Not why you got a downvote, seems like a very reasonable discussion (but I note your prefacing as “taking one for the team”, so perhaps you were expecting the downvote more than I was!).
I’ll give my thoughts on these questions. I’m employed by CEA, but not speaking for CEA here. I’m sure you’d get a mix of thoughts from people who work there, but I also guess that my views aren’t too atypical.
1) It’s unclear whether the amount of money being raised should be regarded as large or small, and seems to depend quite a lot on the reference frame. First, not-so-relevant classes: it’s very large by the standards of an individual (relevant for the “many hundreds of lives” comment), and very small by the standards of society as a whole.
More relevantly, it’s large by the standards of the budgets from last year, or the year before, as you point out. With expansion like that, we should at least be considering whether the value of marginal activities is changing significantly. However, also relevant is that it’s reasonably small by the standards of the effective altruism community (and this is true if you draw the border not just at CEA but including all the EA meta-charities*). Our metrics are a bit crude, but to a first approximation I think CEA has been growing in proportion to the community as a whole.
This last reference class seems one of the most relevant to me. The value produced by the continued growth of the movement is large. It’s fairly clear that the growth happens as a result of things that people do. This includes a lot of things that individuals do without working on it full-time or being part of any organisation. Nonetheless, the average value of these activities is very high (even compared to the very good opportunities given by AMF), and the total amount of them isn’t so high that the marginal value is obviously much lower (though it could be lower, and there is a question mark here). This gives an argument for scaling up meta-charity quite quickly. There are then questions about how much of that ought to be professionalised, and how much of the professional meta-charity ought to be at CEA rather than elsewhere. I won’t try to answer that (but I will note that I think all of your queries have already kicked in by the time we get to the conclusion that metacharity activities should be scaling up roughly in proportion with the community).
2) How confident am I that CEA produces more value than equivalent donations to AMF? This turns out to be a surprisingly tricky question to pin down, because there are a lot of subtly different versions to which I’d give different answers. I’ll try to specify and answer a couple of these, but these probabilities are from the top of my head and might change if you asked me another day. In all cases I’ll include a guess about the opportunity cost of staff time in “equivalent donations”.
2.1) How likely do I think it is that CEA as a whole produces more value than equivalent AMF donations?
80%, but this is a bit misleading. I’m closer to 95% that the bundle of activities is more valuable, but if CEA weren’t pursuing them I think it’s quite likely other individuals or groups would pick up and do more, and that it’s possible that they’d do a better job.
2.2) How likely do I think that marginal CEA activities produce more value than equivalent AMF donations?
Quite dependent on the marginal activity. Averaged over a large enough margin, perhaps 80% again (though the chance that the bundle of marginal activities beats AMF is lower now, maybe 90% -- I think the displacement effects are weaker). But I think that individual CEA activities are significantly higher variance than AMF, so many of them may have low chances (~5-50%?) of outdoing AMF, balanced by the possibility of doing a lot better.
--
*Of the CEA orgs, I think it’s right to consider GWWC, 80k, and EAO as essentially meta-charities. GPP has had a mix of activities, some of which have meta-charity goals, and some of which are aiming at other indirect levers to creating value. Taken as a bundle, I’m reasonably confident that the non-meta-charity activities significantly exceed the bar of AMF, but it seems quite possible that some individual activities don’t.
“However, also relevant is that it’s reasonably small by the standards of the effective altruism community (and this is true if you draw the border not just at CEA but including all the EA meta-charities*). Our metrics are a bit crude, but to a first approximation I think CEA has been growing in proportion to the community as a whole.”
This does seem like the most relevant metric to me, and I’m not sure I agree with you about ‘reasonably small’ here. As Sebastian notes, the total CEA budget for 2016 is approx. $1.8m. That raises two questions for me:
How much of this do we expect to raise from within the EA community, broadly defined? I would expect the answer to be ‘a lot’, maybe 85%, but interested if others have different impressions.
How much money will the EA community, similarly broadly defined, donate this year?
That would get you towards ‘(financial) resources going to CEA as a fraction of total EA resources’, which is a number I care on quite a lot actually.
The EA community, broadly defined, donates a huge amount of money. GiveWell moved more than $20m (excluding GoodVentures) in 2015, source and credits effective altruism as motivating a substantial part of this. Giving What We Can moved more than $3m. FLI committed grants worth about $7m. Leverhulme Foundation granted $15m for existential risk research. This is far from exhaustive, but we’re looking at something on the order of magnitude of $50m fairly easily.
Relative to this, CEA’s $1.8m does not seem nearly as large. I think one of the sources of intuitive surprise is just that the EA movement as a whole seems to be roughly doubling or a bit more in size every year which means that the heuristics we have for thinking about size become out of date very quickly. Relative to EA as a whole, CEA may be shrinking slightly since we have been growing a little slower than doubling.
Most of the projects have significant non-EA funding, but this is something we’re trying to grow (for example by recruiting for a development manager who could expand our non-EA donor base). 80k got a lot of funding through YCombinator and associated leads. GWWC gets a substantial amount from people with a strong interest in development and giving but less in effective altruism itself. GPP gets significant funding from grant sources that wouldn’t otherwise fund EA work. Even EAO got at least $50k from people who have not typically given to EA charities, which is surprising for an organisation focused so heavily on the EA community itself. I’ve gone over our numbers and think 80k may have gotten more than half its budget outside the EA community recently. GPP gets around 40%. This is pretty loose stuff though, because it’s so hard to define what counts as EA money and we don’t have good access to the counterfactuals. Ben also makes a really important point about the donors who move from giving to us to supporting other EA projects.
Although note that you’ve compared 2015 money moved with projected 2016 costs. I think it’s probably more appropriate to compare them in the same year, which if the growth is indeed a doubling every year means the CEA budget is about half the size proportionally.
Overall it looks like meta-charity is in the vicinity of 5% of money moved. My personal (extremely uncertain) guess is that 5-10% would be about right for a movement in equilibrium, but that during rapid growth it’s better to be higher.
I understand how you’re both getting to these numbers and a ratio of around 5%. I’m trying to reconcile them with the fact that for example, Denise and I expect to give around 50% of our donations to meta-charities this year. For everyone like us, there need to be 9 comparably sized donors giving 0% to make the maths line up. But I don’t actually offhand know of anyone planning to give less than 5% to meta except some small donors who don’t want to split their donations. I understood EA donations to be relatively top-heavy, so those small donors shouldn’t affect things that drastically. So that suggests my circles are skewed in some way, but I’m not sure how.
All I can think of is that the really big donors give less than 5% (though not 0) to meta-charities and they skew the entire pool. But that doesn’t quite seem right either, e.g. Good Ventures donated $15m to Givewell charities in 2014 and Givewell operations were $3m total in that year. Annoyingly I can’t seem to find how much GV donated to GW for their operations that year, but I assume it was not significant less, and probably more, than 5% * $15m = $0.75m.
In any case, if it is the case that the low overall percentage is entirely due to a few large donors contributing low percentages to meta-charity, that still seems worth bearing in mind when you are actually in the process of fund-raising from relatively smaller donors (who are presumably the targets of this post). AFAIK they are already committing way more than 5% to meta-charity.
There’s a big pool of silent donors who give to GiveWell recommended charities. They don’t participate actively in the community so you never hear from them.
So I agree that if you care about “actively participating EAs” then the percentage who give to meta-charities is likely higher.
It’s not clear which figure is most relevant. It makes sense that the people most involved in the community are in the best relative position to support the EA orgs because they know the most about them. It also makes sense that the more persuaded people will be more inclined to support meta-charities. Finally, large donors should give more to startups because they have more time to research (and can specialise in being meta-charity donors); small donors should stick to GiveWell recommended charities. Overall, the large active EA donors should give above the 5% baseline to pull the overall ratio up.
My understanding was that the majority of Givewell donors have essentially nothing to do with EA whatsoever (though this seems to be changing by the sounds of the board meeting Sebastian mentioned linked to above). Is the claim that there’s an intermediate group who, say, have heard of and would identify with EA and donated to GW-recommended charities but don’t actively engage? That doesn’t sound crazy.
I agree with all of the latter points, but again my impression was that donations are already very top heavy. Some estimates that might help pin this disagreement more precisely:
What percentage of donations by dollars do you think come from people donating more than $30k annually?
What percentage do you think those people give to meta-charity?
If your answers are, say, 30% and 20%, that already gets you above 5%, and of course that’s a lower bound at this point.
On the silent donors, I’m not sure. If you’re giving a lot of money to a GiveWell recommended charity or you’re a member of GWWC, then functionally you’re an EA (in my definition). But I agree many of them might not explicitly identify as EAs. I do think there’s a significant intermediate group, but I’m not sure how many.
Perhaps more relevant, even if they don’t identify as EA, where the silent donors give is influenced by EA. So I think there’s still a good case for including them in the money moved.
If we can persuade a big group of people to give to AMF but not metacharities, then it makes sense to do that, and then for the people who are interested in giving to either give to meta charity.
Perhaps more relevant, even if they don’t identify as EA, where the silent donors give is influenced by EA. So I think there’s still a good case for including them in the money moved.
Can you explain why you’re thinking that they’re influenced by EA? It seems at least equally plausible that they’re influenced by GiveWell, which is distinct from most EA meta-orgs, and operates using a different model. Are you thinking that there’s another influence on them, like 80k or GWWC?
They’re influenced by GiveWell, and GiveWell is part of EA.
Or even if you don’t think GiveWell is part of EA, they’re very similar to EA in their approach, and many of the staff are explicitly EAs or supporters of EA. I think GiveWell has also been influenced by other groups in EA, though it’s hard to tell.
I agree that GiveWell could be considered part of EA. Ultimately I see that as a merely semantic question. My and I think AGB’s point is that the donors who follow GiveWell aren’t self-identified members of the “EA movement”, and aren’t giving because of EA outreach specifically. It appears that organizations doing EA outreach specifically get much more than 5% of the money donated by members of the “EA movement” who were inspired to give by those organizations.
Ah, I was trying to look at the ratio of meta-charity work to money being moved at the same time. Since meta-charities fundraise often a year in advance, then as a flow of donations the proportion should be higher.
I do think that “You should measure and give to the most effective measured charities” is a much easier sell than “You should try to understand which of the things we can’t measure well could be even better than the ones we can”, so my prior is that these numbers aren’t surprising. I don’t feel I have a good understanding of the community breakdown; I expect I interact way disproportionately with people who are on board with the second statement. This means my error bars on the kind of estimate that you ask Ben for are large, and they don’t have too much effect on my all-things-considered belief.
From memory, 80k has raised about 40% of its total funding from outside the EA community. We’ve also brought more donors into the EA community, which I’m not including in this figure.
To add an additional clarification of the question:
2.3 How likely is it that marginal CEA activities produce more value than equivalent AMF donations in expectation?
Speaking only for EAO, it seems somewhat likely to me that EAO will ultimately not be able to produce as much value as AMF would have with marginal donations. However, I think we have a much larger potential upside than marginal donations to AMF so, I’m pretty confident donations to EAO are better in expectation.
I would be extremely surprised if donations to 80K and GWWC weren’t better than AMF in expectation.
Hi Denise,
Not why you got a downvote, seems like a very reasonable discussion (but I note your prefacing as “taking one for the team”, so perhaps you were expecting the downvote more than I was!).
I’ll give my thoughts on these questions. I’m employed by CEA, but not speaking for CEA here. I’m sure you’d get a mix of thoughts from people who work there, but I also guess that my views aren’t too atypical.
1) It’s unclear whether the amount of money being raised should be regarded as large or small, and seems to depend quite a lot on the reference frame. First, not-so-relevant classes: it’s very large by the standards of an individual (relevant for the “many hundreds of lives” comment), and very small by the standards of society as a whole.
More relevantly, it’s large by the standards of the budgets from last year, or the year before, as you point out. With expansion like that, we should at least be considering whether the value of marginal activities is changing significantly. However, also relevant is that it’s reasonably small by the standards of the effective altruism community (and this is true if you draw the border not just at CEA but including all the EA meta-charities*). Our metrics are a bit crude, but to a first approximation I think CEA has been growing in proportion to the community as a whole.
This last reference class seems one of the most relevant to me. The value produced by the continued growth of the movement is large. It’s fairly clear that the growth happens as a result of things that people do. This includes a lot of things that individuals do without working on it full-time or being part of any organisation. Nonetheless, the average value of these activities is very high (even compared to the very good opportunities given by AMF), and the total amount of them isn’t so high that the marginal value is obviously much lower (though it could be lower, and there is a question mark here). This gives an argument for scaling up meta-charity quite quickly. There are then questions about how much of that ought to be professionalised, and how much of the professional meta-charity ought to be at CEA rather than elsewhere. I won’t try to answer that (but I will note that I think all of your queries have already kicked in by the time we get to the conclusion that metacharity activities should be scaling up roughly in proportion with the community).
2) How confident am I that CEA produces more value than equivalent donations to AMF? This turns out to be a surprisingly tricky question to pin down, because there are a lot of subtly different versions to which I’d give different answers. I’ll try to specify and answer a couple of these, but these probabilities are from the top of my head and might change if you asked me another day. In all cases I’ll include a guess about the opportunity cost of staff time in “equivalent donations”.
80%, but this is a bit misleading. I’m closer to 95% that the bundle of activities is more valuable, but if CEA weren’t pursuing them I think it’s quite likely other individuals or groups would pick up and do more, and that it’s possible that they’d do a better job.
Quite dependent on the marginal activity. Averaged over a large enough margin, perhaps 80% again (though the chance that the bundle of marginal activities beats AMF is lower now, maybe 90% -- I think the displacement effects are weaker). But I think that individual CEA activities are significantly higher variance than AMF, so many of them may have low chances (~5-50%?) of outdoing AMF, balanced by the possibility of doing a lot better.
--
*Of the CEA orgs, I think it’s right to consider GWWC, 80k, and EAO as essentially meta-charities. GPP has had a mix of activities, some of which have meta-charity goals, and some of which are aiming at other indirect levers to creating value. Taken as a bundle, I’m reasonably confident that the non-meta-charity activities significantly exceed the bar of AMF, but it seems quite possible that some individual activities don’t.
“However, also relevant is that it’s reasonably small by the standards of the effective altruism community (and this is true if you draw the border not just at CEA but including all the EA meta-charities*). Our metrics are a bit crude, but to a first approximation I think CEA has been growing in proportion to the community as a whole.”
This does seem like the most relevant metric to me, and I’m not sure I agree with you about ‘reasonably small’ here. As Sebastian notes, the total CEA budget for 2016 is approx. $1.8m. That raises two questions for me:
How much of this do we expect to raise from within the EA community, broadly defined? I would expect the answer to be ‘a lot’, maybe 85%, but interested if others have different impressions.
How much money will the EA community, similarly broadly defined, donate this year?
That would get you towards ‘(financial) resources going to CEA as a fraction of total EA resources’, which is a number I care on quite a lot actually.
The EA community, broadly defined, donates a huge amount of money. GiveWell moved more than $20m (excluding GoodVentures) in 2015, source and credits effective altruism as motivating a substantial part of this. Giving What We Can moved more than $3m. FLI committed grants worth about $7m. Leverhulme Foundation granted $15m for existential risk research. This is far from exhaustive, but we’re looking at something on the order of magnitude of $50m fairly easily.
Relative to this, CEA’s $1.8m does not seem nearly as large. I think one of the sources of intuitive surprise is just that the EA movement as a whole seems to be roughly doubling or a bit more in size every year which means that the heuristics we have for thinking about size become out of date very quickly. Relative to EA as a whole, CEA may be shrinking slightly since we have been growing a little slower than doubling.
Most of the projects have significant non-EA funding, but this is something we’re trying to grow (for example by recruiting for a development manager who could expand our non-EA donor base). 80k got a lot of funding through YCombinator and associated leads. GWWC gets a substantial amount from people with a strong interest in development and giving but less in effective altruism itself. GPP gets significant funding from grant sources that wouldn’t otherwise fund EA work. Even EAO got at least $50k from people who have not typically given to EA charities, which is surprising for an organisation focused so heavily on the EA community itself. I’ve gone over our numbers and think 80k may have gotten more than half its budget outside the EA community recently. GPP gets around 40%. This is pretty loose stuff though, because it’s so hard to define what counts as EA money and we don’t have good access to the counterfactuals. Ben also makes a really important point about the donors who move from giving to us to supporting other EA projects.
Although note that you’ve compared 2015 money moved with projected 2016 costs. I think it’s probably more appropriate to compare them in the same year, which if the growth is indeed a doubling every year means the CEA budget is about half the size proportionally.
Overall it looks like meta-charity is in the vicinity of 5% of money moved. My personal (extremely uncertain) guess is that 5-10% would be about right for a movement in equilibrium, but that during rapid growth it’s better to be higher.
I understand how you’re both getting to these numbers and a ratio of around 5%. I’m trying to reconcile them with the fact that for example, Denise and I expect to give around 50% of our donations to meta-charities this year. For everyone like us, there need to be 9 comparably sized donors giving 0% to make the maths line up. But I don’t actually offhand know of anyone planning to give less than 5% to meta except some small donors who don’t want to split their donations. I understood EA donations to be relatively top-heavy, so those small donors shouldn’t affect things that drastically. So that suggests my circles are skewed in some way, but I’m not sure how.
All I can think of is that the really big donors give less than 5% (though not 0) to meta-charities and they skew the entire pool. But that doesn’t quite seem right either, e.g. Good Ventures donated $15m to Givewell charities in 2014 and Givewell operations were $3m total in that year. Annoyingly I can’t seem to find how much GV donated to GW for their operations that year, but I assume it was not significant less, and probably more, than 5% * $15m = $0.75m.
In any case, if it is the case that the low overall percentage is entirely due to a few large donors contributing low percentages to meta-charity, that still seems worth bearing in mind when you are actually in the process of fund-raising from relatively smaller donors (who are presumably the targets of this post). AFAIK they are already committing way more than 5% to meta-charity.
There’s a big pool of silent donors who give to GiveWell recommended charities. They don’t participate actively in the community so you never hear from them.
So I agree that if you care about “actively participating EAs” then the percentage who give to meta-charities is likely higher.
It’s not clear which figure is most relevant. It makes sense that the people most involved in the community are in the best relative position to support the EA orgs because they know the most about them. It also makes sense that the more persuaded people will be more inclined to support meta-charities. Finally, large donors should give more to startups because they have more time to research (and can specialise in being meta-charity donors); small donors should stick to GiveWell recommended charities. Overall, the large active EA donors should give above the 5% baseline to pull the overall ratio up.
My understanding was that the majority of Givewell donors have essentially nothing to do with EA whatsoever (though this seems to be changing by the sounds of the board meeting Sebastian mentioned linked to above). Is the claim that there’s an intermediate group who, say, have heard of and would identify with EA and donated to GW-recommended charities but don’t actively engage? That doesn’t sound crazy.
I agree with all of the latter points, but again my impression was that donations are already very top heavy. Some estimates that might help pin this disagreement more precisely:
What percentage of donations by dollars do you think come from people donating more than $30k annually?
What percentage do you think those people give to meta-charity?
If your answers are, say, 30% and 20%, that already gets you above 5%, and of course that’s a lower bound at this point.
On the silent donors, I’m not sure. If you’re giving a lot of money to a GiveWell recommended charity or you’re a member of GWWC, then functionally you’re an EA (in my definition). But I agree many of them might not explicitly identify as EAs. I do think there’s a significant intermediate group, but I’m not sure how many.
Perhaps more relevant, even if they don’t identify as EA, where the silent donors give is influenced by EA. So I think there’s still a good case for including them in the money moved. If we can persuade a big group of people to give to AMF but not metacharities, then it makes sense to do that, and then for the people who are interested in giving to either give to meta charity.
Can you explain why you’re thinking that they’re influenced by EA? It seems at least equally plausible that they’re influenced by GiveWell, which is distinct from most EA meta-orgs, and operates using a different model. Are you thinking that there’s another influence on them, like 80k or GWWC?
They’re influenced by GiveWell, and GiveWell is part of EA.
Or even if you don’t think GiveWell is part of EA, they’re very similar to EA in their approach, and many of the staff are explicitly EAs or supporters of EA. I think GiveWell has also been influenced by other groups in EA, though it’s hard to tell.
I agree that GiveWell could be considered part of EA. Ultimately I see that as a merely semantic question. My and I think AGB’s point is that the donors who follow GiveWell aren’t self-identified members of the “EA movement”, and aren’t giving because of EA outreach specifically. It appears that organizations doing EA outreach specifically get much more than 5% of the money donated by members of the “EA movement” who were inspired to give by those organizations.
Ah, I was trying to look at the ratio of meta-charity work to money being moved at the same time. Since meta-charities fundraise often a year in advance, then as a flow of donations the proportion should be higher.
I do think that “You should measure and give to the most effective measured charities” is a much easier sell than “You should try to understand which of the things we can’t measure well could be even better than the ones we can”, so my prior is that these numbers aren’t surprising. I don’t feel I have a good understanding of the community breakdown; I expect I interact way disproportionately with people who are on board with the second statement. This means my error bars on the kind of estimate that you ask Ben for are large, and they don’t have too much effect on my all-things-considered belief.
From memory, 80k has raised about 40% of its total funding from outside the EA community. We’ve also brought more donors into the EA community, which I’m not including in this figure.
To add an additional clarification of the question:
Speaking only for EAO, it seems somewhat likely to me that EAO will ultimately not be able to produce as much value as AMF would have with marginal donations. However, I think we have a much larger potential upside than marginal donations to AMF so, I’m pretty confident donations to EAO are better in expectation.
I would be extremely surprised if donations to 80K and GWWC weren’t better than AMF in expectation.