by Carl V Phillips
A new paper in the normally more-respectable BMC Public Health, by never-respectable ANTZ at the University of California (San Francisco) reports research that mostly showed that, if people were given disinformation claiming (nonexistent) health effects from smokeless tobacco and e-cigarettes, accompanied by gory pictures, then they will be tricked into to thinking the risk was higher. Surprise!
Well, of course, it is no surprise that people can be tricked and no surprise that UCSF “researchers” would conduct such unethical research. It is rather more of a surprise that the non-ANTZ BMC Public Health would publish it and that an ethics committee would allow it to be done. Ok, maybe not the latter — the ethics committees are pretty much in the pocket of public health. That committee at UCSF probably would never allow, say, Farsalinos’s survey of e-cigarette users, and would trump up some claim that it was a threat to the study subjects, whereas they allowed serial liars Lucy Popova and Pamela M Ling a free hand to tell people they might as well smoke.
Anyway, Clive Bates was first off the block in responding to this travesty, and he covered the breadth of it well, so I am not going to reinvent the wheel here. Go read what he wrote first. Then come back to this, wherein I go deeper into a few specific points.
The title of the article refers to warning labels and the authors use that term to refer to the gory graphics that they showed people. But those are not warning labels. “Warning” implies the transmission of information about risks that someone might decide to act upon. A graphic warning would not be a scary photograph but a picture that actually conveyed warning information. The graphic might help it stand out from the text, such as those that appear on ladders with a stick-figure standing on the top step with a red slash through it. Or it might be for communicating with the pre-literate, like the “Mr. Yuck” image on a poisonous household products, or across language barriers.
Most important, a warning is something that communicates factual information that allows people to more accurately assess risks. The warning on the ladder does not convey the message “if you step on the top step, you will die from Ebola tomorrow.” If it did (imagine for the moment people would believe it), then it would definitely keep people off of the top step, but not because it warned them of the risk, but because it lied to them about a non-existent risk. A statement on a low-risk tobacco product that says, in effect, “this is as harmful as smoking” is not a warning, then, but a manipulative lie. Popova and Ling were testing the effects of lies, not warnings.
Moreover the gory graphics that “public health” people are so fond of, and that were used in the present study, are worse still: they do not convey information at all, true or false. That is why they always come with a caption — there is no information content in the picture. Instead the picture is emotional violence designed to manipulate people’s non-rational feelings. We[*] wrote about that at some length in a comment to the FDA on the topic and while the FDA ignored that, a lot of my language and reasoning found its way into the court ruling that blocked FDA for imposing graphic “warnings”, so apparently Judge Leon’s clerks read it, at least.
[*We = my THRo research shop which was later merged into CASAA but was not part of it at the time.]
It is not exactly news that you manipulate people’s behavior with emotional violence, and that emotional violence and other emotional manipulation is more effectively delivered via photographs rather than text or information-containing abstract graphics. Public health people are just borrowing the tools that have been perfected by warmongers and tyrants, people with the same willingness to sacrifice human welfare for their cause.
The authors’ main conclusion is:
Regulatory agencies should not allow “lower risk” warning labels, which have similar effects to the “FDA Approved” label, which is prohibited, and should consider implementing graphic warning labels for smokeless tobacco products and e-cigarettes.
Set aside the patently idiotic reasoning (“truthful information X had a similar effect to false claim Y by our simplistic measures, and therefore it must be bad to tell the truth”) and focus on should. That work implies that based on some stated goal, they believe that the evidence shows that the particular actions would further the goal to a degree that justifies their cost. As I pointed out in my comment addendum to my essay on the history of the public health pseudo-ethic (which some might have thought was already long enough without it :-), it is basically never ethical to make a policy recommendation in a research report. A report on a single piece of research does not even show that the scientific evidence as a whole supports the simple immediate conclusions that might be drawn from the one study in isolation. It certainly never contains an analysis of the complete ramifications — costs, benefits, real-world effectiveness — of a proposed policy. Thus recommendations like this are always unethical.
But this one is far, far worse than average.
I read through the introduction and rambling tangential final bits of the paper with this conclusion in mind. (And, yes, it was painful. Normally, when I am interested in research results, I never read those sections — and I recommend that strategy to everyone. They basically never contain useful information.) Their implicit justification for all of this are the authors’ declarations in the first paragraph that “tobacco use” is very unhealthy, which of course was all about the effects of smoking. They offer no justification at all in this paper about why someone would ever want to discourage smoke-free tobacco product use, whatever the method. None.
They report on countries that have imposed emotional violence graphics on packaging or are considering it, perhaps thinking that this is an argument it must be a good idea, and undoubtedly not having the intellect to figure out this would also serve as an argument that beheading people for witchcraft or drug possession is a good idea. They then recount efforts to try to get the accurate messages across, such as RJR’s initiative to change the inaccurate warnings on smokeless tobacco. They never once say that such efforts are a bad thing, let alone justify such a claim. They never deny that RJR was seeking to have the labels changed to being more accurate. They just assume that everyone else shares their religious extremist views about beheading …er… making people believe falsehoods about tobacco products, and therefore any attempt to change that must be evil.
The study, as they interpreted it, showed that changing the label to RJR’s proposed messaging would significantly lower people’s perception of the risk from smokeless tobacco. That is, it would move them closer to believing the truth, though they would probably still grossly overestimate the risk (this is not clear because the methodology is such a joke it is difficult to interpret the results). The authors obviously do not like people knowing the truth, but they never even say this is a bad thing, they just assume that their target audience already thinks that.
That is the extent of their (non)justification of their conclusions. There is no statement of goals. There is no actual statement (only innuendo) that it is better if people believe false negative information. There is no discussion of real-world effects of labels or whether their artificial and vague study results seem like they translate into anything real. There is not even a review of whether other research tends to support their specific conclusions (they lie with the usual “this is the first research to do exactly what we did” line to imply there is no other relevant research in the world; there is tons of relevant research — they were just too lazy to bother with it).
In other words, with no stated goal or justification, and no analysis of what acting on these recommendations might really do, they nevertheless make the recommendations.
I discussed at length the implicit and never-justified pseudo-ethic in public health — pursuing the longevity and purity of the bodies they wish to protect from the owners of those bodies, whatever the cost. But these authors make even that look good by comparison. They are making pronouncements about what should be done without even a hint of analysis that suggests it would fulfill even that goal.
In a field crowded with contenders, this is a solid candidate for the most unethical public health research I have ever seen. The mission of the research was purely political as were the conclusions, making it unethical for those reasons alone. There was no possible apparent benefit to the “research” — the way the study was designed, there was no possible way it could tell us anything useful, even setting aside the politics and misuse of the results. And yet it harmed the subjects who participated in it. They were forced to look at horrifying graphics that are not easily forgotten. (If scenes from the movie version of A Clockwork Orange are coming to mind, I am sure you are not alone.) Subjects were given messages that were flatly false and would tend to cause harm if believed, and apparently not briefed at the end that these were fictitious. The reader of the paper is also never told that such claims in the experiment as “e-cigarettes cause oral cancer” were made up from whole cloth by the authors.
The research was “approved by the Committee on Human Research (the IRB) at
the University of California San Francisco.” So either this ethics committee made the blatantly unethical decision to allow all of the above, or the researchers lied to the ethics committee about the research having value and the messages not being deceptive and emotionally violent (and if the latter, then the committee was still unethical due to its gross negligence in believing this). Moreover, this was not some random flight of insanity that was self-funded by the fanatic authors (as if these people would ever do something without getting paid for it — ha!). It was supported by a grant from the National Cancer Institute, making them complicit in the unethical behavior.
On top of that, BMC Public Health published it, a glaring failure of the journal review process. The research was unethical. The authors’ stated and implicit premises were false. The conclusions did not follow from the research, even if the false premises were accepted. The research was technically flawed (for reasons that pale in comparison to the ethical problems, but are nonetheless significant). And yet the “peer review” system that public health loves so much seemed to have no problem with it.
It follows from this that the NCI should be defunded, UCSF should be placed in the hands of a magistrate to keep them from doing further unethical research, the editorial board of BMC Public Health should be replaced, and the authors of this paper should be forced to personally apologize to everyone who was exposed to their misleading message. Because these authors, who are extremely unethical people, are in favor of warning labels, all warning labels of any kind should be eliminated.
Ok, that does not actually follow from the analysis. Though if you accepted the pattern of “reasoning” used by Popova and Ling, it would.