I realized the other day that even at this late point in the pandemic there is a basic thing many people still don't understand about diseases in general that underpins a lot of how we react to epidemics.
It is a simple idea, but most people don't realize it on their own. I am a pretty clever guy myself, and while it's obvious once you're told, very few people reason this out on first principles alone, including me. It's this:
A disease with high transmissibility but low fatality rate ("IFR") results in a HIGHER collective death count than a disease with a high fatality rate and lower (or sometimes even equal) transmissibility.
This runs counter to most peoples' intuitions but is a really key thing to understand.
If a disease kills its host quickly, it will not spread as much. The ultimate number of people killed is more about how many people are infected than it is about how quickly or how "severely" people are killed.
Let's say you are a disease, and your goal is to kill as many people as possible.
You don't want to kill your host quickly (or even often). You actually want minimal symptoms in your hosts so that they will continue their regular activities and spread you around as much as possible.
Once you kill the host, it's game over for you too.
(Yeah, sometimes corpses are a vector, but corpses are way less effective at spreading disease - they're immobile, most people avoid them, etc.)
(Yeah, sometimes corpses are a vector, but corpses are way less effective at spreading disease - they're immobile, most people avoid them, etc.)
As a disease, if you kill 90% of the people you infect, and especially if you do it quickly, they won't have much time to spread you around to others.
If instead you only kill 0.1% of them and most only have very mild symptoms, they will spread the disease to many, many, many more hosts, and even though you only kill 0.1% of them, the total number of people who die is much higher.
For example, if you cause only mild (or no) symptoms in 99.9% of people you infect, but each person can infect 10 other people, and you only kill 0.1% of them, you can cause millions and millions of deaths.
(There's an iOS game called Plague Inc where you play as a disease and can adjust these sorts of variables - your goal is to kill as many people as possible, and indeed the most successful strategy is to be a disease with high transmissibility, few symptoms, and low death rate)
This leads to a very unique policy divergence between public policy (minimizing total deaths) and individual policy (minimizing risk/cost to oneself).
Take a covid-like disease as an example (because covid managed to hit this almost perfectly):
For any ONE individual, your individual risk of dying from infection is fairly low: under 0.1% if you're healthy. Thus, the "rational" strategy for the average person is to do NO mitigation and just continue living life.
Certainly as an individual, you are much more concerned about the risk of catching e.g. Ebola, which has a death rate of about 50%.
Yet, if everyone acts according to that individual strategy, almost everyone will catch COVID, and 0.1% of everyone is a very very large number - the collective number of deaths will be in the millions.
For almost all people, the calculus of "millions of other people will die" is nowhere near as pressing as "I'm almost certainly not going to die."
You know that thought experiment about "if you were given a button that if you press it, it will give you a million dollars and someone you don't know will die, would you press the button?"
Most people have an answer (usually "yes"), and they think that's the end of the thought experiment.
No, the interesting thing is when you vary the amount of money and amount of deaths, and see what peoples' answer is.
For instance, most people will still press the button for half that amount, i.e. $500k for 1 death.
For instance, most people will still press the button for half that amount, i.e. $500k for 1 death.
In fact, most people will press the button for $5 even with as many as 100 or 1000 people they don't know dying. $5 is roughly the cost to an individual of wearing a mask.
In many threads about covid measures, you've probably seen the guy who comments, "This is stupid! The disease has a survival rate of 99.9%!"
Well, yes. It is precisely BECAUSE the disease has a survival rate of 99.9% that its collective impact will end up being much much much worse if we do nothing, while its individual impact "rationally" implies that the average individual should do nothing different.
If the survival rate were 50%, it would almost certainly kill FEWER people in total.
The idea that "a disease that is unlikely to kill YOU is more collectively dangerous in the total number of deaths caused in the end" is a very unintuitive fact that I think most people never realize on their own, even if it's becomes obvious once said.
(I learned it many years ago and I've often forgotten that while it is obvious ONCE you know it, it's one of those things that very few people figure out on their own. Typically, a disease is bad/scary or deserving of concern if it's likely to kill YOU)
A disease is not like a murderer who threatens to kill you, where someone who isn't being threatened doesn't need to do anything.
A disease is a threat that can only be fought by collectively modifying individual behaviors of people who aren't being directly threatened.
A disease is a threat that can only be fought by collectively modifying individual behaviors of people who aren't being directly threatened.
(I am eliding things like long-haul covid here. My wife has had long-haul since March 2020 so I am well aware - I am focusing on a central point about an unintuitive result in epidemiology, hence I describe it as "a covid-like" disease, i.e. presuming no post-acute sequelae)
If you're interested in the game I mentioned, here it is: apps.apple.com
By the way, these are the real-life numbers for Omicron.
What follows is all sorts of consequences about what happens when a disease like that meets various societies etc etc ... any midwit can opine on those, so I'll end here.
But if you see that "this disease has a 99.9% survival rate!" comment, consider that most people have no inkling that a less fatal disease is likely to cause more deaths total than a more fatal disease.
Feel free at that point to link to this thread if you want.
"Wow, Yishan has such intriguing ideas! What other interesting thoughts does he have?"
I'm glad you asked!
I'm glad you asked!
Follow and retweet my threads, because they will get you lots of followers! No, really.
Loading suggestions...