But I feel strongly that it is not tenable to have a general community norm against something your most influential actors are doing without pushback, and checking the comments on the linked post Iâm not seeing that pushback.
I hear this as âyou canât complain about FarmKind, because you didnât complain about OpenPhilâ. But:
Jeff didnât complain about the GiveWell match at the time it was offered, because he didnât notice it. I donât think we can draw too much adverse inference from any specific person not commenting on any specific situation.
A big part of my motivation to leave a comment on the FarmKind post was anticipating that they might not have heard about common objections to donation matching, whereas I think Claire Zabel probably has heard of them.
Similarly, one might have differing expectations about how much one vs. the other organisation would be interested in your feedback, or likely to change path based on feedback (I donât really have a considered view on how FarmKind vs. OpenPhil score here)
idk, sometimes I feel like leaving comments and sometimes I donât?
I think itâs better to focus on the actual question of whether matches are good or bad, or what the essential features are for a match to be honest or not. Based on that question, we can decide âit was a mistake not to push back more on OpenPhilâ or âwhat OpenPhil did was fineâ if we think thatâs still worth adjudicating.
Iâm sorry you hear it that way, but thatâs not what it says; Iâm making an empirical claim about how norms work /â donât work. If you think the situation I describe is tenable, feel free to disagree.
But if we agree it is not tenable, then we need a (much?) narrower community norm than âno donation matchingâ, such as âno donation matching without communication around counterfactualsâ, or Open Phil /â EAF needs to take significantly more flak than I think they did.
I hoped pointing that out might help focus minds, since the discussion so far had focused on the weak players not the powerful ones.
But if we agree it is not tenable, then we need a (much?) narrower community norm than âno donation matchingâ, such as âno donation matching without communication around counterfactualsâ, or Open Phil /â EAF needs to take significantly more flak than I think they did.
While I think a norm of âno donation matchingâ is where we should be, I think the best weâre likely to get is âno donation matching without donors understanding the counterfactual impactâ. So while Iâve tried to argue for the former Iâve limited my criticism of campaigns to ones that donât meet the latter.
If youâre just saying âthis other case might inform whether and when we think donation matches are OKâ, then sure, that seems reasonable, although Iâm really more interested in people saying something like âthis other case is not bad, so we should draw the distinction in this wayâ or âthis other case is also bad, so we should make sure to include that tooâ, rather than just âthis other case existsâ.
If youâre saying âwe have to be consistent, going forward, with how we treated OpenPhil /â EA Funds in the pastâ, then surely no: at a minimum we also have the option of deciding it was a mistake to let them off so lightly, and then we can think about whether we need to do anything now to redress that omission. Maybe now is the time we start having the norm, having accepted we didnât have it before?
FWIW having read the post a couple of times I mostly donât understand why using a match seemed helpful to them. I think how bad it was depends partly on how EA Funds communicated to donors about the match: if they said âthis match will multiply your impact!â uncritically then I think thatâs misleading and bad, if they said âOpenPhil decided to structure our offramp funding in this particular way in order to push us to fundraise more, mostly you should not worry about it when donatingâ, that seems fine, I guess. I looked through my e-mails (though not very exhaustively) but didnât find communications from them that explicitly mentioned the match, so idk.
In support of this view: There was a lot going on with the Open Phil EAIF/âLTFF funding drawdown & short-term match announcement. Those funds losing a lot of their funding was of significant import to the relevant ecosystems, and likely of personal import to some commenters who expected to seek EAIF/âLTFF funding in the future. The short-term matching program was only a part of the larger news story (as it were). So commenter attention on that post was fragmented in a way that was not the case with FarmKindâs post.
In non-support of this view: Given AIMâs involvement with FarmKindâs launch, I would be really surprised if its founders were unaware of the critiques that had been levied against donation-matching programs over the past 10-15 years in EA. It is also unclear how much FarmKind could change course based on feedback (other than giving up and shutting down); matching is core to what it does in a way that isnât true of Open Phil.
Other than giving up and shutting down, they could have put offsetting front and center. I think it might be psychologically compelling to some who donât want to give up meat to be able to undo some of their contributions to the factory farming system. I actually became aware of their calculator from your quick take, as currently it is pretty hard to find.
An interesting idea! I think an offset-based strategy has some challenges, but Iâd be interested in seeing how it went. It tries to sell a different affective good than most traditional efforts (guilt reduction, rather than warm fuzzies), and a fundraising org probably has to choose between the two.
On the plus side, it probably appeals to a different set of donors. On the down side, youâd find it hard to collect more than the full-offset amount from any donor.
(Working out the ethics of quilting oneâs prospective donors is too complex for a comment written on commuter rail....)
Someone could not just eliminate their contribution to FF but be part of the solution if their contribution is a greater than one multiple of the offset. I think people might like a 1.5 to 2X offset potentially for the warm fuzzies.
It is also unclear how much FarmKind could change course based on feedback (other than giving up and shutting down)
There are lots of options helping animals (through raising money or otherwise) that donât involve this kind of competition around the impact of donations. Itâs common for startups to pivot if their first product doesnât work out.
Itâs common for startups to pivot if their first product doesnât work out.
I think FarmKind would be constrained by the need to keep faith with its donors, though. I agree about pivoting when there has been a material change in facts or circumstances. Thatâs an understood part of the deal when you donate to a charity. But I donât see any evidence of the product on which FarmKind raised its seed funding was not âwork[ing] outâ in practice. It had just launched its platform. The critiques you and Ben offered were not novel or previously unknown to FarmKind (e.g., their launch info cites Giving Multiplier as a sort of inspiration).
It is of course good to change your mind when you realize that you made an error. But from the seed donor perspective, changing paradigms so early would look an awful lot like a bait and switch in effect if not in intent. If an organization fundraises based on X and then promptly decides to do something significantly different for reasons within the organizationâs control without giving X a serious try, I think it should generally offer people who donated for X idea to have their donations regranted elsewhere. Whether its donors would chose that is unknown to me.
Independently, itâs not clear that a pivot would be practically feasible for FarmKind (or most other CE incubatees a few months after launch). It launched on a $133K seed grant with two staff members about three months ago (not including time spent in the CE program). Even assuming they could get 100% donor consent for repurposing funds, they would still be going almost back to square one. Itâs not clear to me that it could come up with a new idea, spend the time to develop that, relaunch, and show results to get funding for the rest of year 1 and beyond (which has been a big challenge for many CE-incubated orgs).
Given those constraints, I think it is fair to say that the extent to which FarmKind could realistically change course in response to criticism (other than winding down) remains unclear.
Yeah, in retrospect maybe it was kind of doomed to expect that I might influence FarmKindâs behaviour directly, and maybe the best I could hope for is influencing the audience to prefer other methods of promoting effective giving.
I hear this as âyou canât complain about FarmKind, because you didnât complain about OpenPhilâ. But:
Jeff didnât complain about the GiveWell match at the time it was offered, because he didnât notice it. I donât think we can draw too much adverse inference from any specific person not commenting on any specific situation.
A big part of my motivation to leave a comment on the FarmKind post was anticipating that they might not have heard about common objections to donation matching, whereas I think Claire Zabel probably has heard of them.
Similarly, one might have differing expectations about how much one vs. the other organisation would be interested in your feedback, or likely to change path based on feedback (I donât really have a considered view on how FarmKind vs. OpenPhil score here)
idk, sometimes I feel like leaving comments and sometimes I donât?
I think itâs better to focus on the actual question of whether matches are good or bad, or what the essential features are for a match to be honest or not. Based on that question, we can decide âit was a mistake not to push back more on OpenPhilâ or âwhat OpenPhil did was fineâ if we think thatâs still worth adjudicating.
Iâm sorry you hear it that way, but thatâs not what it says; Iâm making an empirical claim about how norms work /â donât work. If you think the situation I describe is tenable, feel free to disagree.
But if we agree it is not tenable, then we need a (much?) narrower community norm than âno donation matchingâ, such as âno donation matching without communication around counterfactualsâ, or Open Phil /â EAF needs to take significantly more flak than I think they did.
I hoped pointing that out might help focus minds, since the discussion so far had focused on the weak players not the powerful ones.
While I think a norm of âno donation matchingâ is where we should be, I think the best weâre likely to get is âno donation matching without donors understanding the counterfactual impactâ. So while Iâve tried to argue for the former Iâve limited my criticism of campaigns to ones that donât meet the latter.
If youâre just saying âthis other case might inform whether and when we think donation matches are OKâ, then sure, that seems reasonable, although Iâm really more interested in people saying something like âthis other case is not bad, so we should draw the distinction in this wayâ or âthis other case is also bad, so we should make sure to include that tooâ, rather than just âthis other case existsâ.
If youâre saying âwe have to be consistent, going forward, with how we treated OpenPhil /â EA Funds in the pastâ, then surely no: at a minimum we also have the option of deciding it was a mistake to let them off so lightly, and then we can think about whether we need to do anything now to redress that omission. Maybe now is the time we start having the norm, having accepted we didnât have it before?
FWIW having read the post a couple of times I mostly donât understand why using a match seemed helpful to them. I think how bad it was depends partly on how EA Funds communicated to donors about the match: if they said âthis match will multiply your impact!â uncritically then I think thatâs misleading and bad, if they said âOpenPhil decided to structure our offramp funding in this particular way in order to push us to fundraise more, mostly you should not worry about it when donatingâ, that seems fine, I guess. I looked through my e-mails (though not very exhaustively) but didnât find communications from them that explicitly mentioned the match, so idk.
In support of this view: There was a lot going on with the Open Phil EAIF/âLTFF funding drawdown & short-term match announcement. Those funds losing a lot of their funding was of significant import to the relevant ecosystems, and likely of personal import to some commenters who expected to seek EAIF/âLTFF funding in the future. The short-term matching program was only a part of the larger news story (as it were). So commenter attention on that post was fragmented in a way that was not the case with FarmKindâs post.
In non-support of this view: Given AIMâs involvement with FarmKindâs launch, I would be really surprised if its founders were unaware of the critiques that had been levied against donation-matching programs over the past 10-15 years in EA. It is also unclear how much FarmKind could change course based on feedback (other than giving up and shutting down); matching is core to what it does in a way that isnât true of Open Phil.
Other than giving up and shutting down, they could have put offsetting front and center. I think it might be psychologically compelling to some who donât want to give up meat to be able to undo some of their contributions to the factory farming system. I actually became aware of their calculator from your quick take, as currently it is pretty hard to find.
An interesting idea! I think an offset-based strategy has some challenges, but Iâd be interested in seeing how it went. It tries to sell a different affective good than most traditional efforts (guilt reduction, rather than warm fuzzies), and a fundraising org probably has to choose between the two.
On the plus side, it probably appeals to a different set of donors. On the down side, youâd find it hard to collect more than the full-offset amount from any donor.
(Working out the ethics of quilting oneâs prospective donors is too complex for a comment written on commuter rail....)
Someone could not just eliminate their contribution to FF but be part of the solution if their contribution is a greater than one multiple of the offset. I think people might like a 1.5 to 2X offset potentially for the warm fuzzies.
There are lots of options helping animals (through raising money or otherwise) that donât involve this kind of competition around the impact of donations. Itâs common for startups to pivot if their first product doesnât work out.
I think FarmKind would be constrained by the need to keep faith with its donors, though. I agree about pivoting when there has been a material change in facts or circumstances. Thatâs an understood part of the deal when you donate to a charity. But I donât see any evidence of the product on which FarmKind raised its seed funding was not âwork[ing] outâ in practice. It had just launched its platform. The critiques you and Ben offered were not novel or previously unknown to FarmKind (e.g., their launch info cites Giving Multiplier as a sort of inspiration).
It is of course good to change your mind when you realize that you made an error. But from the seed donor perspective, changing paradigms so early would look an awful lot like a bait and switch in effect if not in intent. If an organization fundraises based on X and then promptly decides to do something significantly different for reasons within the organizationâs control without giving X a serious try, I think it should generally offer people who donated for X idea to have their donations regranted elsewhere. Whether its donors would chose that is unknown to me.
Independently, itâs not clear that a pivot would be practically feasible for FarmKind (or most other CE incubatees a few months after launch). It launched on a $133K seed grant with two staff members about three months ago (not including time spent in the CE program). Even assuming they could get 100% donor consent for repurposing funds, they would still be going almost back to square one. Itâs not clear to me that it could come up with a new idea, spend the time to develop that, relaunch, and show results to get funding for the rest of year 1 and beyond (which has been a big challenge for many CE-incubated orgs).
Given those constraints, I think it is fair to say that the extent to which FarmKind could realistically change course in response to criticism (other than winding down) remains unclear.
Yeah, in retrospect maybe it was kind of doomed to expect that I might influence FarmKindâs behaviour directly, and maybe the best I could hope for is influencing the audience to prefer other methods of promoting effective giving.