I’ve only spoken to a few donors, mostly friends and family. They seemed to especially like the idea of donating to charities that are local to their communities, which most highly cost-effective charities don’t, because they are mostly in low income places, not the USA. They were also concerned with what portion of the donated money would be spent on the charity’s own overhead or administration. And they especially didn’t want to receive spam mail or email from charities as a result of donating.
I suppose my assumption is that I should start with an argument I think would work on myself, if I hadn’t studied global health as much as I have, and then try to figure out what part of that people are resistant to. If I hadn’t studied global health, and wanted to be able to donate effectively without having to study it, I would have been looking for that upper limit with a likelihood of somewhere around 95% that the cost would be below it, even if the upper limit needed to be pretty high in order to reach that level of certainty. So now I’m trying to figure out such an upper limit myself.
I again want to say that I resonate a lot with what you’re trying to do, since I’ve tried (and mostly failed) to do something similar myself before.
I worry a bit that you’re prematurely optimising approach-wise when you conclude the thing to focus on is figuring out the cost-effectiveness upper limit at which you can tell people honestly and confidently that their donation is doing that much good, instead of asking donors how they think about giving (which you did). For instance, Sawyer’s comment reiterates the sentiment I mentioned earlier that
Most potential donors are not really risk neutral, and would rather spend $5,001 to definitely save one life than $5,000 to have a 10% chance of saving 10 lives. Risk neutrality is a totally defensible position, but so is non-neutrality. It’s good to have the option of paying a “premium” for a higher confidence (but lower risk-neutral EV).
Orthogonally, I think most people are willing to pay more for a more legible/direct theory of impact.
“I give $2800, this kid has lifesaving heart surgery” is certainly more legible and direct than a GiveWell-type charity. In the former case, the donor doesn’t have to trust GiveWell’s methodologies, data gathering abilities, and freedom from bias.
and these aren’t necessarily obvious in advance, so if you start from a (simplistic) model of giving advisory effectiveness as being [number of prospective donors in your circle] x [fraction of donors who find argument X persuasive] x [$ per donor] x [expected good per $], then starting with an argument that works on yourself (and on me too — I don’t think we’re that representative of the donor pool) doesn’t let you get empirical input on the 2nd term, while asking donors does.
To some extent, I’m also trying to figure out a good upper limit for myself, so that I have a better idea of how much my donations are really worth. I think if I can increase my confidence that the cost is below some value, I’ll have an easier time motivating myself to avoid spending money so that I can donate more.
Going back to my original post, the only reason I’m concerned with figuring out how to factor in costs from organizations like the World Health Organization, GiveWell, and the Centre for Effective Altruism is because, from an average-based utilitarian theory, I think family planning charities in sub-Saharan Africa can save the life of a kid under 5 years old for somewhere around $275. I would normally guess that costs of organizations like those would be so tiny, that they’re not worth trying to account for. But $275 per life (which might be about $4.50 per DALY) is so cost-effective, that costs from those necessary organizations could have a pretty big impact on the overall cost.
From what I’ve read, health interventions in sub-Saharan Africa often have a cost-effectiveness of about $150 per DALY, or about $50-$80 for one of GiveWell’s top charities. And I think the costs for all of those organizations together might add up to about 5-10% of the money that goes to health interventions in sub-Saharan Africa. If we assume a normal cost per DALY is about $100 and add 5-10% of that to my estimate of Lafiya Nigeria’s cost-effectiveness of $4.50 per DALY, that results in about $10-$15 per DALY, approximately doubling or tripling the cost per DALY.
And it seems like GiveWell’s cost-effectiveness analyses doesn’t account for these kinds of costs. I understand why they wouldn’t want to, since most people looking at their cost-effectiveness analyses are trying to compare one charity to another, not trying to figure out the total cost. But, still I want to try to account for it in my calculations.
Thanks for elaborating David, I think I better understand where you’re coming from now. I still don’t buy the need to incorporate these costs though, so in the interest of potentially changing my mind and to prevent talking past each other let me try to be concrete.
The model reports the effectiveness of an intervention per philanthropic dollar spent, but this is (subtly) incorrect. The decision problem facing GiveWell is not exactly finding the most cost-effective charity per philanthropic dollar, but rather finding the most cost-effective charity that my philanthropic dollar can contribute towards. So, the cost-effectiveness of a particular charity per dollar I personally donate ought to be the effectiveness that that dollar brings, plus any matched additional philanthropic funding that dollar generates – but I’ve still only spent one dollar when it comes to the ‘cost’ part of ‘cost-effectiveness’. The idea is that you might abstractly be interested in the total cost of an intervention vs its effects, or even the total philanthropic cost of an intervention vs its effects, but GiveWell is concretely interested in the good done by a marginal donation, so only the first actor matters. Hopefully the diagram below illuminates more than it confuses!
Applying this approach to (say) Lafiya Nigeria’s figures, Rethink Priorities estimates that if you donated $361 and Lafiya spent it in Q1 ’25 you would have averted one additional death in expectation. Supposing you included costs from other orgs like WHO etc, sure the total cost may increase, but it’s still the case that your $361 helped avert one additional death in expectation! You could think of those other overhead costs as “leveraged funding” or “other people’s money moved by your money” or whatever, but why would any of that be relevant to the fact that an additional death was averted because you decided to donate $361? Shouldn’t that averted death be what ultimately matters?
Maybe you’re thinking from a credit attribution perspective? e.g. you want to say “yes a life was spared, but I shouldn’t get all the credit, since my money moved WHO’s money etc, so if we did proper accounting (maybe using Shapley values) maybe I can only get 40% credit for averting that neonatal death or averting that woman from dying in childbirth, whereas WHO gets 30% credit and other actors get 30% credit”. But in that case I don’t see where this over-focus on credit attribution is coming from, instead of just wanting to avert the death of that child or mother and knowing that you the donor can do so with $361?
I think I understand your perspective, but I think there are two different ways of looking at any particular charity opportunity, and that is only one of them. I think each applies in different circumstances to different extents, and it can be difficult to tell how much each applies, depending on circumstances.
The first way of looking at it is the one that you described, where some other organization has already paid for its part of the health intervention, and would have paid for it regardless of if the donor donated any money. In that case, it might seem reasonable to me to mostly ignore the costs from the other organization, and only look at the cost paid by the donor when doing the CEA.
One example of where I think this would be applicable is when looking at a health intervention that has diminishing returns. For example, looking at AMF, probably the places where the LLINs will be most effective and least costly to distribute tend to receive them first. This could easily result in later LLINs being less cost-effective than earlier LLINs, so it might make sense to just look at what kind of impacting the donor is adding rather than looking at the total cost and impact of AMF.
The second way of looking at it is that the costs paid by the other organization were necessary for the health intervention, so we should account for those costs, although we would also need to somehow account for any other impacts from what the other organization spent. I think this way of looking at it is particularly valid if the money from the other organization will only be spent if the donor makes the donor’s donation, but I don’t think that’s a requirement.
For example, on the AMF website, it says “100% of public donations buys long-lasting insecticidal nets (LLINs). An LLIN costs US$2.00.” But there are actually many other necessary costs that need to be paid in order for LLINs to be effective, like the costs of distribution and the costs of figuring out where additional LLINs are needed. So, if we assumed that $2/LLIN was the only cost being paid, we would end up with incorrect cost-effectiveness numbers.
One thing that I think can help decide which perspective to use is the relative cost-effectiveness of the other organization’s impact alone vs. the donor’s donation alone. If the cost-effectiveness of the other organization’s donations is a lot better than the one for the primary donor’s donation, then I think it might be safe not to account for the costs of the other organization’s donations. But, if not, I think it might make sense to account for the other organization’s costs.
For example, let’s say an organization spends $100 on a health intervention that saves one life. Then the donor comes along and pays another $10 to save a second life, but that donor can only do it because of the $100 that the organization already paid. In this case I think the more accurate cost-effectiveness for saving a life would be ($100 + $10) / 2 = $55, rather than just $10. However, if someone is just looking at which charity to donate to rather than trying to figure out realistically how many lives will be saved by a given donation, I could see the applicable number being $10/life. That’s because I could see someone from the charity saying, “Hey, somebody else made this $100 donation, but didn’t give us the other $10, even though it would make the intervention a lot more cost-effective, and it would be great if you could make up the difference.” That seems like a good donation to make. But if someone asked, “How many lives do your donations tend to save?”, I think the $55/life number would be more accurate.
In my case, I think the cost-effectiveness of Lafiya Nigeria is dramatically better than the cost-effectiveness of most global health interventions, even GiveWell’s top charities, so I think accounting for the overhead costs from other organizations might make sense. Although, maybe only in the context of figuring out how far the money will actually go, not necessarily in the context of trying to figure out which charity to donate to.
Also, to some extent, my cost-effectiveness number for Lafiya Nigeria just seems too low to be possible, so I’m trying to figure out what might be wrong with it. Trying to account for costs from organizations like the WHO seems like a promising path for doing that.
Thanks for all the feedback on this. I think I’ve done a lot more useful thinking regarding this as a result.
I’ve only spoken to a few donors, mostly friends and family. They seemed to especially like the idea of donating to charities that are local to their communities, which most highly cost-effective charities don’t, because they are mostly in low income places, not the USA. They were also concerned with what portion of the donated money would be spent on the charity’s own overhead or administration. And they especially didn’t want to receive spam mail or email from charities as a result of donating.
I suppose my assumption is that I should start with an argument I think would work on myself, if I hadn’t studied global health as much as I have, and then try to figure out what part of that people are resistant to. If I hadn’t studied global health, and wanted to be able to donate effectively without having to study it, I would have been looking for that upper limit with a likelihood of somewhere around 95% that the cost would be below it, even if the upper limit needed to be pretty high in order to reach that level of certainty. So now I’m trying to figure out such an upper limit myself.
I again want to say that I resonate a lot with what you’re trying to do, since I’ve tried (and mostly failed) to do something similar myself before.
I worry a bit that you’re prematurely optimising approach-wise when you conclude the thing to focus on is figuring out the cost-effectiveness upper limit at which you can tell people honestly and confidently that their donation is doing that much good, instead of asking donors how they think about giving (which you did). For instance, Sawyer’s comment reiterates the sentiment I mentioned earlier that
and Jason’s comment seems relevant as well:
and these aren’t necessarily obvious in advance, so if you start from a (simplistic) model of giving advisory effectiveness as being [number of prospective donors in your circle] x [fraction of donors who find argument X persuasive] x [$ per donor] x [expected good per $], then starting with an argument that works on yourself (and on me too — I don’t think we’re that representative of the donor pool) doesn’t let you get empirical input on the 2nd term, while asking donors does.
To some extent, I’m also trying to figure out a good upper limit for myself, so that I have a better idea of how much my donations are really worth. I think if I can increase my confidence that the cost is below some value, I’ll have an easier time motivating myself to avoid spending money so that I can donate more.
Going back to my original post, the only reason I’m concerned with figuring out how to factor in costs from organizations like the World Health Organization, GiveWell, and the Centre for Effective Altruism is because, from an average-based utilitarian theory, I think family planning charities in sub-Saharan Africa can save the life of a kid under 5 years old for somewhere around $275. I would normally guess that costs of organizations like those would be so tiny, that they’re not worth trying to account for. But $275 per life (which might be about $4.50 per DALY) is so cost-effective, that costs from those necessary organizations could have a pretty big impact on the overall cost.
From what I’ve read, health interventions in sub-Saharan Africa often have a cost-effectiveness of about $150 per DALY, or about $50-$80 for one of GiveWell’s top charities. And I think the costs for all of those organizations together might add up to about 5-10% of the money that goes to health interventions in sub-Saharan Africa. If we assume a normal cost per DALY is about $100 and add 5-10% of that to my estimate of Lafiya Nigeria’s cost-effectiveness of $4.50 per DALY, that results in about $10-$15 per DALY, approximately doubling or tripling the cost per DALY.
And it seems like GiveWell’s cost-effectiveness analyses doesn’t account for these kinds of costs. I understand why they wouldn’t want to, since most people looking at their cost-effectiveness analyses are trying to compare one charity to another, not trying to figure out the total cost. But, still I want to try to account for it in my calculations.
Thanks for elaborating David, I think I better understand where you’re coming from now. I still don’t buy the need to incorporate these costs though, so in the interest of potentially changing my mind and to prevent talking past each other let me try to be concrete.
Approach-wise, I think froolow got the accounting right when he was looking into GiveWell’s older (2022) CEAs (emphasis mine):
Applying this approach to (say) Lafiya Nigeria’s figures, Rethink Priorities estimates that if you donated $361 and Lafiya spent it in Q1 ’25 you would have averted one additional death in expectation. Supposing you included costs from other orgs like WHO etc, sure the total cost may increase, but it’s still the case that your $361 helped avert one additional death in expectation! You could think of those other overhead costs as “leveraged funding” or “other people’s money moved by your money” or whatever, but why would any of that be relevant to the fact that an additional death was averted because you decided to donate $361? Shouldn’t that averted death be what ultimately matters?
Maybe you’re thinking from a credit attribution perspective? e.g. you want to say “yes a life was spared, but I shouldn’t get all the credit, since my money moved WHO’s money etc, so if we did proper accounting (maybe using Shapley values) maybe I can only get 40% credit for averting that neonatal death or averting that woman from dying in childbirth, whereas WHO gets 30% credit and other actors get 30% credit”. But in that case I don’t see where this over-focus on credit attribution is coming from, instead of just wanting to avert the death of that child or mother and knowing that you the donor can do so with $361?
I think I understand your perspective, but I think there are two different ways of looking at any particular charity opportunity, and that is only one of them. I think each applies in different circumstances to different extents, and it can be difficult to tell how much each applies, depending on circumstances.
The first way of looking at it is the one that you described, where some other organization has already paid for its part of the health intervention, and would have paid for it regardless of if the donor donated any money. In that case, it might seem reasonable to me to mostly ignore the costs from the other organization, and only look at the cost paid by the donor when doing the CEA.
One example of where I think this would be applicable is when looking at a health intervention that has diminishing returns. For example, looking at AMF, probably the places where the LLINs will be most effective and least costly to distribute tend to receive them first. This could easily result in later LLINs being less cost-effective than earlier LLINs, so it might make sense to just look at what kind of impacting the donor is adding rather than looking at the total cost and impact of AMF.
The second way of looking at it is that the costs paid by the other organization were necessary for the health intervention, so we should account for those costs, although we would also need to somehow account for any other impacts from what the other organization spent. I think this way of looking at it is particularly valid if the money from the other organization will only be spent if the donor makes the donor’s donation, but I don’t think that’s a requirement.
For example, on the AMF website, it says “100% of public donations buys long-lasting insecticidal nets (LLINs). An LLIN costs US$2.00.” But there are actually many other necessary costs that need to be paid in order for LLINs to be effective, like the costs of distribution and the costs of figuring out where additional LLINs are needed. So, if we assumed that $2/LLIN was the only cost being paid, we would end up with incorrect cost-effectiveness numbers.
One thing that I think can help decide which perspective to use is the relative cost-effectiveness of the other organization’s impact alone vs. the donor’s donation alone. If the cost-effectiveness of the other organization’s donations is a lot better than the one for the primary donor’s donation, then I think it might be safe not to account for the costs of the other organization’s donations. But, if not, I think it might make sense to account for the other organization’s costs.
For example, let’s say an organization spends $100 on a health intervention that saves one life. Then the donor comes along and pays another $10 to save a second life, but that donor can only do it because of the $100 that the organization already paid. In this case I think the more accurate cost-effectiveness for saving a life would be ($100 + $10) / 2 = $55, rather than just $10. However, if someone is just looking at which charity to donate to rather than trying to figure out realistically how many lives will be saved by a given donation, I could see the applicable number being $10/life. That’s because I could see someone from the charity saying, “Hey, somebody else made this $100 donation, but didn’t give us the other $10, even though it would make the intervention a lot more cost-effective, and it would be great if you could make up the difference.” That seems like a good donation to make. But if someone asked, “How many lives do your donations tend to save?”, I think the $55/life number would be more accurate.
In my case, I think the cost-effectiveness of Lafiya Nigeria is dramatically better than the cost-effectiveness of most global health interventions, even GiveWell’s top charities, so I think accounting for the overhead costs from other organizations might make sense. Although, maybe only in the context of figuring out how far the money will actually go, not necessarily in the context of trying to figure out which charity to donate to.
Also, to some extent, my cost-effectiveness number for Lafiya Nigeria just seems too low to be possible, so I’m trying to figure out what might be wrong with it. Trying to account for costs from organizations like the WHO seems like a promising path for doing that.
Thanks for all the feedback on this. I think I’ve done a lot more useful thinking regarding this as a result.