September 1, 2020 in Inside Story
Herd mentality or herd immunity?
SHARE: PRINT ARTICLE:
https://doi.org/10.1287/LYTX.2020.05.12
Recently, while waiting in a socially distanced, fully masked check-out line with a few other customers at a pet store (my ancient cat requires special food not available at grocery stores), a woman walked in without wearing a mask. A cashier nearest the door politely told the woman that masks are required, as noted by multiple signs posted around the entrance. Without breaking stride, the woman shouted, “Masks empower pedophiles. Look it up!” and marched full-speed ahead.
Huh? While the two or three other masked customers in line with me stood in stunned silence, it was business as usual for the masked cashier. “Happens all the time,” she said.
As far as I could tell, no one did anything to stop the woman. I’m not sure what they or I could do. Confront her? Shame her? Call the cops? Was it even a crime? In Georgia, where I live and where this happened, for many the motto “live free or die” has morphed into “live free or possibly infect others who may die.” As I write this, Georgia has the highest COVID-19 new infection rate per capita of any state in the country.
All of which brings up a critical question for certain members of the greater analytics community: When mathematical modeling and predicting the spread of the disease caused by the coronavirus, how do you accurately account for human behavior, particularly in this environment where even wearing a mask or social distancing have become political issues?
You can’t, which explains why early on, predictions regarding the number of COVID-19 cases and deaths to expect in the United States were all over the map as modelers struggled to capture human behavior. One notable prognosticator, who happens to be the president, first pegged the numbers at between zero and 15 [1]. Imperial College London, on the other hand, projected many millions of infections and 2.2 million deaths if the U.S. took no action to slow the spread of the virus [2]. The vast majority of scientists and public health and government officials pushed for mitigating measures such as shutdowns, quarantines, mask-wearing, hand-washing and social distancing, which slowed but has yet to stop the deadly virus. In the absence of vaccines and therapeutics, still others suggested quarantining the elderly and medically vulnerable, while letting the younger and stronger members of the human herd run wild until they achieved herd immunity for all. That would certainly solve the behavior modeling problem, but at what cost in terms of human lives?
One thing is for certain: If 100% of the population remained locked down until the disease ran out of human hosts, the number of infections and deaths would be a small fraction of those than if 100% of the population ignored a shutdown mandate. But at what cost to the economy and life as we know it? And good luck getting 100% of the population in a democratic country to agree on anything, especially a shutdown.
Collecting relevant data coupled with privacy issues present still more obstacles to mathematical modelers of pandemics. Such models generally start by assuming folks are rational and do what’s in their best interest, but what about people like the woman in the pet store or the millions and millions of others who continue to gather in large numbers at fraternity parties, neighborhood bars, funerals and weddings, and yes, political rallies, without wearing masks or social distancing? Let’s face it, children like to play, college students like to party, and adults like to socialize, together. After all, they’re human. Which begs the question: Will we reach herd immunity before we have safe, effective and widely available therapeutics and vaccines? The race is on. As I write this, with the number of U.S. cases of COVID-19 approaching 6 million and the number of confirmed deaths nearing 180,000, I’m rooting for anything that stops the menace. My head and heart is hoping for vaccines, but if Georgia is any indication, don’t count out the herd.
References
Peter Horner is the editor of Analytics magazine.
([email protected])