The way I would probably approach the calculation is this:
- Roughly how effective is the average charity compared to GiveWell’s top charities? 10%? 1%? - What is the mean annual revenue of these charities? (Mean, not median, remember the power law) - How many charities do I expect to be able to help with this? - How much more effective do I think I can make them?
I’m not sure how to find out the answers, but that would be a way to approach it. Since the questions are difficult, it might be good to have three calculations—optimistic, average, and pessimistic for plugging in the variables.
I agree this is a calculation that would produce a correct result, but it multiplies a lot of numbers that I’m highly unconfident in.
I assume there’s some clever approach such as:
Ignoring all the ~zero-impact charities.
Asking Givewell:
How many kind-of-ok charities exist?
Did Givewell talk to ~10/~100 such charities? If so, do these charities need help with their tech at all? How important is that problem?
How many charities would I be able to reach? (This specifically is a B2B-sales question, where I have better priors)
There are other blind spots I might have. I’ll give you some examples that I’ve already thought of, but the point is that I assume there are more:
If I help The Association of Pavement Sweepers (I just made them up) more efficient in raising funds and volunteers, will that be at the expense of better charities?
If I help charities advertise their ESG metrics
Am I promoting a bad metric that is “competing” with better ones, such as QALYs (or something else)?
Is this something that capitalizm would have handled anyway, since there is a strong financial incentive for it from the impact investing movement? Unclear, but if so, I prefer pushing the other problems
Could I nudge charities to do better impact evaluations? Maybe create a community where they help each other be more effective with their resources? (Seems hard)
The founder of Wave replied here that maybe I’m focusing on impact too much, and considering remote-working too little
Maybe there is a specific “class” of charities that if I only help them, it is already easily worth it. Specifically, I’m thinking about:
Advocating for governments to do different things. (Though, would I also help the “wrong side”?)
Meta organizations such as the world bank (though, I imagine they’re really tidy already.. they must be, right?)
Disaster relief: Monday helped with an earthquake and with covid vaccinations. I think this is something they do regularly, even though they don’t seem to consider it a focus right now.
I hope this paints the picture of why I see so much potential in getting outside advice on this. There are so many ways to approach the problem, and I totally might be thinking inside some kind of box without being aware of it. Of course, when it’s time to make a decision, I’ll make it with whatever uncertainty remains. So far I passed their technical interview, there are a few more steps in the process.
How useful is it to help a large number of ineffective charities?
How to even approach this calculation?
====
The way I would probably approach the calculation is this:
- Roughly how effective is the average charity compared to GiveWell’s top charities? 10%? 1%?
- What is the mean annual revenue of these charities? (Mean, not median, remember the power law)
- How many charities do I expect to be able to help with this?
- How much more effective do I think I can make them?
I’m not sure how to find out the answers, but that would be a way to approach it. Since the questions are difficult, it might be good to have three calculations—optimistic, average, and pessimistic for plugging in the variables.
I agree this is a calculation that would produce a correct result, but it multiplies a lot of numbers that I’m highly unconfident in.
I assume there’s some clever approach such as:
Ignoring all the ~zero-impact charities.
Asking Givewell:
How many kind-of-ok charities exist?
Did Givewell talk to ~10/~100 such charities? If so, do these charities need help with their tech at all? How important is that problem?
How many charities would I be able to reach? (This specifically is a B2B-sales question, where I have better priors)
There are other blind spots I might have.
I’ll give you some examples that I’ve already thought of, but the point is that I assume there are more:
If I help The Association of Pavement Sweepers (I just made them up) more efficient in raising funds and volunteers, will that be at the expense of better charities?
If I help charities advertise their ESG metrics
Am I promoting a bad metric that is “competing” with better ones, such as QALYs (or something else)?
Is this something that capitalizm would have handled anyway, since there is a strong financial incentive for it from the impact investing movement? Unclear, but if so, I prefer pushing the other problems
Could I nudge charities to do better impact evaluations? Maybe create a community where they help each other be more effective with their resources? (Seems hard)
The founder of Wave replied here that maybe I’m focusing on impact too much, and considering remote-working too little
Maybe there is a specific “class” of charities that if I only help them, it is already easily worth it. Specifically, I’m thinking about:
Advocating for governments to do different things. (Though, would I also help the “wrong side”?)
Meta organizations such as the world bank (though, I imagine they’re really tidy already.. they must be, right?)
Disaster relief: Monday helped with an earthquake and with covid vaccinations. I think this is something they do regularly, even though they don’t seem to consider it a focus right now.
I hope this paints the picture of why I see so much potential in getting outside advice on this. There are so many ways to approach the problem, and I totally might be thinking inside some kind of box without being aware of it. Of course, when it’s time to make a decision, I’ll make it with whatever uncertainty remains. So far I passed their technical interview, there are a few more steps in the process.