My understanding (based on talking to people involved in Wytham and knowing the economics of renting and buying large venues in a lot of detail) is that the sale of Wytham (edit: as done here, where the venue will either be sold at a very large discount or lie empty for a long period of time) does not actually make any economic sense for EV in terms of its mission to do as much good as possible. It is plausible that the initial purchase was a mistake, and that it makes sense to set plans in motion to sell the venue, but my understanding is that it will likely take many years for EV to sell during which the venue will be basically completely empty, or the venue will have to be sold at a pretty huge loss. This means at this point, it’s likely worth it to keep it running.
Also based on talking to some of the people close to these decisions, and trying to puzzle together how this decision was made, it seems very likely to me that the reason why Wytham is being sold is not based in a cost-effectiveness analysis, but the result of a PR-management strategy which seems antithetical to the principles of Effective Altruism to me.
EV (and Open Phil) are supposed to use its assets and funds to help the most people and cause the most good for the world, not to protect their own reputation. Making donations and major financial decisions primarily driven by reputation-concerns is the primary pathology of most of the world’s charity landscape, where vanity projects and complicated signaling games dominate where donations go, and going down this path seems to me a very worrying development for the future of EA.
My sense is that with this move, EV and Open Philanthropy have opened up a huge number of organizations within EA to attacks by any sufficiently large online mob, despite potentially producing enormous value, given that they have demonstrated they are willing to force the leadership of the EA community to give up projects with little concern for their cost-effectiveness if they do not align with the signaling aims of Open Phil and EV.
It is possible that maybe someone made a cost-effectiveness analysis here that turned out negative, and if so I would love to see it since it has large relevance to my work. But I would be extremely surprised that a positive cost-effectiveness analysis here would cause EV to reverse the sale of the property, and in conversations on this topic with people involved it seemed that curiosity and appetite for understanding the actual cost-effectiveness of this project was very low compared to the PR-implications of it.
(To be clear, I am not saying that we should fully blind ourselves to considerations of reputation and public relations. However, I think this kind of reputational optimization is perilous and if is one of the domains where naive consequentialist-type reasoning tends to most often go awry.
I think our reputational strategy should primarily be oriented around acting with integrity and honesty. And on that dimension the central tenet of how the EA community has presented itself is that we make decisions on the basis of what we think will help the most people, and are very much not making decisions on the basis of what will look good to other people, or will put us personally in the most powerful positions.
Imagine GiveWell releasing their recommended charities saying “well, there was one charity that easily defeated AMF in terms of the cost-effectiveness of its program activities, but it was dealing with sanitization issues which are really gross that nobody wants to donate to and we expected that if we recommended it this would overall reduce the donations going through GiveWell. We thought this effect was big enough to cause us to decide to not recommend this charity as our top charity”. I think this would be crazy and clearly violate the principles that GiveWell set out according to which it compiles its recommendations. While weaker, I think something similar is going on in how this decision seems to have been made)
To the extent this is an implied characterization of what happened here, I find it unlikely to be an apt one. It is unlikely that, e.g., EVF and/or OP made an optics-based decision on account of random posters on X. I also see no reason to conclude that the decisionmakers were affected by what their friends thought. Rather, I think the decisionmakers concluded that the expected state of the world was better if EV divested Wytham. For the same reason, I think the reference to “primary pathology of most of the world’s charity landscape, where vanity projects and complicated signaling games dominate where donations go” is overdone. Even if we assume that continued operation was more economically advantageous, this is a project on the periphery of what EA is, not an object-level issue. That reputational effects may have overruled a cost-effective analysis that disregarded those effects in this particular case does not update me on the probability that EA is at risk that vanity and signaling will “dominate,” or even play a major role, in EA funding decisions writ large.
As a practical matter, funders have a huge influence on organizational operations. (This isn’t wholly unique to the charitable world: customers and investors have somewhat analogous influences on for-profits.) Giving some weight to the views of future potential funders—who may be less likely to give to a movement that remains linked to the “castle”—does not strike me as fundamentally different than “letting” current funders’ views and preferences have as much weight as they do.
To the extent criticism is directed at Dustin, Cari, or Open Phil—the EA community does not own its donors’ money, and I see no basis for demanding that they continue to associate themselves with the “castle” if they do not wish to do so. Unless there’s another donor who is willing to incur the capital and operating costs of Wytham, I don’t see any potential room for criticizing EVF itself here. Of course, anyone who thinks Wytham should be reopened is welcome to fundraise for purchasing it or a similar building.
I also submit that wise stewardship and leadership of a social movement includes some consideration of morale amongst the rank and file. I’m guessing that some community builders whose funding has been cut due to financial circumstances may have been salty about “the castle” running while they were being asked to work with fewer resources. They probably were losing some effectiveness—and morale—through having to defend Wytham. The whole situation likely contributed to some people disengaging and/or not engaging. If these sorts of effects should not be considered, then I think there is much else in the meta world that could stand a reevaluation.
Finally, I’m more willing to weight optics on meta stuff than on object-level concerns; I think it would be much more epistemically dangerous to (e.g.) refuse to value farmed animals because of that isn’t seen as legitimate by certain others than it is to sell the “EA castle” because it is getting in the way of maintaining public respect and effectiveness. Moreover, in your GiveWell hypo, the project rejection on PR grounds would be corrosive to GiveWell’s function and value proposition (to be an unbiased, objective recommender in the areas it operates in) in a way that is less true in meta (where securing more money, talent, support, and other resources for object-level work is at least a key penultimate objective).
That reputational effects may have overruled a cost-effective analysis that disregarded those effects in this particular case does not update me on the probability that EA is at risk that vanity and signaling will “dominate,” or even play a major role, in EA funding decisions writ large.
I think Open Philanthropy staff will transparently tell you that they have recently substantially shifted towards considering optics and reputation as a core component of their grants.
I agree that Wytham itself is only one datapoint so should not update you much, though I think if you are curious about this, it wouldn’t be too hard to confirm that there is a broader shift going on (to be clear, I wouldn’t consider it vanity in that case, though broader signaling concerns seem quite substantial).
and I see no basis for demanding that they continue to associate themselves with the “castle” if they do not wish to do so
I agree that the donors should feel free to disassociate themselves from whatever they want, though in this case how the castle is being handled is a decision by EV, the most central EA organization. Also, of course, if a donor chooses to disassociate like that, it’s within the rights of EA community members to think less of them (they might still think positively on-net, but highlighting how someone’s grantmaking ignoring cost-effectiveness concerns in favor of personal reputation managemetn clearly is a valid criticism and should make you less excited about someone’s giving, and also concerned about the secondary effects of their giving)
Finally, I’m more willing to weight optics on meta stuff than on object-level concerns
This seems like a mistake.
I agree that it would seem more legitimate to do things for optics-reasons, but the detrimental effect on incentives and ability to think that come from optics-focused decision-making are just as real in meta work as for object-level work. The reason to not do things for optics-reasons is of course not that people will see you as more legitimate if you don’t, that’s just another optics-concern. The reason is that it affects the incentives on people to do good work, sets up an adversarial epistemic environment, and generally makes decision-making predictably worse. I don’t see why we should make a different tradeoff on those axis for meta work, where figuring out how to have a positive impact is usually substantially harder and messier than in more clear-cut global health and development cases.
I agree that the donors should feel free to disassociate themselves from whatever they want, though in this case how the castle is being handled is a decision by EV, the most central EA organization.
Another place where I have changed my mind over time is the grant we gave for the purchase of Wytham Abbey, an event space in Oxford.
[ . . . .]
Because this was a large asset, we agreed with Effective Ventures ahead of time that we would ask them to sell the Abbey if the event space, all things considered, turned out not to be sufficiently cost-effective. We recently made that request; funds from the sale will be distributed to other valuable projects they run.
Claire had previously noted that “(Proceeds from a sale would be used as general funding within EVF, and that funding would replace some of our and other funders’ future grants to EVF.)” So the plan of reallocating the funds locked up in Wytham to replace some of its general support to EVF if OP wanted out seems to have been the understanding from the get-go.
Based on that, it sounds like the sale was EVF’s decision insofar as it could have refused, damaged its relationship with its predominant funder, and found some other way to plug the resulting massive hole in its budget. In other words: not really. And EVF would still have needed to come up with a non-OP funding source for operating Wytham; surely it was not going to get OP’s funds to directly or indirectly do so after not honoring the request to divest it had agreed in advance OP could make.
Sure, I am not saying that EV should have gone back on some kind of promise.
But generally expect that when I trade with leadership in this ecosystem that they will be held to a standard of cost-effectiveness and not a standard of looking good on optics-grounds. And I think it continues to be good to judge EV and OP for making decisions on grounds that seem inconsistent with EA principles to me.
It shouldn’t cost them like an infinite amount of social capital, but I do think it makes the people a bit less suited to being in long-term EA leadership positions (i.e. if we had an election for EA leadership, the degree to which people optimize for optics instead of cost-effectiveness would hopefully be one of the central virtues on which to evaluate our leadership).
‘I think Open Philanthropy staff will transparently tell you that they have recently substantially shifted towards considering optics and reputation as a core component of their grants.’
the result of a PR-management strategy which seems antithetical to the principles of Effective Altruism to me
This view has been asserted many times before, but to my knowledge, it has never been explicitly defended. Why is it antithetical to the principles of effective altruism to be concerned with the reputational damage certain decisions can cause, when such damage can often severely impact one’s ability to do good?
In a comment explaining his decision to seek funding for the Wytham Abbey project, Owen Cotton-Barratt expresses a similar view. Owen writes that it is “better to let decisions be guided less by what we think looks good, and more by what we think is good.” But, to state the obvious, deciding based on what is good will sometimes require giving a lot of weight to how good the decision will look to others, because those perceptions are among the circumstances affecting the impact of our actions.
There may be a more sophisticated justification for the decision procedure to never, or almost never, allow PR concerns to influence one’s decision-making. Empirically, this doesn’t look true to me, though. For better or worse, we live in a world where PR “scandals” can harm people or movements involved in them to an extreme degree. I think we should take notice of this fact, and act accordingly.
Meta: I’m worried there could be some people-talking-past-each-other here. I never meant to claim that PR concerns shouldn’t influence one’s decision-making, but that they shouldn’t drive one’s decision-making. On this view you should certainly be willing to change direction on relatively unimportant issues for PR reasons, but should be somewhat resolute against doing so when it would change what was otherwise the central important thing you wanted to do.
Similarly, I agree with Habryka’s clarification in the thread above:
I am not saying that we should fully blind ourselves to considerations of reputation and public relations. However, I think this kind of reputational optimization is perilous and if is one of the domains where naive consequentialist-type reasoning tends to most often go awry.
In any case I don’t think that the case in point was primarily a PR-driven decision of the type being objected to. If it had been, I think that that would have been a little corrosive to the reputation of Open Philanthropy of trying to take everything seriously and go after the most fundamental things.
Alexander Berger’s reflections on Wytham are consistent with the thing that seems to me to be desirable. He regretted not weighing possible PR risks more heavily on a grant that they regarded as marginal—that seems fine. If they let such PR risks dominate their decision-making for a non-marginal grant, or mean that they wouldn’t even get to the point of assessing whether a grant was marginal, that would (of course depending on the details) seem more unwise.
But of course you’re right that it would be bad to ~never weigh PR considerations. I used to sometimes find it kind of distasteful (/veering towards ~immoral) to try explicitly to model how other people might think of me or of orgs I was a part of. But I think this was just a mistake.
I think assuming that this is purely based on optics is unwarranted. Like I argued at the time, talk of ‘optics’ is kind of insulting to the everyperson, carrying the implication that the irrational public will misunderstand the +EV of such a decision. Whereas I contend that there’s a perfectly rational Bayesian update that people should do towards an organisation being poorly run or even corrupt when that org spends large sums of money on vanity projects which they justify with a vague claim about having done some CBA that they don’t want to share.
Meanwhile, there’s no guarantee EA will have fresh billionaires any time soon, so even if it takes a couple of years to sell, it might be worth it, given that it a) there are alternative far cheaper-to-run venues like Lightcone and CEEALAR, and b) just recouping the sticker price would fund multiple cash-strapped EA orgs for several years.
To be clear, what I am criticizing here is not operating the venue while the sale is going on, or setting some kind of target for the operators in terms of quality-adjusted-events or estimates of counterfactual events caused, that would allow them to continue operating the venue.
I totally agree that observing someone spending money on a “vanity project” would be evidence that they are poorly run or corrupt, but like, Wytham would not be a vanity project if it were to make economic sense for EV or the EA community at large to operate. So whether a project is a vanity project is dependent on a cost-effectiveness analysis (which I don’t think really has occurred in this case).
My understanding (based on talking to people involved in Wytham and knowing the economics of renting and buying large venues in a lot of detail) is that the sale of Wytham (edit: as done here, where the venue will either be sold at a very large discount or lie empty for a long period of time) does not actually make any economic sense for EV in terms of its mission to do as much good as possible. It is plausible that the initial purchase was a mistake, and that it makes sense to set plans in motion to sell the venue, but my understanding is that it will likely take many years for EV to sell during which the venue will be basically completely empty, or the venue will have to be sold at a pretty huge loss. This means at this point, it’s likely worth it to keep it running.
Also based on talking to some of the people close to these decisions, and trying to puzzle together how this decision was made, it seems very likely to me that the reason why Wytham is being sold is not based in a cost-effectiveness analysis, but the result of a PR-management strategy which seems antithetical to the principles of Effective Altruism to me.
EV (and Open Phil) are supposed to use its assets and funds to help the most people and cause the most good for the world, not to protect their own reputation. Making donations and major financial decisions primarily driven by reputation-concerns is the primary pathology of most of the world’s charity landscape, where vanity projects and complicated signaling games dominate where donations go, and going down this path seems to me a very worrying development for the future of EA.
My sense is that with this move, EV and Open Philanthropy have opened up a huge number of organizations within EA to attacks by any sufficiently large online mob, despite potentially producing enormous value, given that they have demonstrated they are willing to force the leadership of the EA community to give up projects with little concern for their cost-effectiveness if they do not align with the signaling aims of Open Phil and EV.
It is possible that maybe someone made a cost-effectiveness analysis here that turned out negative, and if so I would love to see it since it has large relevance to my work. But I would be extremely surprised that a positive cost-effectiveness analysis here would cause EV to reverse the sale of the property, and in conversations on this topic with people involved it seemed that curiosity and appetite for understanding the actual cost-effectiveness of this project was very low compared to the PR-implications of it.
(To be clear, I am not saying that we should fully blind ourselves to considerations of reputation and public relations. However, I think this kind of reputational optimization is perilous and if is one of the domains where naive consequentialist-type reasoning tends to most often go awry.
I think our reputational strategy should primarily be oriented around acting with integrity and honesty. And on that dimension the central tenet of how the EA community has presented itself is that we make decisions on the basis of what we think will help the most people, and are very much not making decisions on the basis of what will look good to other people, or will put us personally in the most powerful positions.
Imagine GiveWell releasing their recommended charities saying “well, there was one charity that easily defeated AMF in terms of the cost-effectiveness of its program activities, but it was dealing with sanitization issues which are really gross that nobody wants to donate to and we expected that if we recommended it this would overall reduce the donations going through GiveWell. We thought this effect was big enough to cause us to decide to not recommend this charity as our top charity”. I think this would be crazy and clearly violate the principles that GiveWell set out according to which it compiles its recommendations. While weaker, I think something similar is going on in how this decision seems to have been made)
To the extent this is an implied characterization of what happened here, I find it unlikely to be an apt one. It is unlikely that, e.g., EVF and/or OP made an optics-based decision on account of random posters on X. I also see no reason to conclude that the decisionmakers were affected by what their friends thought. Rather, I think the decisionmakers concluded that the expected state of the world was better if EV divested Wytham. For the same reason, I think the reference to “primary pathology of most of the world’s charity landscape, where vanity projects and complicated signaling games dominate where donations go” is overdone. Even if we assume that continued operation was more economically advantageous, this is a project on the periphery of what EA is, not an object-level issue. That reputational effects may have overruled a cost-effective analysis that disregarded those effects in this particular case does not update me on the probability that EA is at risk that vanity and signaling will “dominate,” or even play a major role, in EA funding decisions writ large.
As a practical matter, funders have a huge influence on organizational operations. (This isn’t wholly unique to the charitable world: customers and investors have somewhat analogous influences on for-profits.) Giving some weight to the views of future potential funders—who may be less likely to give to a movement that remains linked to the “castle”—does not strike me as fundamentally different than “letting” current funders’ views and preferences have as much weight as they do.
To the extent criticism is directed at Dustin, Cari, or Open Phil—the EA community does not own its donors’ money, and I see no basis for demanding that they continue to associate themselves with the “castle” if they do not wish to do so. Unless there’s another donor who is willing to incur the capital and operating costs of Wytham, I don’t see any potential room for criticizing EVF itself here. Of course, anyone who thinks Wytham should be reopened is welcome to fundraise for purchasing it or a similar building.
I also submit that wise stewardship and leadership of a social movement includes some consideration of morale amongst the rank and file. I’m guessing that some community builders whose funding has been cut due to financial circumstances may have been salty about “the castle” running while they were being asked to work with fewer resources. They probably were losing some effectiveness—and morale—through having to defend Wytham. The whole situation likely contributed to some people disengaging and/or not engaging. If these sorts of effects should not be considered, then I think there is much else in the meta world that could stand a reevaluation.
Finally, I’m more willing to weight optics on meta stuff than on object-level concerns; I think it would be much more epistemically dangerous to (e.g.) refuse to value farmed animals because of that isn’t seen as legitimate by certain others than it is to sell the “EA castle” because it is getting in the way of maintaining public respect and effectiveness. Moreover, in your GiveWell hypo, the project rejection on PR grounds would be corrosive to GiveWell’s function and value proposition (to be an unbiased, objective recommender in the areas it operates in) in a way that is less true in meta (where securing more money, talent, support, and other resources for object-level work is at least a key penultimate objective).
I think Open Philanthropy staff will transparently tell you that they have recently substantially shifted towards considering optics and reputation as a core component of their grants.
I agree that Wytham itself is only one datapoint so should not update you much, though I think if you are curious about this, it wouldn’t be too hard to confirm that there is a broader shift going on (to be clear, I wouldn’t consider it vanity in that case, though broader signaling concerns seem quite substantial).
I agree that the donors should feel free to disassociate themselves from whatever they want, though in this case how the castle is being handled is a decision by EV, the most central EA organization. Also, of course, if a donor chooses to disassociate like that, it’s within the rights of EA community members to think less of them (they might still think positively on-net, but highlighting how someone’s grantmaking ignoring cost-effectiveness concerns in favor of personal reputation managemetn clearly is a valid criticism and should make you less excited about someone’s giving, and also concerned about the secondary effects of their giving)
This seems like a mistake.
I agree that it would seem more legitimate to do things for optics-reasons, but the detrimental effect on incentives and ability to think that come from optics-focused decision-making are just as real in meta work as for object-level work. The reason to not do things for optics-reasons is of course not that people will see you as more legitimate if you don’t, that’s just another optics-concern. The reason is that it affects the incentives on people to do good work, sets up an adversarial epistemic environment, and generally makes decision-making predictably worse. I don’t see why we should make a different tradeoff on those axis for meta work, where figuring out how to have a positive impact is usually substantially harder and messier than in more clear-cut global health and development cases.
Alexander Berger wrote yesterday:
Claire had previously noted that “(Proceeds from a sale would be used as general funding within EVF, and that funding would replace some of our and other funders’ future grants to EVF.)” So the plan of reallocating the funds locked up in Wytham to replace some of its general support to EVF if OP wanted out seems to have been the understanding from the get-go.
Based on that, it sounds like the sale was EVF’s decision insofar as it could have refused, damaged its relationship with its predominant funder, and found some other way to plug the resulting massive hole in its budget. In other words: not really. And EVF would still have needed to come up with a non-OP funding source for operating Wytham; surely it was not going to get OP’s funds to directly or indirectly do so after not honoring the request to divest it had agreed in advance OP could make.
Sure, I am not saying that EV should have gone back on some kind of promise.
But generally expect that when I trade with leadership in this ecosystem that they will be held to a standard of cost-effectiveness and not a standard of looking good on optics-grounds. And I think it continues to be good to judge EV and OP for making decisions on grounds that seem inconsistent with EA principles to me.
It shouldn’t cost them like an infinite amount of social capital, but I do think it makes the people a bit less suited to being in long-term EA leadership positions (i.e. if we had an election for EA leadership, the degree to which people optimize for optics instead of cost-effectiveness would hopefully be one of the central virtues on which to evaluate our leadership).
‘I think Open Philanthropy staff will transparently tell you that they have recently substantially shifted towards considering optics and reputation as a core component of their grants.’
Why do you think this?
Because I have talked to Open Philanthropy staff who have told me this
This view has been asserted many times before, but to my knowledge, it has never been explicitly defended. Why is it antithetical to the principles of effective altruism to be concerned with the reputational damage certain decisions can cause, when such damage can often severely impact one’s ability to do good?
In a comment explaining his decision to seek funding for the Wytham Abbey project, Owen Cotton-Barratt expresses a similar view. Owen writes that it is “better to let decisions be guided less by what we think looks good, and more by what we think is good.” But, to state the obvious, deciding based on what is good will sometimes require giving a lot of weight to how good the decision will look to others, because those perceptions are among the circumstances affecting the impact of our actions.
There may be a more sophisticated justification for the decision procedure to never, or almost never, allow PR concerns to influence one’s decision-making. Empirically, this doesn’t look true to me, though. For better or worse, we live in a world where PR “scandals” can harm people or movements involved in them to an extreme degree. I think we should take notice of this fact, and act accordingly.
Meta: I’m worried there could be some people-talking-past-each-other here. I never meant to claim that PR concerns shouldn’t influence one’s decision-making, but that they shouldn’t drive one’s decision-making. On this view you should certainly be willing to change direction on relatively unimportant issues for PR reasons, but should be somewhat resolute against doing so when it would change what was otherwise the central important thing you wanted to do.
I defended a similar broad position to this in a post a couple of years back on the perils of optimizing in social contexts.
Similarly, I agree with Habryka’s clarification in the thread above:
In any case I don’t think that the case in point was primarily a PR-driven decision of the type being objected to. If it had been, I think that that would have been a little corrosive to the reputation of Open Philanthropy of trying to take everything seriously and go after the most fundamental things.
Alexander Berger’s reflections on Wytham are consistent with the thing that seems to me to be desirable. He regretted not weighing possible PR risks more heavily on a grant that they regarded as marginal—that seems fine. If they let such PR risks dominate their decision-making for a non-marginal grant, or mean that they wouldn’t even get to the point of assessing whether a grant was marginal, that would (of course depending on the details) seem more unwise.
But of course you’re right that it would be bad to ~never weigh PR considerations. I used to sometimes find it kind of distasteful (/veering towards ~immoral) to try explicitly to model how other people might think of me or of orgs I was a part of. But I think this was just a mistake.
I think assuming that this is purely based on optics is unwarranted. Like I argued at the time, talk of ‘optics’ is kind of insulting to the everyperson, carrying the implication that the irrational public will misunderstand the +EV of such a decision. Whereas I contend that there’s a perfectly rational Bayesian update that people should do towards an organisation being poorly run or even corrupt when that org spends large sums of money on vanity projects which they justify with a vague claim about having done some CBA that they don’t want to share.
Meanwhile, there’s no guarantee EA will have fresh billionaires any time soon, so even if it takes a couple of years to sell, it might be worth it, given that it a) there are alternative far cheaper-to-run venues like Lightcone and CEEALAR, and b) just recouping the sticker price would fund multiple cash-strapped EA orgs for several years.
To be clear, what I am criticizing here is not operating the venue while the sale is going on, or setting some kind of target for the operators in terms of quality-adjusted-events or estimates of counterfactual events caused, that would allow them to continue operating the venue.
I totally agree that observing someone spending money on a “vanity project” would be evidence that they are poorly run or corrupt, but like, Wytham would not be a vanity project if it were to make economic sense for EV or the EA community at large to operate. So whether a project is a vanity project is dependent on a cost-effectiveness analysis (which I don’t think really has occurred in this case).
I really did not get this from
and suggest editing it into the original comment.
Yeah, that’s fair. I’ll edit it in.