It seems to me that OP should just treat it as a grant of £15m (or whatever it was) to EVF. Presumably when you publish a grant report, that’s what it will say.
I think you make reasonable points that OP should consider it to be worth more than the resale difference, but there’s a big range between that and £15m. If there’s, eg, a 10% chance of this significant a break happening in the next 4 years, that would put the costs at 0.9 * resale difference + 0.1 * £15m. Do you think it’s reasonable to put a much higher probability on this?
Plus, EVF provides services to a lot of charities, so the main way I expect OpenPhil would decide they do not want money to remain with EVF would be if they decide that EVF itself is doing a bad job, or EVF otherwise significantly deviates from their values. Which is totally possible! But it seems reasonable to bet against.
I’m afraid I think this is fundamentally the wrong way to look at it. If you make a grant £X to allow another organisation to buy an asset, the fact that the donee ends up with a valuable asset is obviously an argument in favour of the grant, but you’ve still made a grant for £X. You can’t offset the value of the asset, because it isn’t your asset.
It may be that in making the grant for £X, you’re saving yourself making a series of future grants (in the case the cost of hiring venues for the conferences/retreats), and in that sense it may be financially advantageous, but you still have made a grant of £X, which needs to justified on its merits, just as you would have had to justify the counterfactual series of future grants. Fundamentally what seems to have happened here is that value of the events themselves has been taken for granted, even though they were being very significantly prefunded.
How do you put a meaningful probability on this? That 10% number is totally arbitrary.
I could equally say that the growing number of governance problems in EVF discussed in this forum make it 80% likely that it would be disavowed in the coming years. (This number is still meaningless)
This is a pretty uncertain question, and I’m not arguing that there’s a precise or rigorous way to answer it! But you can’t not put a probability on it—ultimately, everything we do is making a decision under uncertainty, and implicitly putting a probability on things. Eg, OpenPhil is implicitly putting a, say, less than 10% probability of EVF taking the money and running.
I’m making 10% explicit for illustrative purposes, and if I were making a grant here I’d be trying to think through the clearest evidence for or against (notably, I think that EVF has a decent track record, even if CEA as a whole doesn’t, has had so for a while, and provides a significant amount of infrastructure to a range of orgs doing good work, so I’d put the probability of such a large break that the money is basically burned as fairly low). And I think it’s generally unwise to make decisions that strongly rest on some specific and brittle assumptions re the right numbers. But if plausible feeling numbers don’t greatly shift the decision, this feels like useful data.
I think you make reasonable points that OP should consider it to be worth more than the resale difference, but there’s a big range between that and £15m. If there’s, eg, a 10% chance of this significant a break happening in the next 4 years, that would put the costs at 0.9 * resale difference + 0.1 * £15m. Do you think it’s reasonable to put a much higher probability on this?
Plus, EVF provides services to a lot of charities, so the main way I expect OpenPhil would decide they do not want money to remain with EVF would be if they decide that EVF itself is doing a bad job, or EVF otherwise significantly deviates from their values. Which is totally possible! But it seems reasonable to bet against.
I’m afraid I think this is fundamentally the wrong way to look at it. If you make a grant £X to allow another organisation to buy an asset, the fact that the donee ends up with a valuable asset is obviously an argument in favour of the grant, but you’ve still made a grant for £X. You can’t offset the value of the asset, because it isn’t your asset.
It may be that in making the grant for £X, you’re saving yourself making a series of future grants (in the case the cost of hiring venues for the conferences/retreats), and in that sense it may be financially advantageous, but you still have made a grant of £X, which needs to justified on its merits, just as you would have had to justify the counterfactual series of future grants. Fundamentally what seems to have happened here is that value of the events themselves has been taken for granted, even though they were being very significantly prefunded.
How do you put a meaningful probability on this? That 10% number is totally arbitrary.
I could equally say that the growing number of governance problems in EVF discussed in this forum make it 80% likely that it would be disavowed in the coming years. (This number is still meaningless)
This is a pretty uncertain question, and I’m not arguing that there’s a precise or rigorous way to answer it! But you can’t not put a probability on it—ultimately, everything we do is making a decision under uncertainty, and implicitly putting a probability on things. Eg, OpenPhil is implicitly putting a, say, less than 10% probability of EVF taking the money and running.
I’m making 10% explicit for illustrative purposes, and if I were making a grant here I’d be trying to think through the clearest evidence for or against (notably, I think that EVF has a decent track record, even if CEA as a whole doesn’t, has had so for a while, and provides a significant amount of infrastructure to a range of orgs doing good work, so I’d put the probability of such a large break that the money is basically burned as fairly low). And I think it’s generally unwise to make decisions that strongly rest on some specific and brittle assumptions re the right numbers. But if plausible feeling numbers don’t greatly shift the decision, this feels like useful data.