Propaganda, Thinly Masked

A few minutes ago, about 6:40am Korea Time, I opened the internet on my computer, which automatically directed me to my MSN homepage, where I was greeted by this top headline: “Model lowers death projection as more wear masks.” 

My first thought was, “That is not a news headline, but a propaganda slogan, revealing yet again how thoroughly in the tank for the administrative state behemoth the American corporate media is.” My second thought, however, was that this was not very effective propaganda, since the headline gives away the game on its own disingenuousness.

For what does it actually mean to say that an artificial predictive model lowers its death projection “as more wear masks”? It means, very directly and unambiguously, that if the statisticians build into their model the presumption that wearing a mask absolutely prevents death from COVID-19, and also factor in the premise that more people are in fact wearing masks, then the model’s output will necessarily show a reduced number of projected deaths. After all, that reduced number was simply programmed into the model. The salient question is — and no predictive model has anything whatsoever to say about this question — “Does wearing a mask actually reduce COVID-19 deaths?”

Consider: If I build a predictive model on the premise that eating Wheaties prevents coronavirus infection, then if I factor into my calculations a recent uptick in the sale of Wheaties, my model will unfailingly return a result of “reduced infections.” How could it do otherwise?

That was what I thought about as I prepared my breakfast, imagining I would blast off a quick post about it, as part of my never-ending series on the corporate media as propaganda wing of global authoritarianism. 

So imagine my momentary dismay, and subsequent delight, when, having sat down with my bowl of bibimbap — you really must try Korean food someday, if you haven’t — I reopened the same MSN homepage and read the headline again: “Death projection lowered as more wear masks.”

Wait a moment! I could have sworn the headline had specifically included the word “model” when I had read it five minutes earlier. But of course, I quickly realized, it had included the word “model.” My entire thought process while preparing breakfast was focused on the absurdity of reporting a computer model’s death projections as major news, while hoping no one in their dumbed-down readership notices that a model as such proves nothing, and is only as reliable as the verifiability and comprehensiveness of its premises.

Obviously, in the five minutes between my first and second readings of that headline, some cleverly devious headline editor at MSN noticed the same thing I did — that their headline was easily assailable as transparent propaganda — and changed it to remove the alarm-deflating effects of the word “model.” 

Apparently, however, the editors at MSN are a little more on the ball than those at U.S. News & World Report, the source of the story linked at that headline. For the U.S. News story itself still (as of this writing) carries the headline, “Coronavirus Model Decreases Death Projection as Mask Wearing Increases.”

Here is the heart of the article:

The model, which is run by the Institute for Health Metrics and Evaluation at the University of Washington, predicts nearly 220,000 deaths in the U.S. by Nov. 1. That number is down nearly 5,000 from its previous estimate a week ago.

Researchers cited an increase in mask mandates for the public as well as more people wearing masks and practicing social distancing without mandates as reasons for the decline.

But the model shows that more work can be done on the issue of masks. If 95% of people wore masks when leaving their homes, nearly 34,000 lives would be saved by early November, according to IHME.

There is absolutely nothing in the article to justify the model’s certainty about the efficacy of masks. There is nothing at all to suggest that the results might be affected by what kind of masks people are wearing, or to indicate the proper or “necessary” contexts in which masks are useful for preventing deaths. We get only this summary claim, coming directly from the people who built the model: “If 95% of people wore masks when leaving their homes, nearly 34,000 lives would be saved by early November.”

This model, you see, is not merely a pseudo-scientific predictor; it is also a moral judge. This is about “saving lives.” If you do not wear a mask every time you leave your home, for any reason, in any context, then you are choosing not to save lives. In other words, you are guilty of causing deaths. Replace this model’s moral reasoning with an equivalent model about traffic deaths or common flu deaths, and imagine how modern life, prosperity, and liberty would have ground to a halt generations ago, had we all assumed the same position on those public dangers that these alleged scientists are pitching with regard to this particular coronavirus.

And while we are highlighting the moral presumptuousness of this model’s creators, I should note the very prominent use, in the group’s explanation of its results, of the word “mandate.” “Researchers cited an increase in mask mandates for the public as well as more people wearing masks and practicing social distancing without mandates as reasons for the decline.” That is, the primary reason for the improved death projection is “mask mandates for the public.”

The implicit argument: Not wearing a mask is immoral and will directly cause deaths; mandatory mask laws will increase mask-wearing; therefore, mandatory mask laws are a moral good.

This is why scientists, statisticians, and other “specialists” should be kept far — very far — from positions of political decision-making authority.

You may also like...