We are presently suffering from a once-in-a-lifetime viral pandemic, and the only way to keep millions from dying is to shut down large parts of the economy. The economic costs are staggering, but we might keep hundreds of thousands of people alive, and millions more from bad health, lung damage, and the like. A lot of people, especially some conservatives and libertarians, are chafing under the restrictions. Their concerns are reasonable: the costs of the shutdowns are staggering! But I fear this is tempting many into conclusion-driven reasoning: “Oh, the reasons for the shutdown are based on bad models, but the economic costs are real, so let’s relax and stop freaking out before we destroy ourselves economically.” People are desperate for some degree of certainty so that we can figure out how to avoid paying all of these costs.This is part of what has driven the recent, severe, and frankly well-earned embarrassment of Richard Epstein for making predictions. He was almost certainly reasoning backwards: economic destruction and lost liberty are bad, and so the viral threat necessitating those costs must be being overblown. I’ve seen this all over Facebook as well.
But there’s also an error in the other direction. “Economic” costs are human costs. Human beings suffer, and often die, when they are poorer and out of work (though oddly it looks like human morality in the US has fallen off a cliff this year). Folks mocking people for caring about the economic costs aren’t taking the economic costs seriously enough and seem to prefer dehumanizing their opponents to figuring out a constructive solution.
Here’s the trouble for both sides, though: we lack the information necessary to determine, or even begin to determine, who is correct. And that’s because there’s a lot we don’t know about COVID-19, in particular its true fatality rate and how widespread it is at present. Because of that, we don’t have a sense for how many people could die, with estimates between a few thousand to millions. If we knew that social distancing would save a million lives in the US, the “economy” side of the debate would be wrong, and if we knew social distancing would only save ten thousand lives in the US, the “human” side would be wrong.
But we not only lack certainty, we don’t even know the probabilities of these outcomes. We are in the condition that Frank Knight dubbed uncertainty. We don’t know that there’s, say, a 20% chance of 1 million people dying, a 60% chance of 100,000 people dying, and a 20% chance of 10,000 people dying. Even in that case, we could make trade-offs, and we could debate which trade-offs were correct.
Until then, we’re all going to desperately grasp for knowledge of probabilities, but we must be careful not to manufacture probabilities without the very best information available. Until then, well, we should … Well, I’m not exactly sure because I don’t know the probabilities. I guess extreme caution is called for? Precautionary principle, save us all!