Still, I’d love to hear: Why is asking for more transparency behind CEA/EV’s expenses a bad idea? Lots of individuals donate to them, so I feel like they deserve to know.
My personal take is that funders owe explanations to their donors, but not necessarily to the public or to the broader EA community (though it’s nice!). In this case, since the grant wasn’t really funded by CEA, this seems totally fine (though I do agree that, if it was bought by CEA, and many EAs donated to CEA, then justifying this publicly is probably good).
If it was a large funder with private funding, like OpenPhil, it feels much less clear to me. My guess is that general transparency is pretty good, and being able to receive high-quality external feedback is high value, but I’m not convinced that high-quality external feedback happens very often (and think that, eg, this post and the surrounding comments don’t meet that bar). I find Holden’s thoughts on this fairly persuasive. And I think that needing to make all decisions externally legible with clear, long justifications, seems plausibly more effort than is worth. Though I’m pretty in favour of the brief public grants databases they have.
Hey Jeroen! I personally would love to see the argument for your idea that CEA should publicly defend expenses over $500k.
I can imagine that the disagree votes can be pretty encouraging, but consider the ‘overall karma’ was positive, I think people (including myself) would be interested in seeing your reasoning—even if they (initially) disagree.
When I read your initial comment, my thoughts were “Yes! Wait, actually no. Maybe? What would be the difficulties?”. So I’d love to see a post elaborating on the ideas you’ve already thought up.
Part of what the idea needs—and was impossible for Jeroen to say in the context of Nathan’s post due to the length limit—is a ramp-up plan and limiting principles to prevent it from becoming more burdensome than it is worth. So it’s not realistic to implement this in a single year. As far as limiting principles: once there has been an explanation for “we spent X on the community health team and here’s a explanation why,” we should not expect another public explanation unless that funding line significantly changes (or perhaps all significant line items come up for re-explanation at 5-10 year intervals). Or perhaps there should be a first-stage listing of expenses, with the time/resources spent to publish a detailed rationale for most of them only if enough stakeholders request that in a poll.
The “detailed expected value calculation” may be a bit much for certain expenses. I realize it is heretical to say so, but not all organization expenses are subject to rigorous cost/benefit analyses.
I was confused by the disagree votes on your proposal and would also like to hear more about the reasoning behind them—I’d even argue that 500k is too high a bar for spending transparency. I feel like the value of this kind of transparency would by far outweigh the costs in time and effort by setting a good example and “practicing what you preach” alone. As a major representative of EA and a major funding body, CEA should hold itself to the same standards of cost effectiveness and transparency that we would expect of other NGOs. Additionally, having to publish a writeup like thet would incentivize rigorous and clear thinking, and seeing the calculations and reasoning in action would make CEA’s general approach clearer. I think it’s reasonable to expect a report on spending choices and the reasoning behind them ~every 3-6 months if doing it on an item-by-item basis is too burdensome.
I would support transparency and explanations of the kind that Hashim already provided. I think the idea that things should be justified by explicit expected value calculations, although it sometimes seems like a core EA idea, is not actually a good one. We usually can’t predict and quantify the kind of outcomes we are trying to achieve, and such attempts are more misguiding than useful.
The instructions on that post were, in my opinion, confusing. They included “Agreevote if you think [the proposed topics] are well-framed” for inclusion in a polis poll. I think that makes interpretation of “disagreevote” in that thread very difficult. For myself, I thought your core idea was a good one but not well-framed (“defended” sounded too adversarial).
Initially this was going to be a post rather than a question, where I would argue all expenses by CEA/EV above $500k should be publicly defended and explained with a detailed expected value calculation. But after pitching something similar under Nathan’s post and getting 17 disagree votes I decided it was probably a bad idea: https://forum.effectivealtruism.org/posts/CnAhPPsMWAxBm7pii/what-specific-changes-should-we-as-a-community-make-to-the?commentId=oa5xFokTTTb3maPwB
Still, I’d love to hear: Why is asking for more transparency behind CEA/EV’s expenses a bad idea? Lots of individuals donate to them, so I feel like they deserve to know.
My personal take is that funders owe explanations to their donors, but not necessarily to the public or to the broader EA community (though it’s nice!). In this case, since the grant wasn’t really funded by CEA, this seems totally fine (though I do agree that, if it was bought by CEA, and many EAs donated to CEA, then justifying this publicly is probably good).
If it was a large funder with private funding, like OpenPhil, it feels much less clear to me. My guess is that general transparency is pretty good, and being able to receive high-quality external feedback is high value, but I’m not convinced that high-quality external feedback happens very often (and think that, eg, this post and the surrounding comments don’t meet that bar). I find Holden’s thoughts on this fairly persuasive. And I think that needing to make all decisions externally legible with clear, long justifications, seems plausibly more effort than is worth. Though I’m pretty in favour of the brief public grants databases they have.
Hey Jeroen! I personally would love to see the argument for your idea that CEA should publicly defend expenses over $500k.
I can imagine that the disagree votes can be pretty encouraging, but consider the ‘overall karma’ was positive, I think people (including myself) would be interested in seeing your reasoning—even if they (initially) disagree.
When I read your initial comment, my thoughts were “Yes! Wait, actually no. Maybe? What would be the difficulties?”. So I’d love to see a post elaborating on the ideas you’ve already thought up.
Part of what the idea needs—and was impossible for Jeroen to say in the context of Nathan’s post due to the length limit—is a ramp-up plan and limiting principles to prevent it from becoming more burdensome than it is worth. So it’s not realistic to implement this in a single year. As far as limiting principles: once there has been an explanation for “we spent X on the community health team and here’s a explanation why,” we should not expect another public explanation unless that funding line significantly changes (or perhaps all significant line items come up for re-explanation at 5-10 year intervals). Or perhaps there should be a first-stage listing of expenses, with the time/resources spent to publish a detailed rationale for most of them only if enough stakeholders request that in a poll.
The “detailed expected value calculation” may be a bit much for certain expenses. I realize it is heretical to say so, but not all organization expenses are subject to rigorous cost/benefit analyses.
I was confused by the disagree votes on your proposal and would also like to hear more about the reasoning behind them—I’d even argue that 500k is too high a bar for spending transparency. I feel like the value of this kind of transparency would by far outweigh the costs in time and effort by setting a good example and “practicing what you preach” alone. As a major representative of EA and a major funding body, CEA should hold itself to the same standards of cost effectiveness and transparency that we would expect of other NGOs. Additionally, having to publish a writeup like thet would incentivize rigorous and clear thinking, and seeing the calculations and reasoning in action would make CEA’s general approach clearer. I think it’s reasonable to expect a report on spending choices and the reasoning behind them ~every 3-6 months if doing it on an item-by-item basis is too burdensome.
I would support transparency and explanations of the kind that Hashim already provided. I think the idea that things should be justified by explicit expected value calculations, although it sometimes seems like a core EA idea, is not actually a good one. We usually can’t predict and quantify the kind of outcomes we are trying to achieve, and such attempts are more misguiding than useful.
The instructions on that post were, in my opinion, confusing. They included “Agreevote if you think [the proposed topics] are well-framed” for inclusion in a polis poll. I think that makes interpretation of “disagreevote” in that thread very difficult. For myself, I thought your core idea was a good one but not well-framed (“defended” sounded too adversarial).