Every so often you hear some hotshot, usually the corporate or government type, say something like "we want to be evidence based." The term "evidence based" gained a stronghold in medicine and had become commonplace in other arenas. The problem is, as I've written in other places, the notion of "evidence based" decision making is far more nuanced than many realize. It cannot be broadly applied in most disciplines outside of those physical, deterministic domains --- e.g. physics, certain areas of medicine.
The concept is typically used with great ignorance of emergent behavior and downside asymmetry (as my tweet says). Emergent behavior means things happen (emerge) in complex systems that cannot be predicted. The interdependence of the factors are governed by power laws. And downside asymmetry simply means there is greater (usually far greater) downside than upside. In other words, emergence means we don't know how the "evidence" is going to change over time; so it doesn't mean much! And downside asymmetry means the price to be paid, the damage that will be done, if we're wrong isn't worth the risk.
My tweet was in response one from the complexity specialist Joe Norman, who was responding to a tweet from the World Health Organization (WHO). WHO had used the term in their public communications about the Coronavirus. The problem is the evidence changes over time. Who here believes we shouldn't take a potential epidemic extremely seriously because it might have negative ramifications for GDP growth? Not using caution with complex systems where we can't possibly predict the outcome is a recipe for all manner of disasters. Luckily (if we're feeling generous in how we want to state this) most of those disasters --- e.g. over prescription of medicines, wars, economic booms and busts --- have been slow moving enough to not wipe out a sizable portion of the planet. At least not all at once, and that's debatable.
When it comes to potential epidemics, because something hasn't hit close to home in a long time is a very bad reason to assume it cannot or will not.
