I like this analysis! Some slight counter-considerations:
Displacements can also occur in donations, albeit probably less starkly than with jobs, which are discrete units. If my highest priority charity announces its funding gap somewhat regularly, and I donate a large fraction of that gap, this would likely lower the expected amount donated by others to this charity and this difference might be donated to causes I consider much less important. (Thanks to Phil Trammell for pointing out this general consideration; all blame for potentially misapplying it to this situation goes to me.)
Also, in the example you gave where about 10% of people highly prioritize cause A, wouldn’t we expect the multiplier to be significantly larger than 0.1 because conditional on a person applying to position P, they are quite likely to have a next best option that is closely aligned with yours? Admittedly this makes my first point less of a concern since you could also argue that the counterfactual donor to an unpopular cause I highly prioritize would go on to fund similar, probably neglected causes.
Displacements can also occur in donations, albeit probably less starkly than with jobs, which are discrete units. (...)
Agreed! This is a good point. I think many causes are still very funding-constrained (e.g. animal welfare and global health and poverty, at least), though, so this effect would be pretty negligible for them. This is a concern for less funding-constrained causes.
Also, in the example you gave where about 10% of people highly prioritize cause A, wouldn’t we expect the multiplier to be significantly larger than 0.1 because conditional on a person applying to position P, they are quite likely to have a next best option that is closely aligned with yours?
Yes, this might be true for the next position Q which A would take. However, the effect there might be small in expectation anyway (before discounting) if it’s another displacement to another position that would have been filled anyway, rather than a displacement into earning to give or into an otherwise unfilled position, and subsequent effects in the chain are discounted more, approaching a factor of 0.1.
I like this analysis! Some slight counter-considerations:
Displacements can also occur in donations, albeit probably less starkly than with jobs, which are discrete units. If my highest priority charity announces its funding gap somewhat regularly, and I donate a large fraction of that gap, this would likely lower the expected amount donated by others to this charity and this difference might be donated to causes I consider much less important. (Thanks to Phil Trammell for pointing out this general consideration; all blame for potentially misapplying it to this situation goes to me.)
Also, in the example you gave where about 10% of people highly prioritize cause A, wouldn’t we expect the multiplier to be significantly larger than 0.1 because conditional on a person applying to position P, they are quite likely to have a next best option that is closely aligned with yours? Admittedly this makes my first point less of a concern since you could also argue that the counterfactual donor to an unpopular cause I highly prioritize would go on to fund similar, probably neglected causes.
Agreed! This is a good point. I think many causes are still very funding-constrained (e.g. animal welfare and global health and poverty, at least), though, so this effect would be pretty negligible for them. This is a concern for less funding-constrained causes.
Yes, this might be true for the next position Q which A would take. However, the effect there might be small in expectation anyway (before discounting) if it’s another displacement to another position that would have been filled anyway, rather than a displacement into earning to give or into an otherwise unfilled position, and subsequent effects in the chain are discounted more, approaching a factor of 0.1.
I’ve updated the section to clarify. Thanks!