Thank you for creating this! I want to understand some possible risks to my value system better. So here’s one scenario that I’ve been thinking about.
I realize that it’s a trust system but if Donor A trusts Donor B on something that is not clear enough to Donor A to be able to ask and that’s so unremarkable to Donor B that they see no reason to tell Donor A about it any more than their espresso preferences, then no one is really at fault if they miscommunicate.
Say Donor A:
is neutral between Rethink Priorities (RP) and the Against Malaria Foundation (AMF) (but Donor B doesn’t know this),
can get tax exemption for a donation to RP but not AMF, and
wants to donate $2k.
And Donor B:
values a dollar to RP more than 100 times as highly as one to AMF (but Donor A doesn’t know this),
can get tax exemption for a donation to AMF but not RP, and
wants to donate $1k.
Without donation swap:
Donor A is perfectly happy, and donates $2k to RP because of the tax exemption as tie breaker (but even if they split the donation 50:50 or donate with 50% probability, this case is still problematic).
Donor B is a bit sad but donates, say, $850 to RP, which comes down to the same cost to them due to the lacking tax break.
In effect: RP gains $2,850 and AMF gains $0. Both donors are reasonably happy with this result, the only wrinkle being the taxes.
But with donation swap:
Donor A loves helping their fellow EAs and so offers a swap even though they don’t personally need it.
Donor B enthusiastically takes them up on the offer to save the taxes, donates $1k to AMF, and Donor A donates $1k to RP. Later, Donor A donates their remaining $1k to RP.
In effect: RP gains $2k and AMF gains $1k. Slightly positive for Donor A but big loss for Donor B.
This seems like a plausible scenario to me but there are other scenarios that are less extreme but still very detrimental to one side and possibly even harder to spot.
So am I overlooking something that alleviates this worry or do donors have to know (commit) and be transparent about where they will donate if no swap happens, in order for the other party to know whether they should take them up on the offer?
The only thing we put in to help with preferences, is to allow donors to “reject the swap”—e.g. if they don’t like the charity they are instructed to donate to. But that is clearly insufficient. They could be given more information before they choose to reject.
I’m trying to think of a way of asking for easily accessible information from donors to avoid these problems. This info would be shared with the match. Would “If you don’t get matched where would your donation go?”
with the options
Same charity, and I’ll get tax benefits
Same charity, and I won’t get tax benefits
Charity X and I’ll get tax benefits
Would that be sufficient to avoid the worst of the problems?
Hi! Thank you! That sounds good (“Charity X” would be a free text field?), but I don’t know whether there are other problems it doesn’t address. To guard against that, an FAQ entry explaining the problem would be best. Generally that’ll be needed because this is probably unintuitive for many people (like me), so even if they have the information about the swap counterfactual, they may not be able to use it optimally without an explanation.
Thank you for creating this! I want to understand some possible risks to my value system better. So here’s one scenario that I’ve been thinking about.
I realize that it’s a trust system but if Donor A trusts Donor B on something that is not clear enough to Donor A to be able to ask and that’s so unremarkable to Donor B that they see no reason to tell Donor A about it any more than their espresso preferences, then no one is really at fault if they miscommunicate.
Say Donor A:
is neutral between Rethink Priorities (RP) and the Against Malaria Foundation (AMF) (but Donor B doesn’t know this),
can get tax exemption for a donation to RP but not AMF, and
wants to donate $2k.
And Donor B:
values a dollar to RP more than 100 times as highly as one to AMF (but Donor A doesn’t know this),
can get tax exemption for a donation to AMF but not RP, and
wants to donate $1k.
Without donation swap:
Donor A is perfectly happy, and donates $2k to RP because of the tax exemption as tie breaker (but even if they split the donation 50:50 or donate with 50% probability, this case is still problematic).
Donor B is a bit sad but donates, say, $850 to RP, which comes down to the same cost to them due to the lacking tax break.
In effect: RP gains $2,850 and AMF gains $0. Both donors are reasonably happy with this result, the only wrinkle being the taxes.
But with donation swap:
Donor A loves helping their fellow EAs and so offers a swap even though they don’t personally need it.
Donor B enthusiastically takes them up on the offer to save the taxes, donates $1k to AMF, and Donor A donates $1k to RP. Later, Donor A donates their remaining $1k to RP.
In effect: RP gains $2k and AMF gains $1k. Slightly positive for Donor A but big loss for Donor B.
This seems like a plausible scenario to me but there are other scenarios that are less extreme but still very detrimental to one side and possibly even harder to spot.
So am I overlooking something that alleviates this worry or do donors have to know (commit) and be transparent about where they will donate if no swap happens, in order for the other party to know whether they should take them up on the offer?
That is a really good point Denis!
The only thing we put in to help with preferences, is to allow donors to “reject the swap”—e.g. if they don’t like the charity they are instructed to donate to. But that is clearly insufficient. They could be given more information before they choose to reject.
I’m trying to think of a way of asking for easily accessible information from donors to avoid these problems. This info would be shared with the match. Would “If you don’t get matched where would your donation go?”
with the options
Same charity, and I’ll get tax benefits
Same charity, and I won’t get tax benefits
Charity X and I’ll get tax benefits
Would that be sufficient to avoid the worst of the problems?
Hi! Thank you! That sounds good (“Charity X” would be a free text field?), but I don’t know whether there are other problems it doesn’t address. To guard against that, an FAQ entry explaining the problem would be best. Generally that’ll be needed because this is probably unintuitive for many people (like me), so even if they have the information about the swap counterfactual, they may not be able to use it optimally without an explanation.