One Part Inspiring, One Part Distressing

The June/July COVID issue of PAW was one part inspiring, one part distressing.

The inspiring part: the ingenuity, generosity, and in some cases physical courage displayed by so many Princetonians in the face of the reported pandemic.

The distressing part: the failure to address the 800-pound gorilla in the room. Never before in history have so many actions with such drastic, far reaching, and long lasting negative consequences for so many been taken on such thin science.

I’ve also rarely seen the news media engage in such a thorough effective censorship of scientists including at least two Nobel Prize winners and tenured professors in relevant disciplines from places like Stanford and Yale. Was there no one of equal stature in the Princeton sphere whose alternative view of the pandemic’s science could have been aired? After all, our motto is “Princeton in the nation’s service,” not “Princeton in lockstep.”

Also, and this needs to be addressed, I will be charitable and assume that Professor Fleurbaey’s article “Should We Save Lives, Or Should We Save the Economy?” was submitted many months before the complete and very public discrediting of Neil Ferguson’s projection of 2.2 million dead in the U.S. without social distancing. Professor Fleurbaey’s projection of 3 million deaths, which he calculated to have a cost of $30 trillion, is off by a very large factor.

In addition to the death projection error, the value of a “statistical life” in the U.S. may be $10 million average as Professor Fleurbaey says, but we have an old and very well established industry that is experienced in calculating the “value” of lives depending on the age of the decedent so his arithmetic is off substantially there as well. Spreadsheet projections are fine, but when their product forms the basis of a ringing endorsement of specific supposedly medically-driven actions that will seriously and negatively impact the lives of hundreds of millions (possibly billions) globally, more rigor should be on display.