Probably a little bit of both. Rather than get into the debate about GiveWell, I'd like to point out a basic principle that the traditional philanthropic sector just doesn't get:
Openness. I've pointed out the different interpretations of openness before, but in the nonprofit sector, here we are pursuing charitable missions, where we want to help people, it sure seems like a no-brainer to be open about what you are doing.
Givewell exhibits two important pieces of openness:
(1) Publish what the charity submitted to Givewell. I'd be curious to know how many of the documents that charities submit to funding agencies can be found on their own web sites. If the material toots your own horn, publish it!
If you are bankrolling an organization, have them share their knowledge with the rest of the sector- more people will be helped.
We all know how backward the philanthropy sector is... we can't even get an aggregated RSS feed of grants made by foundations (Grantsfire seems stillborn). But what if you could get that RSS feed with a link to the application itself? A very powerful research tool for donors.
(2) Publish the selection methodology. My experience in the private foundation world was that the methodology we usually extraordinarily good. However, since there were always grants made that were clearly not motivated by the methodology, publishing the methodology just encouraged people to figure out which board member's in-law was on the board of the organization that got a big grant even though the methodology suggested they were a less effective candidate.
I tends to come down on the side of publish the methodology and don't feel embarrassed that there is favoritism. A foundation should be able to say, we have a relationship with this organizations and choose to fund them. End of story.
But having the methodology out there is huge for the sector.
Constructive debate can now happen. The metric of "Cost per significant life change" is, in my opinion a pretty bogus figure. But someone has bothered to put a methodology out. Now others can offer alternative methodologies. Hopefully the wisdom of crowds comes up with some better metrics.
Other thoughts that I don't have time to flesh out:
Givewell does something similar to the Sunlight Foundation in politics. See Larry Lessig's switch from copyright to corruption. By bringing philanthropic efforts into the light (openness) you reduce the dominance of corruption in philanthropy (which IMHO is almost as corrupt as politics, in Lessig's definition of corruption).
The nonprofit sector is almost architected to stifle innovation. Of course brash outsiders with poor social skills are the guys that innovate.
Charity Navigator (administrative costs are everything) and Givewell ($ per life saved) are probably both wrong. But the fact they are open means eventually they or someone else will get closer to getting it right.
I'm not knowledgeable enough about the sector to confidently critique anything but the lack of transparency and public discussion. And it's this problem that stops any other problems there might be from getting addressed. -Holden KarnofskyOK, just this quote alone makes me want to send them a check.