Cool post, I enjoyed seeing these numbers like this. I share your takeaway that global health seems more overrepresented than I expected.
“Do you feel that the numbers I’m using are misrepresentative?”
One consideration here is whether the appropriate figures to consider are “EA funding” or “all funding.” What’s the case for the former? Just that you expect EA funding to be substantially more cost-effective?
Maybe. But even then you’d ideally include non-EA funding with some sort of discount, e.g. each non-EA dollar is only counted as 0.5, 0.1, or 0.01 EA dollars. I appreciate also that EA dollars are easier to count.
I was thinking along very similar lines with ‘Limitations’ #1. It would be much better to model this as a contribution function in four dimensions, rather than only counting ‘EA dollars’.
Not only would this require more data, but one would need to assign a multiplier to each separate intervention à la GiveWell moral weights. What fraction of a ‘Global Health’ dollar is counted when Bill Gates funds vaccine research? Could be interesting for future work.
Cool post, I enjoyed seeing these numbers like this. I share your takeaway that global health seems more overrepresented than I expected.
“Do you feel that the numbers I’m using are misrepresentative?” One consideration here is whether the appropriate figures to consider are “EA funding” or “all funding.” What’s the case for the former? Just that you expect EA funding to be substantially more cost-effective? Maybe. But even then you’d ideally include non-EA funding with some sort of discount, e.g. each non-EA dollar is only counted as 0.5, 0.1, or 0.01 EA dollars. I appreciate also that EA dollars are easier to count.
I was thinking along very similar lines with ‘Limitations’ #1. It would be much better to model this as a contribution function in four dimensions, rather than only counting ‘EA dollars’.
Not only would this require more data, but one would need to assign a multiplier to each separate intervention à la GiveWell moral weights. What fraction of a ‘Global Health’ dollar is counted when Bill Gates funds vaccine research? Could be interesting for future work.
Another would be to add up the budgets of all EA and EA-recommended organizations (or EA-recommended programs).