It was the choice of “Money gap—Large (~$86 million” in the summary that got me. It just seems immediately odd that if you think that Earning To Give to some global poverty charities is on a par with other common EA career choices in terms of marginal impact (i.e. assuming you think “poverty” should be on the table at all for us), that the size of this funding gap is the equivalent of ~$0.086pp for the bottom billion. And in fact the linked post gives a funding gap of something more like $400 million for GiveWell’s top charities alone (on top of expected funding from Good Ventures and donors who aren’t influenced by GiveWell), with GiveDirectly able to absorb “over 100 million dollars”. But it’s not so odd if you think that the expected value of donating to GiveWell-recommended charities is several orders of magnitude greater compared to the average global poverty charity. I’m aware that heavy-tailed distributions are probably at play here, but I’m very skeptical that GiveWell has found anywhere near the end of that tail (although I think they’re the best we have).
Regardless of what the author meant, I think I see this kind of thinking in EA fairly regularly, and it’s encouraged by giving the “neglectedness” criterion such prominence, perhaps unduly.
And yes, I also want to thank the author for encouraging people to think and talk about this in a more nuanced way.
I had a similar reaction.
It was the choice of “Money gap—Large (~$86 million” in the summary that got me. It just seems immediately odd that if you think that Earning To Give to some global poverty charities is on a par with other common EA career choices in terms of marginal impact (i.e. assuming you think “poverty” should be on the table at all for us), that the size of this funding gap is the equivalent of ~$0.086pp for the bottom billion. And in fact the linked post gives a funding gap of something more like $400 million for GiveWell’s top charities alone (on top of expected funding from Good Ventures and donors who aren’t influenced by GiveWell), with GiveDirectly able to absorb “over 100 million dollars”. But it’s not so odd if you think that the expected value of donating to GiveWell-recommended charities is several orders of magnitude greater compared to the average global poverty charity. I’m aware that heavy-tailed distributions are probably at play here, but I’m very skeptical that GiveWell has found anywhere near the end of that tail (although I think they’re the best we have).
Regardless of what the author meant, I think I see this kind of thinking in EA fairly regularly, and it’s encouraged by giving the “neglectedness” criterion such prominence, perhaps unduly.
And yes, I also want to thank the author for encouraging people to think and talk about this in a more nuanced way.