by Carl V Phillips
In the previous post, I quoted from Jon Stewart’s farewell monologue in which he alluded to how it is usually relatively easy to identify utterly bullshit claims and call them out. This includes utterly junk science. There are stories of master fraudsters in science, who carefully cook data and convince the world for years they have made game-changing discoveries, only getting caught after too much contrary evidence piles up. For some immediately detectable cases of junk science, it requires a bit of clever expert analysis to detect it. But these cases should not distract from the fact that most junk science is junk on its face.
Consider the claims by Glantz et al. about a gateway effect from e-cigarettes. In my recent paper, I show that a truth-seeking analysis of the same data Dutra and Glantz used in one of their papers produces results opposite what would be predicted by the gateway hypothesis. This was a pretty obvious analysis to me, and the number crunching was easily carried out by a couple of colleagues who had mastered working with that dataset. But I understand that it was nothing the average reader could work out or would even think to suggest. Analysis like that is my speciality. Here and here are a couple of other recent examples — fairly simple and obviously right, but not trivial or obvious to most readers — that Brad Rodu and I published about other anti-THR claims. Here is my favorite among the non-proprietary such analyses I have done on other topics.
But the thing is, this level of analysis is not needed to clearly identify that the Glantz et al. claims are bullshit. Their claim that e-cigarette use leads to smoking among teenagers is based entirely on the observation that there is an association between using/trying one product and the other. It requires nothing more than this observation to show that they are spouting pure bullshit. If teenagers are using e-cigarettes for THR, there will be such an association (i.e., the causation flows in the opposite direction). If it is merely the case that each product is likely to be tried or used by the same type of people, there will be such an association (i.e., confounding). There is no doubt that the confounding exist and there are good reasons to believe the opposite-direction causation exists. Therefore it is impossible to draw any conclusions about gateway effects from their observation. QED, bullshit.
The main lesson that we can draw from this — I guess it is a scientific conclusion about politics rather than a science lesson, but whatever — is that there is no reality check for the scientific claims about tobacco. Such blatant bullshit could not survive any truth-seeking scrutiny. This is not exactly news to anyone familiar with the field, but it is still interesting to come at it from this angle. If there were any demand for accuracy from anyone with authority or substantial power, such absolutely blatant bullshit would not be the norm.
Contrast with this a bit of anti-bullshit that has been making the rounds in THR circles. It seems that an Australian Senate committee absolutely excoriated long-time anti-THR and anti-tobacco liar Simon Chapman. The only problem is, it was not about tobacco (nevertheless, click and read it — it is great). It seems that Chapman, in spite of being one of those guys who pushes the “Big Tobacco are all evil!!!” trope long after it stopped being true, has spent a lot of time over the last few years carrying water for Big Energy (who actually are evil). Specifically, he produced junk science for them that is designed to engineer doubt about the overwhelming evidence that large wind power generators create health problems for nearby residents. (More details on Chapman’s various evils: via me, via Snowdon.)
For those who may not know, industrial-scale wind power is one of the greatest boondoggles of our time. It is ridiculously expensive compared to other source of energy — subsidized out of our pockets, with the money flowing to big engineering and energy companies. Those who profit have tricked the eco-types into believing it is a good way to reduce carbon emissions, but it turns out that after its various impacts and inefficiencies are considered, the net effects there are estimated to be about a wash. It turns out that there is no scientific basis for the claim it reduces carbon emissions; people just repeat the claim because if you do not understand the technology it seems like it ought to be true. Meanwhile, wind turbines do enormous real environmental damage and cause health problems for many nearby residents. Big wind turbines make a lot of low-frequency noise, which is known to cause the collection of stress-related diseases that have been observed.
Enter Chapman and his bizarre crusade to use “Big Tobacco”-style tactics to deny the obvious cause of some serious health problems (by which I refer to the tobacco companies of c.1970, of course — this has not occurred for decades), and to pursue a campaign of attacking and ridiculing the people who are suffering the health problems. The mind just boggles at the latter: Even if one is convinced that people are wrong about the source of their disease, what kind of twisted individual laughs at disease victims? We might agree that Curt Schilling et al. are not just wrong but harmful fools for blaming their cancers on smokeless tobacco, and certainly that the supposed experts who amplify their baseless claims are liars who deserve to be brutally criticized But can you imagine jesting about the cancer victims themselves? That is exactly what Chapman did.
Chapman is a perfect epitome of the dishonesty and hatred that permeates the core of “public health”. So it was delightful to see the Australian Senate point out that Chapman was drawing conclusions that are beyond his skill set (as he does about tobacco), making numerous claims that are bright-line false (as he does about tobacco), basing his “logic” on identifiable bits of scientific nonsense (as… well, you get the idea), and relying on word games (as…).
But consider the contrast: I am aware of no similar condemnation by government officials of him or the others who commit the same blatant crimes against science and humanity regarding tobacco use. Ever notice how anti-tobacco people try very hard to stay out of court, except for state court cases related to the simple liability claims about cigarettes? If they believed what they were saying, they would be filing federal lawsuits left and right. But — all lawyer jokes aside — federal judges generally have the skills to see through blatant bullshit and are reasonably likely to issue a ruling that exposes some of it. By contrast, even those politicians who can see through the junk seldom dare stand up to the tobacco control industry. So political hearings and captured agencies insulate the ANTZ bullshit from real scrutiny.
There is a widespread belief that more good scientific research about tobacco products and their use can win the day for e-cigarettes and THR more generally. But the science already shows that many core ANTZ claims about tobacco are the same patent bullshit cited in that indictment of Chapman. Thus the widespread belief is pretty clearly wrong. I obviously do not know for sure what can win the day, but I am fairly certain that the most useful scientific analysis is at the meta level, analyzing the claims themselves and those making the claims, rather than producing more science that informs accurate claims.
As an epilogue to this, one thing that was missing from the indictment of Chapman (as far as I know) was that deeper level of analysis about the fundamental anti-scientific nature of his “research” in the area. Its omission is my fault — it is my analysis and I did not take the time to contribute a comment to this particular tribunal — though in fairness to myself, it might have been too technical to make the cut when he was caught blatantly contradicting simple facts and established science.
Chapman’s “research” on this topic consisted of gathering the various published case reports from people who believe they are suffering the health effects caused by wind turbines and then looking at the noise rather than the signal, exactly opposite of what a real scientist or truth-seeker would do. To explain: There is a clear constellation of diseases — emotional stress, insomnia, inability to concentrate, and the like — that are known to be caused by stimuli like low-frequency noise and that are observed in many people living near wind turbines. These happen to be diseases for which cause and effect can be convincingly determined by the case-crossover experiments found in these individual case studies.
But such unstructured reports about people’s health outcomes are inevitably going to include various other events, like cases of cancer, that just happened to occur at the same time. Indeed, an ideal case report lists all of these, including those that are not believed to be caused by the exposure in question, because that is potentially useful information for various reasons (among them, it might let scientists discover an apparent effect of the exposure that they did not predict). A real scientist would sort through these reports to find the common outcomes, looking especially for those that are most plausibly causal, and this has been done. What Chapman did was make a list of all the random disease outcomes that were mentioned in one or more reports and his “analysis” basically consisted of saying, “ha ha, if you believed these idiots you would think that wind turbines caused every disease under the sun! Therefore we should conclude they cause no diseases.”
Yes, I am sure there is a case report out there where someone insists that her colon cancer was caused by nearby wind turbines, rather than merely mentioning the random outcome for completeness, just as Schilling insists his cancer was caused by smokeless tobacco. But so what? That is noise. The signal is the cluster of related diseases that appear over and over again. That is what a real scientist looks for. What Chapman did is similar to those who try to deny the effects of smoking with silliness like “some smokers live to 100” or “lots of people who never smoked get lung cancer”; that is noise, but we also have signal.
In retrospect, I do wish I had gotten that point on the record in this case. It is not a huge deal in a practical sense; my analysis of scientific epistemology hardly matters when piled on top of the indictment of Chapman’s bullshit on simpler grounds. Still, I think there is a lot of value in the long run in pointing out exactly how ostensible scientific analysis was corrupted to generate the bullshit.