The Proper Response to Uncertain Risk is Overreaction

 At almost two years into a pandemic, it's only natural that we are seeing more and more takes on the policy choices (and even some personal choices) made in response to the threat of a deadly contagion. Also as expected, most of these takes are using the clarity of hindsight to criticize specific choices which were made in the face of uncertainty. While analyzing what worked, what didn't, and what we should do the next time is necessary and valuable, it can only be done properly if we accept the timeliness of the unknown vs. known.

I think the simplest way to demonstrate this concept is to quantify things, even if only an arbitrary manner. These are all actually fuzzy and complicated subjects, but the premise is the same. When faced with a new risk, even if it's just a variation on known risks, the uncertainty that comes along with it needs to be factored into any response.

So let's say we're at the start of a new, deadly virus spreading around the globe. How bad is it going to be? Some people might say it's only a 5 - about the same as a bad flu season. Others might say it's a 50 - approaching an extinction-level event. Even the level-headed experts will disagree, but if you follow the science and listen to a consensus, you can get an approximate level to use as your basis. Let's say the weighted average comes out to a 15. Great, we have a number to work with. 

But you really shouldn't base your response on a 15. If 15 is the average, then there's around a 50% chance it will be worse than that. If you react as if it's going to be a 15, there's a very strong likelihood you will end up underreacting and face grave consequences. If you wait for more certainty you will respond too late and suffer because of it, especially when facing a risk known to grow exponentially. Instead, you factor in the uncertainty by assuming it's going to be a 20 (with 90% confidence it will be less than that), react at that level knowing you will likely overreact a little bit, which generally has much lower costs than any level of underreaction. So you deliberately overreact compared to the consensus expectation.

And yes, when it ends up being a 17, some people will criticize your overreaction. Some will even say foolish things like: I predicted it would be a 17 and you should have listened to me. Of course, my cousin Jimmy predicted Tampa Bay would win the Superbowl before last season. The fact that he was right doesn't change the more important fact that he predicts that every year, so to have listened to him last season (and not all the seasons before and since) would have been foolish even though he turned out to be right.

Hindsight is valuable, and we have made a lot of improvements on our reactions and we've learned from experience and analysis. With more certainty comes better reactions. We don't worry about masking outside as much. We aren't obsessing over wiping down every surface every fifteen minutes. We know we have to take more care with the elderly, especially in group home situations. We even can keep kids in the classroom as long as we take precautions proven effective: universal masking, vaccines for those eligible, improved ventilation, and frequent testing. But it was absolutely the right call to close down schools when all those things were uncertain because we've also found out that when schools are open without proper precautions things go bad really fast. Our overreactions were good, and with the clarity of hindsight I see a lot of places, a lot of policy, and a lot of people who underreacted based on wishful thinking of assuming a level 10 just because someone, somewhere said so. With a million people dead in the U.S., tens of millions more with long-term health damages (we still don't know the extent of), those who look back and speak with certainty are not the voices to listen to. Cousin Jimmy speaks with confidence and certainty, but at the end of the day, he's from Florida. Do you really want to listen to him?