I recently realized, thanks to a FB comment by Paul Christiano, that this is thinking about things in kind of the wrong way. R is approximately 1 because society is tamping down infection rates when infections are high and ‘loosening’ when infections are low. So, by infecting people, you cause some chain of counterfactual infections that perhaps ends when society notices and tamps down infection, but also you cause the rest of society to do less fun interacting in order to tamp down the virus. So the cost of infecting somebody is to cause everybody else to be more conservative. I’m still not quite sure how to think about that cost tho.
I recently realized, thanks to a FB comment by Paul Christiano, that this is thinking about things in kind of the wrong way. R is approximately 1 because society is tamping down infection rates when infections are high and ‘loosening’ when infections are low. So, by infecting people, you cause some chain of counterfactual infections that perhaps ends when society notices and tamps down infection, but also you cause the rest of society to do less fun interacting in order to tamp down the virus. So the cost of infecting somebody is to cause everybody else to be more conservative. I’m still not quite sure how to think about that cost tho.