Why are employers & schools slow to close? They are weighing benefits & costs. And both have wide confidence intervals, if not ambiguity. In this context, information is marginally very valuable. And what others do provides info. So we'll see a lot of herd behavior. Eg, ... 1/6
schools all closing at once after each dithering for a while. The delay seems costly to those convinced of benefits of distancing. But not to those actors! Their decision is rational given their info set. If you are convinced we need to act fast, you need to understand this. 2/6
It'll help you figure out how to convince employers & schools to act sooner. Implications?
1/Govt can lead by example. Close your own workplaces. Even if you don't have legal authority to quarantine generally. 3/6
1/Govt can lead by example. Close your own workplaces. Even if you don't have legal authority to quarantine generally. 3/6
2/Don't call people stupid for not acting sooner. Honey better than vinegar, etc. Last thing you want is pride stopping people from updating based on your behavior. 4/6
3/We'll very likely over or under-react. Herding models say that herds don't lead to biased inference, but they do generate excess variation. Ie, false positive & negatives. 5/6
If we took an average across 10000 parallel universes with #covid19, on average they'd get the right inference. But not in any 1 universe.
PS. I think we should social distance NOW. But I'm also an actor in the game. And I admit I'm not sure which parallel universe I am in. 6/6
PS. I think we should social distance NOW. But I'm also an actor in the game. And I admit I'm not sure which parallel universe I am in. 6/6
