this goal by claiming that if it were adopted as an ethic it would make the world a better place. It clearly would not do so. It certainly does not reflect the empirical reality of how anyone chooses to live their life.
As for what they characterized of a libertarian ethics, it is narrowed somewhat from just suggesting that “liberalism” is a well-defined rule. But it is still presented sloppily. It is only later still (for some reason, in the section entitled “e-cigarette safety”) that they imply that they are actually referring to informed autonomy:
…promoting autonomy, or the right to make individual decisions with regards to one’s life choices, requires the provision of information concerning the risks and benefits associated with a given behaviour and with voluntary choice .
So after a little scavenger hunt, we can conclude that what they mischaracterize as liberal ethics is actually the informed autonomy principle. Sadly, reference 32 is not to one of the classic statements about this principle, as you might hope, but from a tobacco controller article from 2005. They continue:
This rights-based position is compelling given that the majority of e-cigarette users are current smokers attempting to quit or reduce their number of cigarettes smoked .
Having finally let us know what they mean by the liberal ethic, they immediately fail the test of knowing what that rule means. As I already noted, it is not entirely a liberty-based position, in that it imposes affirmative obligations on some actors. But still it is a deontological rule, meaning that the ethical rule applies to actions themselves, not their consequences. It is about rights and obligations that exist apart from the details of the situation. Thus the entire rest of the sentence is nonsense. What they have said here is analogous to “‘thou shall not kill!’ is compelling in this particular case because the person you were going to kill is a nice guy.” Either the principle is an overarching ethical rule or not. If you are assessing the results it would bring about in a particular circumstance before deciding, you are engaging in consequentialist ethical reasoning of some sort, not following a deontological rule.
(And, incidentally, reference 33 is to a convenience sample survey and therefore cannot support the claim attributed to it. I will now return to ignoring them getting the science wrong.)
While autonomy may be compromised through the influence of nicotine addiction, the consequences may be less pronounced where this choice consists of selecting between alternative sources of nicotine (of potential equal or similar satisfaction), rather than choosing between indulgence and abstinence.
And with that, they resolve the question of whether they even understand what autonomy is. A common trope in some of the anti-tobacco junk science is to declare that “addiction”, whatever that means, causes (or perhaps is) a loss of autonomy. But those effects are not a matter of autonomy, but a change in the preferences that motivate how that autonomy is likely to be exercised. This misuse of the term implies that autonomy only exists if someone is utterly indifferent about what choices to make (which, ironically, would render autonomy materially worthless). Once again, the misuse of this term elsewhere in the public health literature is bad enough. But when someone is trying to write about ethics, it is yet another fatal flaw to get it wrong. Whatever the people making those “loss of autonomy” claims think the word means, in the context of ethics it refers to prohibitions against one actor constraining the actions of another (and, perhaps, also to manipulations that are not literally force but have the same effect).
As for the rest of that sentence, whaaaat? Again, consequences. I suppose we can charitably interpret this as the authors proposing to assess the consequences of adhering to the informed autonomy ethic. That, unlike the previous sentence where they proposed to judge the validity of the deontological rule based on its consequences in the particular case, is perfectly reasonable. However, they do not actually assess the implications of that rule, as applied to a particular act, based on the consequential rule (the pseudo-ethic) that they are purporting to analyze. Rather, this is some weird sludge of judging the consequences of some phenomenon that stands apart from ethics (because there is no act in sight) based on a non-consequentialist ethic. Huh?
This continues for a bit, but I cannot stand it anymore.
Anyway, patching all this together, we can surmise that the supposed goal of this paper is not an analysis of two competing ethics, but about the tension between the ethically indefensible goal implicit in most public health discourse and the actual primary health ethic. Ok, that is an interesting exercise. Unfortunately, they fail in their analysis of that question also.
Reconsider the above-quoted passage about lying. It is now clear that the reference to my paper was arguably defensible because they are not actually talking about a libertarian ethic, but about the informed autonomy rule I briefly addressed in that paper. (Too bad they did not pick up on how to properly describe it.)
But they still have a serious problem. Assessing whether particular lying might further the public health pseudo-ethic is a complicated empirical question, which I did not address in that paper. It is easy to make a blanket argument that misinforming people inevitably lowers their welfare because they can then not correctly assess their tradeoffs. (This is a bit of an oversimplification in various ways that I will not go into here, but it is a solid starting point: people cannot maximize their welfare if they have bad information.) However, while unmitigated lying about THR pretty clearly harms not just welfare but also the goal of the public health pseudo-ethic, it is entirely possible it might be that some particular act or policy of disinformation could further the (ethically indefensible) public health goal.
For example, it could be that misleading people into thinking that e-cigarettes are 5% as harmful as smoking, as in the current campaign from certain quarters of public health, could dissuade some nonsmokers from taking up e-cigarette use. (The implication that this furthers the pseudo-ethic goal assumes that e-cigarettes cause a net positive increase in disease risk, which is not a foregone conclusion, but for present purposes assume it is true.) I have suggested that it this claim was concocted to facilitate supporting e-cigarettes as a “cure” for smoking (5% is a lot lower than 100%) while still maintaining justification for severe government controls on e-cigarettes (5% of the risk from smoking would make it easily the second-most harmful product many people choose to consume in the West). But a closely related explanation it is an attempt to mislead people in order to maximize the goal of the public health pseudo-ethic. That is, the goal of the claim is to trick some nonsmokers who might experience welfare gains from taking up vaping (based on realistic estimate of the health risk) into believing that would be worse off doing so because of the substantial risk, and thus increase their adherence to what public health wants them to do. Notice that this is a fail in terms of a genuine utilitarian ethic, since it would lower their welfare, but is a gain according to the pseudo-ethic.
That would be interesting to analyze. But the paper fails to do that or anything similar. Indeed, they do not analyze any particular policy or action, and so all of their conclusions are based on nothing — again, they are missing one of the fundamental steps in doing an ethical analysis of a policy, identifying the policy.
In the only(!) other mention of the comparison they claim to be exploring, they write:
A useful paradigm that reconciles liberalism and utilitarianism in illustrating the impact of displacing a high-risk activity with a low-risk one is the risk/use equilibrium. For instance, if e-cigarettes reduced a smoker’s risk by 99 %, for every smoker who switched to e-cigarettes, 100 non-smokers would need to initiate e-cigarettes…. Consequently, it is unlikely that e-cigarettes would result in net public health harms despite the inevitable uptake of the product in a non-smoking fraction of the population. …. In practice, sound public health policy can sustain autonomous choices with deleterious consequences to the extent that these do not outweigh net public health benefits.
So they present the obvious simple calculation that shows that the mere existence of e-cigarettes results in a net gain in terms of the public health goal, (for the short-run, that is). But what “sound public health policy” are they talking about? We all know that e-cigarettes and snus reduce population risks compared to a world with just cigarettes. But that is just background information, not ethical analysis. Unless they are proposing a policy of eliminating those products from the world, while for some reason keeping cigarettes, this calculation has no relevance to ethics (which is to say, it would not even have relevance to ethics even if the public health pseudo-ethic were a defensible ethic). Assessing the consequences of a material fact about the world according to some goals is an analysis similar to consequentialist ethical analysis, but it is not a matter of ethics. Even more of a problem, it does not make any sense whatsoever to talk about deontological ethical rules in the context of phenomena that involve no acts that could be subject to deontological rules.
If they actually analyzed a particular realistic action or policy, the arithmetic they present could be used to show that the predicted result (which itself would have to be assessed based on what an analysis showed to be the real, not fantasized, impacts of that policy) would further the public health pseudo-ethic. If the action also fulfilled the requirement for informed autonomy, then they would have shown what they claimed (for that one action). They did not do this.
Their comparison of the two rules continues with…, er, nope, that was the entire content that related to their stated mission. The paper mostly consists of a meandering survey of things that someone once asserted about the science, politics, or history related to the topic, or merely furled their brows about.
They have only a couple of other ethical assertions, which are even weaker than their claimed analysis:
Beyond any immediate emission concerns however, the ethical arguments surrounding second-hand vapor exposure are those that apply to tobacco cigarettes: exposure to e-cigarettes should not be imposed upon those who do not choose to use them, providing a strong argument for use restrictions in public places.
How’s that again? Is this bold assertion based on the public health pseudo-ethic? That seems like a stretch since the environmental emissions are clearly approximately harmless but a policy restricting vaping could encourage smoking, thereby scoring less well according to the pseudo-ethic. Is it based on the ethics of informed autonomy? It is difficult to see how it could be, since it is basically the antithesis of autonomy. So apparently they are pulling another unidentified ethical rule out of thin air, asserting it is valid, and judging this question based on it rather than the rules they claim to be analyzing, without ever telling the reader what the rule is. In other words, they have descended back into the usual relationship between public health people and ethics: “whatever we personally feel like is the right conclusion is in a particular case”.
But perhaps in order to make the above assertion look like legitimate analysis, they also write:
Regardless of their industry ownership, e-cigarette companies would nevertheless have a vested interest in maximizing the number of long-term product users. The ethical onus then falls on governments to restrict the influence of industry through appropriate regulations targeting product manufacturing, availability, and use, devised in light of public health interests.
What ethical concern? What ethic? What is an ethical onus? What is the justification for curtailing people’s free choice? This is supposed to be an analytic paper, not just another worthless statement of personal opinion. So where is the analysis?
I have no idea, and I will leave it there. Given that this is what passes for ethical analysis in health sciences, it is little wonder how poor the ethics of public health people are. It is not all that difficult to find a legitimate ethical analysis and use it as a guide for writing a new one. It is even easier to find a colleague with a decent liberal arts education who can explain what “utilitarian” and “liberal” actually mean. But they did not even try, presumably due to the Dunning-Kruger-fueled arrogance that permeates public health. And, as I said, this paper is above average for what passes for ethical analysis in public health.