Peter, thanks for writing out your reasons—it’s really useful.
I agree the one large donor point is worrying, but you shouldn’t set the donations given by the large donor to zero. There’s a reasonable chance GWWC recruits another large donor in the next couple of years and you should take account of that.
In fact, because donor size generally follows a power law, the top donor determines the mean, so if you ignore the top couple of large donors, your expectation value is always going to be too low.
The even bigger point is that I think it’s extremely conservative to set all future donations to zero. Yes, talk is cheap, so drop out rates could be much higher than we imagine, but as your example shows, to get to the point where the value of future pledges is zero requires extreme pessimism. Like on the order of 98% dropping out within a couple of years. There’s no evidence people are dropping out at anywhere near that rate.
Finally, you’re only considering downwards adjustments. What probability would you put on GWWC ‘optimistic’ scenario in the evaluation doc? I wouldn’t be super surprised if the leverage ratio turns out to be over 100. Let’s say that’s a 10% probability. Then the expected leverage ratio is already at least 10, even if I ignore all the other scenarios.
Then there’s also the impact that results from GWWC that isn’t just money moved (e.g. EA movement building), and the fact that funding now also lays the groundwork for long-term growth, which is where most of the expected value lies.
I’m personally not worried about “networks only run so deep” is a concern. In the early days it’s true GWWC spread in large part through the networks of the staff, but I don’t think this has been true for the last couple of years.
Finally, there’s an element of GWWC can’t win going on here. You say that the cost per member has got so cheap you’re skeptical that it’s real, so not convinced to donate. But I imagine if the cost per member had instead gone up, you’d be saying it’s diminishing returns, and also not convinced!
Peter, thanks for writing out your reasons—it’s really useful.
I agree the one large donor point is worrying, but you shouldn’t set the donations given by the large donor to zero. There’s a reasonable chance GWWC recruits another large donor in the next couple of years and you should take account of that.
In fact, because donor size generally follows a power law, the top donor determines the mean, so if you ignore the top couple of large donors, your expectation value is always going to be too low.
The even bigger point is that I think it’s extremely conservative to set all future donations to zero. Yes, talk is cheap, so drop out rates could be much higher than we imagine, but as your example shows, to get to the point where the value of future pledges is zero requires extreme pessimism. Like on the order of 98% dropping out within a couple of years. There’s no evidence people are dropping out at anywhere near that rate.
Finally, you’re only considering downwards adjustments. What probability would you put on GWWC ‘optimistic’ scenario in the evaluation doc? I wouldn’t be super surprised if the leverage ratio turns out to be over 100. Let’s say that’s a 10% probability. Then the expected leverage ratio is already at least 10, even if I ignore all the other scenarios. Then there’s also the impact that results from GWWC that isn’t just money moved (e.g. EA movement building), and the fact that funding now also lays the groundwork for long-term growth, which is where most of the expected value lies.
Two other points.
I’m personally not worried about “networks only run so deep” is a concern. In the early days it’s true GWWC spread in large part through the networks of the staff, but I don’t think this has been true for the last couple of years.
Finally, there’s an element of GWWC can’t win going on here. You say that the cost per member has got so cheap you’re skeptical that it’s real, so not convinced to donate. But I imagine if the cost per member had instead gone up, you’d be saying it’s diminishing returns, and also not convinced!