Various incarnations of Dr. Michele Gelfand's work examining cultural styles and pandemic consequences have appeared for the last month in assorted venues. I'm recommending a read of her paper in The Lancet (Planetary Health) titled "The relationship between cultural tightness-looseness and Covid-19 cases and deaths: a global analysis." The paper was published January 29. I'm also recommending an opinion piece by Gelfand published in The Guardian on February 1. This brief but informative latter piece fleshes out the implications of Gelfand's work for a general audience and is titled, "Why countries with 'loose,' rule-breaking cultures have been hit harder by Covid."
Gelfand's work provides a broad, expert perspective on the effects cultural personas have had on real world virus consequences. The articles listed above broach the issue of how cultures that have not experienced recent threats fail to react to new threats in a timely, prudent manner.
These pieces by Gelfand make their own arguments clearly, and I'm not going to obfuscate their clarity with my personal prattling.
Simplistic Superficial Stats
I wanted to tackle talking points that have popped up here in the states regarding relationships between Covid-19 stats and shut-downs. The talking points are usually in defense of red-tinged locales' decision-making or lack thereof. These talking points argue against many of Gelfand's conclusions, so I'll address them.
People who are severely critical of shut downs (I'll call them "shut-down deniers") argue that U.S. stats don't currently demonstrate that states' looseness or tightness regarding shut downs and mask-wearing affected infection rates or deaths. The shut-down deniers are saying, in essence, that California and New York had the most onerous restrictions so they should have had lower infection and death rates. I realize that this seems a very naive position, but I have actually encountered professional gamblers (quite fluent in practical statistics) who make these arguments until they are blue in the face. Their political worldviews color their ability to ask the most basic statistical questions.
This is one of those situations where the fact that no correlation exists between more stringent rules and better virus results actually makes the point that the more stringent rules worked. Some of the reasons are obvious.
First, the more stringent rules and longer shutdown areas contain large urban centers. Because urban centers have higher population densities, the transmission opportunities and larger daily virus load exposures should lead to higher transmission and death rates. Shutdowns and tighter regulations have the net effect of creating an environment that mimics lower population density. If locales containing urban centers don't have massively higher transmission and death rates, this demonstrates that the regulations were indeed effective.
Second, because urban areas are landing points for international visitors carrying the virus initially and now carrying variants, the expectation should again be that locales with urban centers should yield much higher transmission and death rates.
Third, regulations in all of the United States, both red and blue, were quite relaxed compared to almost all other countries, and especially compared to countries rated "tight" by Gelfand. The perception of U.S. rule variation is exaggerated by the very parochial American perspective. Most everything in the U.S. has, compared to other countries, been undisciplined and loose.
Fourth, the variants are likely embedded deeply in major urban areas and have not yet penetrated much into places like Montana or the Dakotas.
Fifth and most important, the shutdown deniers here in the U.S. have to forcibly ignore all of the evidence from everywhere else in the world. That's a tough hurdle, but self-absorbed Americans seem to have a knack for it.
All of these points are, again, quite obvious, and actually make a strong case that more stringent rules have been useful.
Conclusion
The lack of overall context inherent in the deniers' position reminds me of a parallel argument in sport science. About 15 years ago, researchers argued against the existence of a real world "hot hand" in basketball and claimed it was all an illusion. The very basic NBA stats used in the analyses were presented without context. A lack of correlation was taken as evidence to prove the position. When context was painted in, however, the lack of simplistic correlation turned out to be evidence for a real world "hot hand," although not for a simple statistical "hot hand."
Sometime in the next month, since it's March Madness, I'll return to the hoops studies and explain their debunking. It's one of my favorite statistical rants.
For now, however, I'll wrap things up by saying that a lack of simple, clear-cut correlations between U.S. city or state regulations and transmission/death rates does not mean that those regulations had no consequences. The lack of simple correlations is actually evidence that the more stringent rules and shutdowns worked. Political prisms are generating arguments without statistical context, and the arguments are largely wrong.
Bob Dietz
March 2, 2021
