As you explain, the key tradeoff is organizational stability vs. donor flexibility to chase high-impact opportunities. There are a couple different ways to strike the right balance. For example, organizations can try to secure long-term commitments sufficient to cover a set percentage of their projected budget but no more, e.g., 100% one year out; 50% two years out; 25% three years out [disclaimer: these numbers are not considered].
Another possibility is for donors to commit to donating a certain amount in the future but not to where. For example, imagine EA organizations x, y, and z are funded in significant part by donors a, b, and c. The uncertainty for each organization comes from both (i) how much a, b, and c will donate in the future (e.g., for how long do they plan to earn to give?), and (ii) to which organization (x, y, or z) will they donate. The option value for the donors comes primarily* from (ii): the flexibility to donate more to x, y, or z depending on how good they look relative to the others. And I suspect much (if not most) of the uncertainty for x, y, and z comes from (i): not knowing how much “EA money” there will be in the future. If that’s the case, we can get most of the good with little of the bad via general commitments to donate, without naming the beneficiary. One way to accomplish this would be an EA fund.
I say “primarily” because there is option value in being able to switch from earning to give to direct work, for example.
This is an empirical question, but at the moment my intuition goes the other way: that the fraction of the benefits of committing coming from commitments to specific organisations is larger than the corresponding fraction of costs. I have two main reasons for thinking this:
The first reason for this is that commitments from existing donors won’t shed that much light on the total amount of “EA money” that will be available: in the first place because personal financial uncertainty means they shouldn’t make firm commitments about all of their giving; in the second place because the amount of EA money that is available will depend in significant part on the rate of influx of new donors, and a lot of the uncertainty will relate to that.
The second reason is that I think organisational decision-making is particularly helped by pushing uncertainty towards zero. The effective difference between a 59% and a 79% chance of getting enough funding may be rather smaller than the difference between a 79% and 99% chance. Organisations may be reluctant to take on new staff if that gives a realistic chance of having to let existing staff go. So the benefits of having organisational certainty rather than just sector-certainty are large. (If we believe this is false, we should perhaps think that orgs should stop holding significant reserves.) I’m not expert in this, would be interested to hear thoughts from people more directly involved in financial planning for EA orgs.
As you explain, the key tradeoff is organizational stability vs. donor flexibility to chase high-impact opportunities. There are a couple different ways to strike the right balance. For example, organizations can try to secure long-term commitments sufficient to cover a set percentage of their projected budget but no more, e.g., 100% one year out; 50% two years out; 25% three years out [disclaimer: these numbers are not considered].
Another possibility is for donors to commit to donating a certain amount in the future but not to where. For example, imagine EA organizations x, y, and z are funded in significant part by donors a, b, and c. The uncertainty for each organization comes from both (i) how much a, b, and c will donate in the future (e.g., for how long do they plan to earn to give?), and (ii) to which organization (x, y, or z) will they donate. The option value for the donors comes primarily* from (ii): the flexibility to donate more to x, y, or z depending on how good they look relative to the others. And I suspect much (if not most) of the uncertainty for x, y, and z comes from (i): not knowing how much “EA money” there will be in the future. If that’s the case, we can get most of the good with little of the bad via general commitments to donate, without naming the beneficiary. One way to accomplish this would be an EA fund.
I say “primarily” because there is option value in being able to switch from earning to give to direct work, for example.
This is an empirical question, but at the moment my intuition goes the other way: that the fraction of the benefits of committing coming from commitments to specific organisations is larger than the corresponding fraction of costs. I have two main reasons for thinking this:
The first reason for this is that commitments from existing donors won’t shed that much light on the total amount of “EA money” that will be available: in the first place because personal financial uncertainty means they shouldn’t make firm commitments about all of their giving; in the second place because the amount of EA money that is available will depend in significant part on the rate of influx of new donors, and a lot of the uncertainty will relate to that.
The second reason is that I think organisational decision-making is particularly helped by pushing uncertainty towards zero. The effective difference between a 59% and a 79% chance of getting enough funding may be rather smaller than the difference between a 79% and 99% chance. Organisations may be reluctant to take on new staff if that gives a realistic chance of having to let existing staff go. So the benefits of having organisational certainty rather than just sector-certainty are large. (If we believe this is false, we should perhaps think that orgs should stop holding significant reserves.) I’m not expert in this, would be interested to hear thoughts from people more directly involved in financial planning for EA orgs.