Policy recommendations as lies

posted by Carl V Phillips

Finishing up the series on Stephen Hecht’s latest alleged research, we focus on his policy recommendations.  How can someone’s recommendation be a lie, you might well ask.  When it is presented as if it follows from particular research, but it does not actually follow.

The juxtaposition of a scientific analysis and a policy recommendation clearly is meant to imply that the analysis is sufficient to justify the recommendation, and indeed that is often stated explicitly.  This is false even for honest recommendations in epidemiology research reports, and even more so for toxicology reports that are pretended to be epidemiology because a policy recommendation needs to be based on the full body of evidence about the effects of the exposure, as well as an analysis of the other costs and benefits of the policy.  Such information is not present in a report about one study’s results, and is seldom even asserted or referred to.

This means that most, indeed almost all, policy recommendations that appear in health science are lies, so there is nothing unusual about the THR context.  But such lies are less harmful in contexts where only scientists are reading the papers and decision-making filters through a reasonably expert process.  No one is going to react to some off-the-wall tacked-on recommendation about workplace safety or how best to deploy a screening test, and indeed, scientific readers typically just scan past those sentences like they would an ad.  But in contexts where people alter their behavior based on bad recommendations, and the science is really about politics, then such lies matter.

A recommendation might follow from the research if it is very modest and general (e.g., “this suggests we should reduce the exposure to the extent that it is easy” rather than “should be avoided completely [at all costs]” or “should be reduced to 10 ppb”) and appropriately contingent (“if the entire body of research on this topic comes to the same conclusion as this study”).  Those of you who ever look at these research reports will know that such modest phrasing is almost never used.  Instead, there is the bald lie, “these results show that we should….”

Hecht’s version of the bald lie in this case is:

“Obviously, we need to decrease the levels of this material in all smokeless tobacco products — or eliminate it altogether.”

Yeah, obviously.  The only thing that is obvious is that this conclusion does not follow from the observation that this chemical is bad for rats in large doses.  Sugar, therapeutic drugs, and cars also cause serious health problems in rats in large doses, so obviously we need to decrease them or eliminate them all together.  Since the relevant smokeless tobacco products do not actually cause measurable rates of cancer, the need to change them to reduce some apparently non-cancer-causing carcinogen seems rather less than obvious.

Hecht adds that removing (S)-NNN from these products is feasible. In fact, some products on store shelves today have reduced levels of the carcinogen.

Credit is due for these two sentences at least, a huge improvement over most “public health” activist science which does not even acknowledge that the feasibility of a recommendation matters.  Still, this is presented as an aside rather than a core part of the policy recommendation, and it is obviously not enough information (e.g., we do not know whether further reductions might be highly costly).

Still, if he were “these products are already very low risk, but all else equal, there might be some health benefit from reducing this chemical”, there is nothing wrong with that.  But rather than telling the truth about the low risk, he is instead communicating the lie “these products might be acceptable if changes were made, but they are horribly dangerous now.”  As for “all else equal”:

Hecht explained that the U.S. Food and Drug Administration has the authority to regulate tobacco products, but no regulations on the levels of specific carcinogens exist yet. “My suggestion is that levels of (S)-NNN in smokeless tobacco be decreased to below 10 parts per billion. That would make it more consistent with the levels of nitrosamines in food products,” he said. (S)-NNN also is in cigarettes and other smoked tobacco items, and he suggested that the substance be regulated in these products, as well.

Ah, there it is.  Someone not familiar with Hecht and his co-conspirators would have no idea of the subtext here, which just trying impose any rule they can rationalize on THR products (and on cigarettes).  There is absolutely nothing in his research that justifies this 10 ppb threshold, or any other specific threshold.  How could giving a megadose to rats possibly inform us about what the exact right level is?  But the goal here is just to require something that is difficult because it is difficult not because it is useful.  Then they can complain when it does not happen (“we think harm reduction is a fine idea in principle, but the products have not met the necessary conditions that we arbitrarily imposed”).  Better still (in their minds), complying might punish the consumers of the product by increasing its price or lowering its quality.

Hecht, of course, did not acknowledge his activist views in the press release — a lie of omission in itself.  A non-expert in the politics of the area would not recognize the mention of the FDA as being a statement about a particular anti-THR activist strategy, though its inclusion was obviously irrelevant to the reporting of the science.  It is a classic example of dog whistle politics, designed to signal his goal to his co-conspirators while hiding it from the vast majority of readers who would not approve of it.

A final aside on this thread:  Many people consider it acceptable to inflict pain and death on non-human animals in pursuit of knowledge that can improve the health and medical treatment of people.  There is a spectrum of a beliefs about the necessary balance of suffering to justify potential knowledge gain.  But only an extreme tail of that distribution — basically those who feel no compassion toward animals at all — would support Hecht inflicting suffering on animals in pursuit of a political hobby project which offers no conceivable benefit.  He is basically hurting animals to look for carcinogens which are not causing cancer.  A little education can save people from succumbing to his lies, but it cannot help the animals he tortures to produce his propaganda.