I think this argument sometimes fails when everyone is using identical or highly-correlated strategies for choosing charities, in the presence of declining returns to scale / limited room for funding of a top charity, and no coordination. Correlated errors also can matter here—and in the notional example, you’ve essentially assumed you know the exact relationship between the charitable opportunities conditional on the coin flip. Often, there is a more complex uncertainty structure, especially in the presence of moral uncertainty across domains.
Aside from this, in the presence of very large donors, the most impactful thing to do is sometimes to look where everyone else isn’t focused, since there are often high-impact opportunities that can’t scale. (With the strong caveat that you should ask someone without any stake in the question to do an impartial BOTEC to decide if there really is an opportunity.)
I would add to the caveat that you should look for an explanation for why no EA-related funds, including funds that take applicants like the EA Funds, will support the opportunity. If other funds saw the opportunity and would have been able/allowed to support it (within scope and no other restrictions), but decided not to, that’s a bad sign for the opportunity, since it suggests it’s below their bar for funding. EA Funds can make small grants to individuals and anonymous grants. They can make some political donations (e.g. Sentience Politics for their ballot initiatives in Switzerland and the Conservative Animal Welfare Foundation in the UK, as you can see in their database), but maybe not donations to parties or party candidates in particular (or maybe with limits on how much). Some opportunities might fall out of the cause scope of each fund, although sometimes the Infrastructure Fund can pick these up, e.g. Effective Environmentalism, but I haven’t seen any others. Presumably they won’t knowingly fund crime, but neither should you...
I agree with this, so essentially “Diversify your portfolio” where the “your portfolio”=the portfolio of EA as a whole
For example, let’s say you think Strong Mind’s is neglected by the community, then you, as an individual, would be of sound mind to donate all of your donations to strong minds.
Yeah, there’s a strong portfolio argument, but I do worry that it potentially makes it too easy to donate only to less effective things for personal giving, on the grounds that other EAs neglect them. To combat this, I think it makes sense for individual EAs to give in proportion to where they think the community overall should give, unless there is a specific non-scalable neglected opportunity that they are pursuing funding.
I think this argument sometimes fails when everyone is using identical or highly-correlated strategies for choosing charities, in the presence of declining returns to scale / limited room for funding of a top charity, and no coordination. Correlated errors also can matter here—and in the notional example, you’ve essentially assumed you know the exact relationship between the charitable opportunities conditional on the coin flip. Often, there is a more complex uncertainty structure, especially in the presence of moral uncertainty across domains.
Aside from this, in the presence of very large donors, the most impactful thing to do is sometimes to look where everyone else isn’t focused, since there are often high-impact opportunities that can’t scale. (With the strong caveat that you should ask someone without any stake in the question to do an impartial BOTEC to decide if there really is an opportunity.)
I would add to the caveat that you should look for an explanation for why no EA-related funds, including funds that take applicants like the EA Funds, will support the opportunity. If other funds saw the opportunity and would have been able/allowed to support it (within scope and no other restrictions), but decided not to, that’s a bad sign for the opportunity, since it suggests it’s below their bar for funding. EA Funds can make small grants to individuals and anonymous grants. They can make some political donations (e.g. Sentience Politics for their ballot initiatives in Switzerland and the Conservative Animal Welfare Foundation in the UK, as you can see in their database), but maybe not donations to parties or party candidates in particular (or maybe with limits on how much). Some opportunities might fall out of the cause scope of each fund, although sometimes the Infrastructure Fund can pick these up, e.g. Effective Environmentalism, but I haven’t seen any others. Presumably they won’t knowingly fund crime, but neither should you...
I agree with this, so essentially “Diversify your portfolio” where the “your portfolio”=the portfolio of EA as a whole
For example, let’s say you think Strong Mind’s is neglected by the community, then you, as an individual, would be of sound mind to donate all of your donations to strong minds.
Yeah, there’s a strong portfolio argument, but I do worry that it potentially makes it too easy to donate only to less effective things for personal giving, on the grounds that other EAs neglect them. To combat this, I think it makes sense for individual EAs to give in proportion to where they think the community overall should give, unless there is a specific non-scalable neglected opportunity that they are pursuing funding.