I already donated them a few thousand dollars. But how do I know if me donating a few thousand dollars makes a difference or not? I’m sure if I gave them a few hundred thousand dollars all at once I could be sure I saved some extra lives but I am not so sure for a few thousand or if it’s like voting in an election.
It sounds like you want to be the one who creates a tipping point. E.g. in your election analogy, if 51⁄100 people voted for your preferred candidate, you’d want to be the 51st person who made the difference between them winning and losing. Or if it cost $5000 to save a life (every time, not on average), you’d want to be sure you donated the 5000th dollar.
That’s totally understandable (if I’m interpreting you correctly), but I’m not sure it’s a helpful way of thinking about things in this case.
Yes, marginal cost effectiveness is important— that’s why we think about neglectedness. But once you’ve selected a cost effective charity, the usefulness of the concept breaks down a bit. (At least in IMO).
Sure, perhaps in some sense the 5000th dollar is more valuable than the other 4999, because it ticked the counter over. But it wouldn’t have been the 5000th dollar if the first dollar hadn’t been donated. In that sense, each dollar is equally valuable— they all have an equal chance of creating that tipping point.
That’s why we think probabilistically. It’s impossible to know ahead of time which dollar will make the marginal difference. So we look for the charities which give the best odds. That’s part of what GiveWell does— their recommended charities are ones they think can make the best use of marginal funding.
At that point, every dollar counts, whether it’s the 1st or the 5000th. This is a team effort, and your contributions aren’t less valuable just because someone else happened to give the final dollar 💙
I’ll try to phrase this question as: let’s say the standard EA cost-effectiveness threshold for saving a life is $5000 (that’s roughly what it is). Are there any charities (presumably EA ones) where if I donate $5,000 (enough to save a life on average), I can be
nearly certain (let’s say 95% sure) that my cash has been spent in ways that caused at least one life to be saved
that would not have been saved if I had not given them $5,000?
Great question, and I’m not sure there actually is one at present, because charities that are significantly more cost-effective than this threshold or that can demonstrate conclusive lifesaving at about this threshold tend to acquire donation reserves so additional money to them doesn’t result in them being able to spend it—that’s basically why the threshold exists, it’s sort of determined by the amount of money EA-inclined donors have available to direct.
Although actually now I think of it there might be something to do with HIV treatment now that PEPFAR money has been pulled. If someone with HIV doesn’t get HIV treatment they will die (unless they are very rarely the person who is naturally immune). And there’s a huge funding gap in that space. So if I were so inclined to care about concreteness in the way you do, that’s where I’d go hunting.
I already donated them a few thousand dollars. But how do I know if me donating a few thousand dollars makes a difference or not? I’m sure if I gave them a few hundred thousand dollars all at once I could be sure I saved some extra lives but I am not so sure for a few thousand or if it’s like voting in an election.
It sounds like you want to be the one who creates a tipping point. E.g. in your election analogy, if 51⁄100 people voted for your preferred candidate, you’d want to be the 51st person who made the difference between them winning and losing. Or if it cost $5000 to save a life (every time, not on average), you’d want to be sure you donated the 5000th dollar.
That’s totally understandable (if I’m interpreting you correctly), but I’m not sure it’s a helpful way of thinking about things in this case.
Yes, marginal cost effectiveness is important— that’s why we think about neglectedness. But once you’ve selected a cost effective charity, the usefulness of the concept breaks down a bit. (At least in IMO).
Sure, perhaps in some sense the 5000th dollar is more valuable than the other 4999, because it ticked the counter over. But it wouldn’t have been the 5000th dollar if the first dollar hadn’t been donated. In that sense, each dollar is equally valuable— they all have an equal chance of creating that tipping point.
That’s why we think probabilistically. It’s impossible to know ahead of time which dollar will make the marginal difference. So we look for the charities which give the best odds. That’s part of what GiveWell does— their recommended charities are ones they think can make the best use of marginal funding.
At that point, every dollar counts, whether it’s the 1st or the 5000th. This is a team effort, and your contributions aren’t less valuable just because someone else happened to give the final dollar 💙
I’ll try to phrase this question as: let’s say the standard EA cost-effectiveness threshold for saving a life is $5000 (that’s roughly what it is). Are there any charities (presumably EA ones) where if I donate $5,000 (enough to save a life on average), I can be
nearly certain (let’s say 95% sure) that my cash has been spent in ways that caused at least one life to be saved
that would not have been saved if I had not given them $5,000?
Great question, and I’m not sure there actually is one at present, because charities that are significantly more cost-effective than this threshold or that can demonstrate conclusive lifesaving at about this threshold tend to acquire donation reserves so additional money to them doesn’t result in them being able to spend it—that’s basically why the threshold exists, it’s sort of determined by the amount of money EA-inclined donors have available to direct.
Although actually now I think of it there might be something to do with HIV treatment now that PEPFAR money has been pulled. If someone with HIV doesn’t get HIV treatment they will die (unless they are very rarely the person who is naturally immune). And there’s a huge funding gap in that space. So if I were so inclined to care about concreteness in the way you do, that’s where I’d go hunting.
Any specific recommendations?