"This seems like it specifically addresses the "..."

https://arbital.com/p/76b

by Benjamin Hoffman Dec 31 2016


This seems like it specifically addresses the "lumpy" case where a program only makes sense above a certain magnitude, but isn't so relevant for the case where lowering the expected cost of evaluating giving opportunities is the main benefit. (Carl and Paul seem to mainly have had in mind the case of returns to scale due to better evaluation, rather than charities with increasing returns to funding.)