I enjoyed thinking and writing about incorporating empathy into economic models in this post, but one consideration I did not talk about was the effect of adding more people to a situation. For example, I looked at the ‘empathetic‘ prisoners’ dilemma and showed that the cooperation problem could always exist for any level of empathy < 100%. If you wanted to expand this to a general collective action problem though, the relative weight of empathy increases (ceteris paribus) simply from there being more people affected and empathized with. As a result, having an empathy of 50% with two other people is the same as having 100% empathy with one other person.

I think this is an interesting line of thought, but I skipped it in the post because it seemed to open a whole can of worms regarding diminishing vs constant empathy at scale.

I also skipped a number of related topics like information, the efficiency of cash transfers vs paternalism, and moral hazard (e.g., threatening damage to yourself to extract gains from empathetic players).