Note: I'm sort of meta-replying to all your posts in this thread here, so some of the bits I talk about come from other posts too. It seems that your analytica

Note: I'm sort of meta-replying to all your posts in this thread here, so some of the bits I talk about come from other posts too. It seems that your analytical process in this is only one-dimensional - maximize impact per dollar. This is reasonable, however it seems to ignore many possible definitions of "impact". Impact could be "lives saved" or "infections avoided" or any number of things that fit nicely in an optimization like you suggest. However, there are other forms of impact which could be considered: * Instead of maximizing number of people impacted, there is a maximization on "best impact" per person. * There are non-monetary forms of impact, in the form of giving people a feeling that someone really cares and respects them. A token donation, or some "bare minimum" bit of charity leaves many people feeling worse than they would without it, particularly when done poorly. Perhaps the optimization is on "make those who can get it, feel they really have someone on their side" * And so on. Further, as I mentioned, impact alone is not a bellwether dimension in how to figure this stuff out. There are considerations of: * what does the giver feel comfortable giving (e.g. does the target cause ethical conflict w/in the giver) * Does the giver have any non-monetary contributions as well? * Does the giver want to be semi-anonymous? * Does the giver want direct contact? * What does the giver get as a result of this? It may sound "unaltruistic" or some other BS, but just handing over cash does not satisfy some people the way direct helping does. Basically, my point is, while you have a decent metric, it is not the only one, nor is it worth discussing in isolation of the rest.