TL;DR: GiveWell just announced ~$900M of giving recommendations this year, which they estimate are all at least 10 times as effective as cash transfers (10x cash). They only expect their donors to contribute $600M to these opportunities, leaving ~$300M unfunded. I think donors who have to GiveDirectly a year ago on GiveDirectly’s recommendation would have done ~10x more good by giving to GiveWell instead (now, or earlier in the past year). I think this was very foreseeable and was disappointed by GiveDirectly’s post here saying the opposite when I saw it last year.
Hi GiveDirectly,
I’m commenting on this post from a year ago because the results of GiveWell’s research are finally in, so it’s a good time for GiveDirectly to do a retrospective on this post.
(I’m also commenting because I was very disappointed by this post when I saw it last year—I felt it went against the principle of effective altruism that says you should try to do the most good, rather than settle for doing some good, and I want to speak my mind about that finally.)
GiveWell wrote the following in its newsletter that was sent out today, November 23, 2022 (bold emphasis added by me):
We’re thrilled that our research team has identified $900 million in highly cost-effective funding opportunities this year in both the core interventions implemented by our top charities—malaria prevention, incentives for vaccination, and vitamin A supplementation (VAS)—and newer-to-us areas, such as water treatment, iron fortification, and maternal syphilis screening and treatment.
This year, we don’t expect to raise enough funds to support all of these excellent programs. If donor support is in line with our forecasts, we’ll be leaving $300 million of identified funding gaps unfilled.
As we share below, we think GiveWell is thinking too small, undervaluing what can be achieved today, underestimating the costs of waiting, overestimating how much better they’ll allocate funds in the future, and not accounting for the perspectives of people living in poverty. In short, we think the world has an unprecedented opportunity to eliminate most of extreme poverty in our lifetimes, and we worry GiveWell’s decision hinders that effort.
I strongly disagreed with this and GiveDirectly’s condemnation of GiveWell’s 2021 decision to rollover funds.
Let’s look at the two concrete claims above that we now have new information on:
1.
overestimating how much better they’ll allocate funds in the future
GiveWell’s funding bar was 5-8x cash last year, and decided to rollover funds that could not meet that funding bar to this year because GiveWell expected to identify significantly more giving opportunities above that bar. As it turns out, GiveWell did not overestimate how much better they’d allocate funds in the future. GiveWell’s announcement today says:
As expected last year, any donors who gave to GiveDirectly last year rather than wait until today to give to GiveWell’s 2022 recommendations lost out on ~90% of the impact they could have had with their money.
2. estimate their top charities at 10x cash. So it seems they did not overestimate how much better they’ll allocate funds.
undervaluing what can be achieved today
I’m highly suspicious of GiveDirectly’s claim that “[GiveWell may have been] underestimating the effects of cash transfers by 2.6x and overstating the benefits of waiting by the same amount,” but for the sake of argument assume GiveDirectly is correct. Even cash transfers have always been 2.6x as valuable as GiveWell estimates, waiting a year to do 3-4 times as much good per dollar is still totally worth it.
GiveDirectly’s claim that GiveWell was “underestimating the costs of waiting” still seems clearly false. Waiting a year to do 3-4 times as much good is a fantastic return.
But as a thought experiment, which is more likely?
Case A: A donor comes to GiveDirectly today and says, “I’m disappointed that the best giving opportunity I can identify today is GiveWell’s recommendation which is 10x as effective as cash. I wish I could go back in time a year and give my money to GiveDirectly instead.”
Case B: A donor comes to GiveDirectly today and says, “I donated to GiveDirectly last year because GiveDirectly recommended giving now after I heard about GiveWell’s decisions to rollover funds, but I just learned that if I had waited just a year I could have done ~10 times as much good with my money.”
TL;DR: GiveWell just announced ~$900M of giving recommendations this year, which they estimate are all at least 10 times as effective as cash transfers (10x cash). They only expect their donors to contribute $600M to these opportunities, leaving ~$300M unfunded. I think donors who have to GiveDirectly a year ago on GiveDirectly’s recommendation would have done ~10x more good by giving to GiveWell instead (now, or earlier in the past year). I think this was very foreseeable and was disappointed by GiveDirectly’s post here saying the opposite when I saw it last year.
Hi GiveDirectly,
I’m commenting on this post from a year ago because the results of GiveWell’s research are finally in, so it’s a good time for GiveDirectly to do a retrospective on this post.
(I’m also commenting because I was very disappointed by this post when I saw it last year—I felt it went against the principle of effective altruism that says you should try to do the most good, rather than settle for doing some good, and I want to speak my mind about that finally.)
GiveWell wrote the following in its newsletter that was sent out today, November 23, 2022 (bold emphasis added by me):
(See GiveWell’s blog post for the full details: Our recommendations for giving in 2022.)
A year ago in this post, GiveDirectly wrote:
I strongly disagreed with this and GiveDirectly’s condemnation of GiveWell’s 2021 decision to rollover funds.
Let’s look at the two concrete claims above that we now have new information on:
1.
GiveWell’s funding bar was 5-8x cash last year, and decided to rollover funds that could not meet that funding bar to this year because GiveWell expected to identify significantly more giving opportunities above that bar. As it turns out, GiveWell did not overestimate how much better they’d allocate funds in the future. GiveWell’s announcement today says:
As expected last year, any donors who gave to GiveDirectly last year rather than wait until today to give to GiveWell’s 2022 recommendations lost out on ~90% of the impact they could have had with their money.
2.
estimate their top charities at 10x cash. So it seems they did not overestimate how much better they’ll allocate funds.
I’m highly suspicious of GiveDirectly’s claim that “[GiveWell may have been] underestimating the effects of cash transfers by 2.6x and overstating the benefits of waiting by the same amount,” but for the sake of argument assume GiveDirectly is correct. Even cash transfers have always been 2.6x as valuable as GiveWell estimates, waiting a year to do 3-4 times as much good per dollar is still totally worth it.
GiveDirectly’s claim that GiveWell was “underestimating the costs of waiting” still seems clearly false. Waiting a year to do 3-4 times as much good is a fantastic return.
But as a thought experiment, which is more likely?
Case A: A donor comes to GiveDirectly today and says, “I’m disappointed that the best giving opportunity I can identify today is GiveWell’s recommendation which is 10x as effective as cash. I wish I could go back in time a year and give my money to GiveDirectly instead.”
Case B: A donor comes to GiveDirectly today and says, “I donated to GiveDirectly last year because GiveDirectly recommended giving now after I heard about GiveWell’s decisions to rollover funds, but I just learned that if I had waited just a year I could have done ~10 times as much good with my money.”