I’m not entirely sure that I believe the premise of this game. Essentially, the claim is that 20 of SingInst’s regular donors have extra money lying around that they are willing to donate to SingInst iff someone else donates the same amount. What do the regular donors intend to do with the money otherwise? Have they signed a binding agreement to all get together and blow the money on a giant party? Otherwise, why would they not just decide to donate it to SingInst at the end of the matching period anyway?
Five: US tax law prohibits public charities from getting too much support from big donors.
Under US tax law, a 501(c)(3) public charity must maintain a certain percentage of “public support”. As with most tax rules, this one is complicated. If, over a four-year period, any one individual donates more than 2% of the organization’s total support, anything over 2% does not count as “public support”. If a single donor supported a charity, its public support percentage would be only 2%. If two donors supported a charity, its public support percentage would be at most 4%. Public charities must maintain a public support percentage of at least 10% and preferably 33.3%. Small donations—donations of less than 2% of our total support over a four-year period—count entirely as public support. Small donations permit us to accept more donations from our major supporters without sending our percentage of public support into the critical zone. Currently, the Singularity Institute is running short on public support—so please don’t think that small donations don’t matter!
Here’s my totally non-binding plan for my $1100 extra dollars that really were just lying around, budgeted but projected to not be spent: If we meet the full challenge, donate $1100 to SingInst and have Microsoft match it as well. If we meet only e.g. 80%, donate 80% of $1100 and have Microsoft match it, and spend the rest on a party I wouldn’t have had otherwise and link y’all to tasteful pictures. That’s a x3 multiplier on ~1% of the $125,000.
Before your post, bentarm, my plan was somewhat different but I estimate it gave at least a 2.9x multiplier.
Status affiliation, feeling even better about donating, creating an “event” which can be linked to and discussed, providing encouragement to others, and likely other reasons I’m not thinking of right now.
I’m betting Peter Thiel also used it to judge popular support for the idea when he ran a $400,000 challenge in 2007. IIRC, they didn’t even make it halfway by the deadline, which likely reduced his subsequent donations.
Essentially, the claim is that 20 of SingInst’s regular donors have extra money lying around that they are willing to donate to SingInst iff someone else donates the same amount.
Bear with my broad strokes here.
Let the utility u of donating x dollars to SingInst be diminishing: u = x^(1/3).
Assume that hedonistic spending can be spread over enough options to make diminishing returns negligible: u = x
The donor then donates their first dollar to SingInst, whereupon hedonistic spending provides them more utils.
The utility u of providing x dollars of a dollar-for-dollar service for SingInst donors is: u = (2x)^(1/3), and the donor gets up to ~1.415 dollars (to gain 2.83 dollars worth of donating utility!) before hedonistic spending is a better option. So this answers the first claim: they are willing to donate extra money iff someone else donates, and they intend to spend it on some other form of utility otherwise. The nature of dollar-for-dollar matching increases the utility they gain from their dollar.
As for the second part… there’s no need to commit to a binding agreement to not donate the leftover to SingInst at the end. Say the donor donated a dollar by themselves, then offered up to 41 cents matching contribution. Other donors take advantage of this offer by 20 cents—the donor still has another 21 cents of positive utility before hedonistic spending takes over, but the matching period ends at this point in time. (Why end the matching period? I’ll get there in a second). At this point, those 21 cents will generate .21 utils if spent hedonistically, but only .06 utils if donated without matching.
So the nature of diminishing utility will produce this behaviour. It’s not so suspicious a premise, in the end.
Now, why end the matching period? A utility-maximising donor ought to offer their 41 cents for as long as it takes—the utility of those 41 cents is always going to be higher than any 41-cent section of hedonistic spending (which is where the money is coming from). The answer lies partly in human psychology. A limited-time offer prompts action in a way that open-ended, time-independent offers do not. A progress bar prompts action too. It’s also partly from status: successfully completing campaigns like this raises the prestige of SingInst. Meeting the target before the specified end date raises the prestige of both SingInst and its donors.
These factors are convincing enough reasons to time-limit the matching offer, mostly because they solve the problem of the pledge not being fulfilled. Odd quirk of our brains to blame here.
The rational donor wouldn’t offer to match x for y time; they would simply offer to match x. So the donor would simply offer to match the first dollar and forty-one cents donated. There’s no risk of accidentally being required to donate more than you’d want to.
I thought humans did that a lot on financial markets (make an offer that potentially incurs an infinitely large obligation), in particular when they sell (“write”) certain option contracts?
I’m not entirely sure that I believe the premise of this game. Essentially, the claim is that 20 of SingInst’s regular donors have extra money lying around that they are willing to donate to SingInst iff someone else donates the same amount. What do the regular donors intend to do with the money otherwise? Have they signed a binding agreement to all get together and blow the money on a giant party? Otherwise, why would they not just decide to donate it to SingInst at the end of the matching period anyway?
This seems relevant:
Yes
Here’s my totally non-binding plan for my $1100 extra dollars that really were just lying around, budgeted but projected to not be spent: If we meet the full challenge, donate $1100 to SingInst and have Microsoft match it as well. If we meet only e.g. 80%, donate 80% of $1100 and have Microsoft match it, and spend the rest on a party I wouldn’t have had otherwise and link y’all to tasteful pictures. That’s a x3 multiplier on ~1% of the $125,000.
Before your post, bentarm, my plan was somewhat different but I estimate it gave at least a 2.9x multiplier.
I like that—it makes the match slightly better than 2:1.
Status affiliation, feeling even better about donating, creating an “event” which can be linked to and discussed, providing encouragement to others, and likely other reasons I’m not thinking of right now.
I’m betting Peter Thiel also used it to judge popular support for the idea when he ran a $400,000 challenge in 2007. IIRC, they didn’t even make it halfway by the deadline, which likely reduced his subsequent donations.
They made more like $270K IIRC,
I sent an internet email to User:Kevin describing the terms of my matching donor pledge, and then signed the internet email with my GPG key.
It’s better to sign the message before you send it, so that the recipient can view the signature.
I did.
Bear with my broad strokes here.
Let the utility u of donating x dollars to SingInst be diminishing: u = x^(1/3).
Assume that hedonistic spending can be spread over enough options to make diminishing returns negligible: u = x
The donor then donates their first dollar to SingInst, whereupon hedonistic spending provides them more utils.
The utility u of providing x dollars of a dollar-for-dollar service for SingInst donors is: u = (2x)^(1/3), and the donor gets up to ~1.415 dollars (to gain 2.83 dollars worth of donating utility!) before hedonistic spending is a better option. So this answers the first claim: they are willing to donate extra money iff someone else donates, and they intend to spend it on some other form of utility otherwise. The nature of dollar-for-dollar matching increases the utility they gain from their dollar.
As for the second part… there’s no need to commit to a binding agreement to not donate the leftover to SingInst at the end. Say the donor donated a dollar by themselves, then offered up to 41 cents matching contribution. Other donors take advantage of this offer by 20 cents—the donor still has another 21 cents of positive utility before hedonistic spending takes over, but the matching period ends at this point in time. (Why end the matching period? I’ll get there in a second). At this point, those 21 cents will generate .21 utils if spent hedonistically, but only .06 utils if donated without matching.
So the nature of diminishing utility will produce this behaviour. It’s not so suspicious a premise, in the end.
Now, why end the matching period? A utility-maximising donor ought to offer their 41 cents for as long as it takes—the utility of those 41 cents is always going to be higher than any 41-cent section of hedonistic spending (which is where the money is coming from). The answer lies partly in human psychology. A limited-time offer prompts action in a way that open-ended, time-independent offers do not. A progress bar prompts action too. It’s also partly from status: successfully completing campaigns like this raises the prestige of SingInst. Meeting the target before the specified end date raises the prestige of both SingInst and its donors.
These factors are convincing enough reasons to time-limit the matching offer, mostly because they solve the problem of the pledge not being fulfilled. Odd quirk of our brains to blame here.
don’t forget the simple explanation: It’s risky to offer to match infinite dollars.
The rational donor wouldn’t offer to match x for y time; they would simply offer to match x. So the donor would simply offer to match the first dollar and forty-one cents donated. There’s no risk of accidentally being required to donate more than you’d want to.
I thought humans did that a lot on financial markets (make an offer that potentially incurs an infinitely large obligation), in particular when they sell (“write”) certain option contracts?