In my post “Inequality Aversion Utility Functions: Would $1000 Mean More to a Poorer Family than $4000 to One Twice as Rich?” I use math and survey data on inequality aversion to argue that the big gains from redistribution are from taking care of the desperately poor. GiveWell is a website that rates charities in a way consistent with that criterion. Take a look. 

I learned about GiveWell from Michael Huemer’s excellent book The Problem of Political Authority.