I’ll give an answer that considers the details of the Copenhagen Consensus Center (CCC) and geoengineering, rather than being primarily a priori. I’ve spent a day and a half digging around and have zero prior knowledge. I spent too much time reading Lomborg and CCC in retrospect, so I mention him disproportionately relative to other sources.
1. Lomborg and his CCC seem very cost-benefit focused in their analysis. A few others are too, but see point 4. Basically, it’s easy to compare climate interventions to other interventions, but hard to figure out how much damage warming and other climate change will cause, so you can’t really figure out the benefit part of the cost-benefit analysis.
2. Lomborg and his CCC has recieved a ton of criticism for systematically making errors that underestimate the effect of climate change, and never making errors that overestimate the effect of it. One detailed account of him making such an error that could not have been made in good faith, is given here. He also literally lies in his cost-benefit analysis (by more than 10x).
There’s a lot of articles about Lomborg e.g. taking a 700k salary and getting donations from Koch and Exxon, which showed up before I found the above examples of him lying about data. I reacted to the info on his salary/funding by saying, “this is indicative of him being sketchy, but instead of just changing my estimate of how likely he is to be sketch by however much (“updating”) and calling it a day, I’m going to take this as a cue to dig into things until I have a firm understanding of whether this guy is systematically lying or not”. Turns out he’s a liar (see previous paragraph).
To place SRM 1 [a plan which, by definition, reduces the amount of heat from the sun that stays in the earth’s atmosphere by 1 watt per square meter] in perspective, 1 W m-2 is about 0.3% of the [average over the earth] incoming solar radiation of 341 W m-2 (Kiehl and Trenberth, 1997; Trenberth et al., 2009).
Which corresponds to a 0.6 C average temperature change. This should put in perspective what a huge effect a 0.3% change in how much heat gets stays in earth’s atmosphere has on global temperatures. If all the glaciers melted, the earth’s temperature would rise 10 C, because glaciers are good at reflecting heat back away from the earth (back of envelope, me); if the earth became totally frozen, the temperature would drop by 55 C (see this page), which is why the thing this post is about, MCB, can realistically change global temperatures by a couple degrees C. Note that 1⁄2 of the global temperature change that already happened was because some glaciers and snow already melted and stopped reflecting heat away from the earth. Basically, the science behind MCB is pretty solid and I’d expect it to basically work.
AFAICT the IPCC estimates of how much warming there will be in the future seem to take into account the fact that the melting of the glaciers will further speed warming in itself, in addition to the warming you get from rising CO2 levels. Except they don’t ever explicitly say whether that was considered as a factor, so I can’t be totally sure they took that into account, even though it sounds like the most obvious thing to consider. (I skimmed this whole damn thing, and it wasn’t said either way!) I guess my next course of action could be to annoy an author about it, though I think I’ll be lazy and not.
4. We have little enough data on “how much economic damage has global warming done so far” that we can’t make decent extrapolations to “how much economic damage will global warming do later”. Like, you have papers saying that the economic damage from 3 C of warming could be 1%, 5-20%, 23%, or 35% of the GDP. When you have zip for data, you fall back on your politics.
5. The obvious game theory consideration of, “it’s better if someone other than you spends money on global warming”. The normal lefty position of, “our institutions aren’t set up to coordinate well on this sort of problem, and every action against climate change, until we change, will predictably be a stopgap measure”. The unusual conservative position of, “just do the cost-benefit analysis for MCB”. How much damn energy I’ve spent filtering out the selectivity in what scraps of data scientists and economists want to show me. /rant
Here’s what I’m taking away from all that:
CCC isn’t reliable in general, but others have made estimates of the cost of worldwide MCB. I’m inclined to believe CCC about MCB in particular, as their numbers match up with others’. MCB is the most cost-effective climate intervention by a ~50x margin, and the estimated cost of worldwide MCB is 750M-1.5B USD annually. The exact technology needed to do MCB hasn’t been fleshed out yet, but could be engineered in a straightforward way.
By CCC’s own analysis, deploying worldwide MCB is >10x more cost-effective than standard global poverty interventions, and the fact that OPP and Givewell have far more funding than they know what to do with (even though they’re lying and saying they don’t), makes MCB even more attractive than this in practice.
Personally, I suspect that fleshing out the details of how MCB could be done in practice, would be more cost-effective than instituting a full-blown implementation of MCB, as having a well-defined way to implement it would reduce the friction for other(s) to implement it. Once I have hella money, it’s something I’d fund (the research on how to do it, but certainly not the actual MCB). Like, to get things to the point of having a written plan, “hey government, here’s exactly how you can do MCB if you want, now you can execute this plan as written if/when you choose”. I expect other interventions (re: factory farming) to be more effective than the actual MCB at preventing suffering.
Thanks for reading, and thanks for bringing MCB to my attention. Stay awesome.
I do think you might have put too much energy into thinking about the CCC though, haha. Maybe I should apologise for having mentioned them, without mentioning that I knew they’d taken money from dirty energy and I never got good epistemic vibes from them.
When I saw that stuff, I just read that as one of the many things we’d expect to see if MCB was legit, like, there would be a think-tank funded by dirty energy singing its praises, and even if that thinktank were earnest, I would still expect anyone who actually gave a shit about solving the problem to take the dirty money, because this kind of research ought to be rateable on the basis of whether or not it is true, rather than who paid for it, an extravagantly costly purity allegiance signal such as rejecting money from the richest people who benefit from your research should not carry a lot of discursive weight (and I’m still fairly sure it wouldn’t have).
What set off alarm bells for me was when I realised just how much the individuals in the CCC were taking in salary for the kind of work they’re doing. It would seem to me that genuine activists never get paid like that. I’m still not sure what that means, though. They live in world I don’t know much about.
I suppose part of the reason I mentioned them is that they seemed to be gathering a lot of interesting heresies that our friends might like to know about, not just MCB. Did you find that was the case?
Thanks! This all sounds right. “CCC has interesting heresies”—was there stuff other than MCB and global poverty? It’s an interesting parallel to EA—that they have interesting heresies, but are ultimately wrong about some key assumptions (that there’s room for more funding/that MCB is sufficient to stop all climate change, respectively. And they both have a fetish for working within systems rather than trying to change them at all.)
Kinda a shame that leftists are mostlynot coming to the “how can we change systems that will undo any progress we make” thing with an effectiveness mindset, though at least these people are.
He also literally lies in his cost-benefit analysis (by more than 10x).
What’s the literal lie, here? The link seems to say that a group led by Lomborg made misleading statements about how they made their prioritisations, but I can’t see any outright falsehoods.
b/c of doing the analysis and then not ranking shit in order.
Further down the list, we find a very controversial project, that is geo-engineering to reduce the intensity of incoming solar radiation to counteract global warming. According to a background paper, such investments would give a return rate of about 1,000. In spite of this enormous return rate, this is given moderate priority, apparently because it is deemed rather uncertain if this will actually work as intended.
> The lowest ranking accepted project, project no. 16, is called “Borehole and public hand pump intervention”. This has an estimated benefit-cost-ratio of loess than 3.4.
Next, we come to priority no. 17, the highest ranking not-accepted project. This is “Increased funding for green energy research and development”. According to the authors of the background paper, this has benefit-cost-ratios of 10 or more if the time horizon is slightly more than 1 decade. It is therefore a bit strange that this is placed below a project with a clearly less favourable benefit-cost-ratio.
do your own research if you disagree, but if you use “apparently because it is deemed rather uncertain if this will actually work as intended.” as an excuse to rate something poorly because you wanted to anyways rather than either do more research and update it, or even just make a guess, then wtf?
We are not playing, “is this plausibly defensible”, we are playing, “what was this person’s algorithm and are the systematically lying”.
I prefer to reserve “literally lying” for when people intentionally say things that are demonstrably false. It’s useful to have words for that kind of thing. As long as things are plausibly defensible, it seems better to say that he made “misleading statements”, or something like that.
Actually, I’m not even sure that this was a particularly egregious error. Given that they never say they’re going to rank things after the explicit cost-effectiveness estimates, not doing that seems quite reasonable to me. See for example givewell’s why we can’t take expected value estimates literally. All the arguments in that article should be even stronger when it’s different people making estimates across different areas. If you think that people should “make a guess” even when they don’t have time to do more research, that’s a methodological disagreement with a non-obvious answer.
I still think it’s plausible that some of the economists were acting in bad faith (it’s certainly bad that they don’t even give qualititive justifications for some of their rankings). But when their actions are plausibly defensible in any particular instance, you need several different pieces of evidence to be confident of that (like where they get their funding from, if they’re making systematic errors in the same direction, etc). If someone are saying things that I would classify as “literal lies”, that’s significantly stronger evidence that they’re acting in bad faith, which means you can skip over some of that evidence-gathering. I thought that you were claiming that Lomborg had made such a statement, and the fact that he hadn’t makes a large difference from my epistemical point of view, even if you have heard sufficiently much unrelated evidence to belive that he’s systematically acting in bad faith.
If you want to spend time predictably spinning in circles in your analysis because you can’t bring yourself to believe someone is lying, be my guest.
As for the specific authors: the individual reports written seem fine in themselves, and as for the geoengineering one, I know a guy who did a PhD under the author and said he’s generally trustworthy (I recall Vaniver was in his PhD program too). Like what I’m saying is the specific reports, e.g. Bickel’s report on geoengineering, seem fine, but Lomborg’s synthesis of them is shit, and you’re obscuring things with your niceness-and-good-faith approach.
Less of this, please. From what Lanrian is citing Lomborg does not come close to outright lying. (there might be more in the link, I have not read anything but the comments.) Accusing somebody of literally lying is a very strong accusation and should only be done in the egregious cases for all the usual reasons.
You are clearly well-informed about this matter. Your earlier comment was helpful and updated me in various directions. You could make me update me even more by applying the Principle of Charity.
I’ll give an answer that considers the details of the Copenhagen Consensus Center (CCC) and geoengineering, rather than being primarily a priori. I’ve spent a day and a half digging around and have zero prior knowledge. I spent too much time reading Lomborg and CCC in retrospect, so I mention him disproportionately relative to other sources.
Cross-posted to my blog.
Here’s what I notice:
1. Lomborg and his CCC seem very cost-benefit focused in their analysis. A few others are too, but see point 4. Basically, it’s easy to compare climate interventions to other interventions, but hard to figure out how much damage warming and other climate change will cause, so you can’t really figure out the benefit part of the cost-benefit analysis.
2. Lomborg and his CCC has recieved a ton of criticism for systematically making errors that underestimate the effect of climate change, and never making errors that overestimate the effect of it. One detailed account of him making such an error that could not have been made in good faith, is given here. He also literally lies in his cost-benefit analysis (by more than 10x).
There’s a lot of articles about Lomborg e.g. taking a 700k salary and getting donations from Koch and Exxon, which showed up before I found the above examples of him lying about data. I reacted to the info on his salary/funding by saying, “this is indicative of him being sketchy, but instead of just changing my estimate of how likely he is to be sketch by however much (“updating”) and calling it a day, I’m going to take this as a cue to dig into things until I have a firm understanding of whether this guy is systematically lying or not”. Turns out he’s a liar (see previous paragraph).
3. Page 33 of this CCC paper notes that,
Which corresponds to a 0.6 C average temperature change. This should put in perspective what a huge effect a 0.3% change in how much heat gets stays in earth’s atmosphere has on global temperatures. If all the glaciers melted, the earth’s temperature would rise 10 C, because glaciers are good at reflecting heat back away from the earth (back of envelope, me); if the earth became totally frozen, the temperature would drop by 55 C (see this page), which is why the thing this post is about, MCB, can realistically change global temperatures by a couple degrees C. Note that 1⁄2 of the global temperature change that already happened was because some glaciers and snow already melted and stopped reflecting heat away from the earth. Basically, the science behind MCB is pretty solid and I’d expect it to basically work.
AFAICT the IPCC estimates of how much warming there will be in the future seem to take into account the fact that the melting of the glaciers will further speed warming in itself, in addition to the warming you get from rising CO2 levels. Except they don’t ever explicitly say whether that was considered as a factor, so I can’t be totally sure they took that into account, even though it sounds like the most obvious thing to consider. (I skimmed this whole damn thing, and it wasn’t said either way!) I guess my next course of action could be to annoy an author about it, though I think I’ll be lazy and not.
4. We have little enough data on “how much economic damage has global warming done so far” that we can’t make decent extrapolations to “how much economic damage will global warming do later”. Like, you have papers saying that the economic damage from 3 C of warming could be 1%, 5-20%, 23%, or 35% of the GDP. When you have zip for data, you fall back on your politics.
5. The obvious game theory consideration of, “it’s better if someone other than you spends money on global warming”. The normal lefty position of, “our institutions aren’t set up to coordinate well on this sort of problem, and every action against climate change, until we change, will predictably be a stopgap measure”. The unusual conservative position of, “just do the cost-benefit analysis for MCB”. How much damn energy I’ve spent filtering out the selectivity in what scraps of data scientists and economists want to show me. /rant
Here’s what I’m taking away from all that:
CCC isn’t reliable in general, but others have made estimates of the cost of worldwide MCB. I’m inclined to believe CCC about MCB in particular, as their numbers match up with others’. MCB is the most cost-effective climate intervention by a ~50x margin, and the estimated cost of worldwide MCB is 750M-1.5B USD annually. The exact technology needed to do MCB hasn’t been fleshed out yet, but could be engineered in a straightforward way.
By CCC’s own analysis, deploying worldwide MCB is >10x more cost-effective than standard global poverty interventions, and the fact that OPP and Givewell have far more funding than they know what to do with (even though they’re lying and saying they don’t), makes MCB even more attractive than this in practice.
Personally, I suspect that fleshing out the details of how MCB could be done in practice, would be more cost-effective than instituting a full-blown implementation of MCB, as having a well-defined way to implement it would reduce the friction for other(s) to implement it. Once I have hella money, it’s something I’d fund (the research on how to do it, but certainly not the actual MCB). Like, to get things to the point of having a written plan, “hey government, here’s exactly how you can do MCB if you want, now you can execute this plan as written if/when you choose”. I expect other interventions (re: factory farming) to be more effective than the actual MCB at preventing suffering.
Thanks for reading, and thanks for bringing MCB to my attention. Stay awesome.
Thank you for looking into this! <3
I do think you might have put too much energy into thinking about the CCC though, haha. Maybe I should apologise for having mentioned them, without mentioning that I knew they’d taken money from dirty energy and I never got good epistemic vibes from them.
When I saw that stuff, I just read that as one of the many things we’d expect to see if MCB was legit, like, there would be a think-tank funded by dirty energy singing its praises, and even if that thinktank were earnest, I would still expect anyone who actually gave a shit about solving the problem to take the dirty money, because this kind of research ought to be rateable on the basis of whether or not it is true, rather than who paid for it, an extravagantly costly purity allegiance signal such as rejecting money from the richest people who benefit from your research should not carry a lot of discursive weight (and I’m still fairly sure it wouldn’t have).
What set off alarm bells for me was when I realised just how much the individuals in the CCC were taking in salary for the kind of work they’re doing. It would seem to me that genuine activists never get paid like that. I’m still not sure what that means, though. They live in world I don’t know much about.
I suppose part of the reason I mentioned them is that they seemed to be gathering a lot of interesting heresies that our friends might like to know about, not just MCB. Did you find that was the case?
Thanks! This all sounds right. “CCC has interesting heresies”—was there stuff other than MCB and global poverty? It’s an interesting parallel to EA—that they have interesting heresies, but are ultimately wrong about some key assumptions (that there’s room for more funding/that MCB is sufficient to stop all climate change, respectively. And they both have a fetish for working within systems rather than trying to change them at all.)
Kinda a shame that leftists are mostly not coming to the “how can we change systems that will undo any progress we make” thing with an effectiveness mindset, though at least these people are.
What’s the literal lie, here? The link seems to say that a group led by Lomborg made misleading statements about how they made their prioritisations, but I can’t see any outright falsehoods.
b/c of doing the analysis and then not ranking shit in order.
do your own research if you disagree, but if you use “apparently because it is deemed rather uncertain if this will actually work as intended.” as an excuse to rate something poorly because you wanted to anyways rather than either do more research and update it, or even just make a guess, then wtf?
We are not playing, “is this plausibly defensible”, we are playing, “what was this person’s algorithm and are the systematically lying”.
I prefer to reserve “literally lying” for when people intentionally say things that are demonstrably false. It’s useful to have words for that kind of thing. As long as things are plausibly defensible, it seems better to say that he made “misleading statements”, or something like that.
Actually, I’m not even sure that this was a particularly egregious error. Given that they never say they’re going to rank things after the explicit cost-effectiveness estimates, not doing that seems quite reasonable to me. See for example givewell’s why we can’t take expected value estimates literally. All the arguments in that article should be even stronger when it’s different people making estimates across different areas. If you think that people should “make a guess” even when they don’t have time to do more research, that’s a methodological disagreement with a non-obvious answer.
I still think it’s plausible that some of the economists were acting in bad faith (it’s certainly bad that they don’t even give qualititive justifications for some of their rankings). But when their actions are plausibly defensible in any particular instance, you need several different pieces of evidence to be confident of that (like where they get their funding from, if they’re making systematic errors in the same direction, etc). If someone are saying things that I would classify as “literal lies”, that’s significantly stronger evidence that they’re acting in bad faith, which means you can skip over some of that evidence-gathering. I thought that you were claiming that Lomborg had made such a statement, and the fact that he hadn’t makes a large difference from my epistemical point of view, even if you have heard sufficiently much unrelated evidence to belive that he’s systematically acting in bad faith.
If you want to spend time predictably spinning in circles in your analysis because you can’t bring yourself to believe someone is lying, be my guest.
As for the specific authors: the individual reports written seem fine in themselves, and as for the geoengineering one, I know a guy who did a PhD under the author and said he’s generally trustworthy (I recall Vaniver was in his PhD program too). Like what I’m saying is the specific reports, e.g. Bickel’s report on geoengineering, seem fine, but Lomborg’s synthesis of them is shit, and you’re obscuring things with your niceness-and-good-faith approach.
Less of this, please. From what Lanrian is citing Lomborg does not come close to outright lying. (there might be more in the link, I have not read anything but the comments.) Accusing somebody of literally lying is a very strong accusation and should only be done in the egregious cases for all the usual reasons.
You are clearly well-informed about this matter. Your earlier comment was helpful and updated me in various directions. You could make me update me even more by applying the Principle of Charity.