Classes of thing I don’t consider particularly important
Adverse selection.
There’s a battle between adverse selection and deference.
If you do zero deference while also prioritizing neglectedness, then the expected utility of your grantees will be worse than you think on average, because neglectedness is evidence that other grantmakers saw downsides that you missed.
If you do too much deference, then you’re double-counting evidence, and too much money will be granted based on other grantmakers’ incorrect views.
As far as I can figure, there’s no way around this. You just have to do it the hard way, which is to say, you have to figure out the correct amount to defer and then do that.
There’s a battle between adverse selection and deference.
If you do zero deference while also prioritizing neglectedness, then the expected utility of your grantees will be worse than you think on average, because neglectedness is evidence that other grantmakers saw downsides that you missed.
If you do too much deference, then you’re double-counting evidence, and too much money will be granted based on other grantmakers’ incorrect views.
As far as I can figure, there’s no way around this. You just have to do it the hard way, which is to say, you have to figure out the correct amount to defer and then do that.