I agree with the Overall statement of this Post. Regarding a “GiveWell of X” type of organization I believe it would have to function quite differently, ideally only working on-demand instead of doing broadly aimed research for the following 2 reasons:
The target group of such an organization would have to be succeptible enough to cost-effectiveness arguments to be moved by their research, but resilient enough not to already pivot to a GiveWell charity instead. That sounds very nieche. Starting out with the mission of an established organization and convincing them that there is a better way to achieve their goal sounds like a more promising approach than “more broadly aimed research”.
In case this organization presents itself similar to GiveWell (easily browsable website with their main results, etc.) this could dillute the perceived effectiveness discrepancy to GiveWell charities. But then again most donors with the tiniest bit of alignment to EA probably Look at the numbers too much for this to become an issue.
Hey fx,
even though I definitely cannot speak on behalf of EA as a whole, I would like to add my 2 cents to this discussion. In my opinion, this sort of question is difficult because it aims at comparing different donation outcomes, that are very hard to compare. (That is a very common problem in EA and I believe that any answer given will be somewhat controversial because of this.) How many years of school attendance generated are equally good to one death prevented? How many times would you have to stub your toe such that the combined pain is more than keeping your hand in boiling water for a second? In a lot of cases, comparing different interventions in Animal Welfare combines mutiple of these “controversial comparison steps”. For example, when comparing the open wing alliance (OWA) and the shrimp welfare project, we need to decide
how much we care about shrimp in comparison to chicken and
how many minutes of living in a battery cage instead of the conditions the OWA managed to achieve are equal to one minute o f slowly suffocating to death as a shrimp.
Rethink Priorities worked a lot on the first problem.
Unfortunately shrimp do not die in a matter of seconds. Their death lasts an average of 20 minutes. Crunching these numbers, MHR comes to the conclusion, that these different interventions are surprisingly close in their effectiveness. Considering just how controversial each intermediate step was to come to this conclusion, I find it quite reasonable that people with slightly different assumptions think, that the shrimp welfare project is much more cost-effective.
I also wanted to use the opportunity to link this post from Benthams Bulldog, which I found quite nice, even though it doesn’t exactly match your question.