A little over a year ago, we ballparked the cost to save a life from malaria at $200. Now we think it’s closer to $1000. Last August, I wrote that K-12 education is my favorite cause. A week ago, I owned up to moving it to the bottom of my list, and far preferring global health (even if it’s 4x as costly as I originally guessed). And that’s not even mentioning our December Board meeting, where I walked in with three suggestions as to how we should grant and was convinced to change my mind on two of them.
If you like a man who knows where he stands, I’m not your type. I’m a newcomer to all of the many areas we’re studying, and I have a lot to learn. You’re going to see me change my mind again and again, sometimes going in a circle to come back to what seemed all along like common sense (though with updated information and reasoning behind it). Then again, the history of most areas of human knowledge can be described the same way. That’s what you can expect when you (a) have a lot to learn; (b) are willing to learn it.
The reason I’ve been flip-flopping so violently, and will probably continue to do so, is the same reason this project is worthwhile: there is not enough information out there, nor are there the right kind of information aggregators, for a donor to become well-informed in a reasonable period of time. Currently, the only option I know of for an individual donor is to, in effect, take wild guesses about whom to help, how to help them, and which organization to work with in order to do so. Because of this, GiveWell doesn’t need to provide infallible analysis in order to provide a valuable service; instead, we’re working to make our guesses better and better informed. We’re sharing our progress in real-time because we want as many donors as possible to be able to see what we’re finding and improve on their own guesses. We don’t guarantee that our recommendations will stay the same over time; but probabilistically speaking, they will be better and better.
I’m aware that I often sound authoritative and 100% confident, even when I’m not (this is an idiosyncrasy I’ve had all my life, and I am working on it). But if you look past the tone, you’ll see a flip-flopper; and given the high-complexity, low-information work we’re doing, you shouldn’t expect any less.