Hello everyone. Well, this forum has blown up, and we (GG) have taken some punches. I want to list a few take-aways on my end:
One thing we’re hearing loud and clear is that there is a lot of worry among this group that having recommendations in categories that are not the most cost-effective will do more harm than good. I think this is worth considering, though I don’t totally agree. What I do agree with is that our site could be designed to lead people to the most cost-effective recommendations, and make deviations from this ideal more obvious. Based on this feedback, we’ve made some changes already to the site, changing ordering and adding text to be clear that we believe policy is the most effective donation category. We would like to make deeper changes in the future.
One argument in this thread that particularly resonated with me is that there would really be a lot of value for conducting some more rigorous cost-benefit analysis for the offset recommendations. While we still believe at this moment that our offset recommendations are solid, I think there is value in performing more explicit cost-effectiveness analysis and using this to re-evaluate our choices. This is something we intend to do in the near term.
There’s clearly a lot of disagreement over activism and our recommendation of the Sunrise Movement Education Fund (TSM) in particular. I’m still on the side of defending this recommendation, but hear and respect the arguments put forth and believe they deserve consideration. At the very least, we need to re-work our materials from this recommendation to lay out our assumptions and arguments better. I’ll write briefly more about activism and TSM a bit below.
Do you want to help? We’re currently at a place of limited funding, so we are not actively conducting research at the moment. That makes it hard to do all the improvements we’d like (and have promised above). Our plan was to spend the next couple of months fundraising before turning back to research, but this thread is making us reconsider that a bit. We’d be interested in taking on a volunteer or short term contract from the EA community to help put some of the legwork in shoring up our existing research (and looking at a couple of new things.) Interested? Have some experience in climate issues and thinking through cost-effectiveness in an EA framework? Reach out at givinggreen@idinsight.org
One thing that I’d like to make clear: our mission at Giving Green is to create the most social impact we can through our recommendations. This, notably, does not necessarily mean that we just find the most cost-effective charity and recommend it solely. I truly believe that by offering ‘best in class’ recommendations in different categories, we can reach a wider audience and increase the pie of money that we can influence. That’s why we do offsets. That is (partially) why we felt it important to search for donation options outside of the narrow mechanism of action advocated by Founder’s Pledge (“(i) advocates (ii) focused on accelerating (iii) neglected yet critical decarbonisation and carbon removal technologies” in the US). While we find FP’s arguments clear and compelling, I don’t share the confidence that this is the mechanism of action with the highest expected value of donations (more later on this). This lack of confidence plus the ability to widen the pool of donors we influence is why we think there is a lot of value in exploring other donation categories. I do understand the concern about dilution from the most cost-effective places, but believe this can be mitigated through clear (better than now) design and information on the website.
Now, just a little bit more on the meat of the disagreement, which is the TSM recommendation. If I understand Johannes’ argument, he agrees with the importance of activism for making policy, both for putting electoral pressure on legislators as well as agenda-setting for current legislators. But, if I were to summarize, there are two main arguments against activism: timing, neglectedness, and one more argument against TSM in particular: strategy.
First timing: Johannes declares “Biden has declared climate as one of his top four priorities,” as an argument that activist energy is no longer important in the current situation. First, I think it’s pretty clear the TSM can take huge credit for this being the case. They flexed their political muscles in the last election, won a seat at the Bernie-Biden task force, heavily influenced Biden’s climate strategy, and succeeded in putting it at the top of the agenda. While I understand the argument that this victory is now in place, I don’t think we should take this climate priority as a given and assume that it will stay in place without further activist pressure. Yes Biden has taken some executive actions on climate already, but there’s no concerted push (as far as I can tell yet) for concerted legislation. At one point Johannes asked what success looks like for TSM. My answer: It’s ensuring climate policy is a priority (in executive and regulatory action, as well as legislation), and indeed shifting the Overton window of climate policy. (Agree with Johannes that the “Green New Deal” is not a realistic piece of legislation.) And for that you need continued pressure.
Second, Neglectedness: Ok, we hear you that some of those numbers in our report on activism spending are out of date. Sorry. The late 2020 numbers from the ClimateWorks Foundation helps, but it appears that the quoted category of “Public engagement” spans a lot more than policy-focused activism, so I’m not sure this is so helpful either. We should probably do some work to get some better numbers, but overall I’m not so convinced that the amount of money spent tells us all that much about the marginal benefit of the extra dollar. For TSM in particular, their budget really ballooned in 2020 (financial reports aren’t out but we heard almost $10 million), so that’s a reason to give some pause. That being said, I don’t see a compelling case for diminishing returns at this level of funding, especially since activism relies on sheer numbers to be effective. But yes, I hear this argument and I think it’s something we should consider more closely.
Finally, strategy: A lot of the previous arguments have revolved around TSM potentially having a negative effect due to its polarization of the climate policy. I really think strategy (ie polarization vs compromise) is a deeply difficult issue that is the subject of many debates within US politics. It’s important to realize that this is a feature, not a bug. TSM has chosen to take a “radical” position and linked with other progressive priorities as a way of building a strong and cohesive political movement. It’s a strategy that has historical precedent, and has borne dividends at numerous other times in US History (civil rights movement, tea party). Does that mean it will work this time? Unclear. Is it risky? Sure. But to argue that you need a bipartisan strategy to influence policy in DC I believe ignores political realities. I agree that sometimes the solutions put forth by TSM (no CCS, etc.) are not ideal, but again this is part of the strategy to build a fired-up coalition. Like Johannes said, Sunrise are not going to be the ones at the table hammering out policy anyway. It’s just their job to get people to the table, and put power in the hands of the climate negotiators.
Our recommendation of TSM also stems on our philosophy of supporting multiple theories of change. There are lots of different ways that political change happens, and I think that the best way to enact the desired change is to accept this, be humble about deciding ex-ante which theory is right, and support multiple avenues.
All of this doesn’t really address Johannes’ main concern, which is that one needs to make the case not just for TSM to be “Good”, but to really be better than CATF at the margin. Based on our approach of providing recommendations in multiple categories and subscribing to multiple theories of change, I don’t think we need to show perfectly that TSM is better or equal to CATF. But we should be able to provide a compelling argument that impact is in the same ballpark. Johannes thinks it is not. I think it is. But also I think we need to do a better job of crafting this argument in our documents. That’s something we are going to work on.
Finally, one more thing. I think that there is a little bit of risk as an EA community of over-committing to a single method of action (“(i) advocates (ii) focused on accelerating (iii) neglected yet critical decarbonisation and carbon removal technologies” in the US) and very small set of organizations dedicated to this method. Like I said, I find FP’s arguments compelling and think their central recommendation of CATF is very solid. But I would be careful about overconfidence here. In our research, we interviewed 50+ experts from various sectors of the climate space. At some point in the conversation we always asked “If you could allocate money in the most cost-effective matter to fight climate change, what would you do?” we got a lot of different answers, but notably no one said CATF unprompted. We usually followed up and asked about CATF directly. When prompted, some people said “Oh yeah, they are great.” But others didn’t agree, and pointed out various issues with their approach or influence. I know this is not a scientific approach and don’t want to put too much weight on it, but I do just think it’s important for EAs as a community to approach this space with humility.
Thanks again, Dan & team, for your gracious and constructive comment! This is what I love about this community most. I think there are still lots of misunderstandings on the nature of the criticisms and severe and consequential disagreements on epistemics and empirics to which I reply to below. But before I do so, just a meta-point on why I engaged in this criticism in the first place.
Why I engage in this criticism
I do not enjoy criticizing. The fact that I engage in criticism is impact-related and not personal. Indeed, John (Halstead) and I spent much of the last year criticizing each other’s reasoning which is one of the reasons I believe the FP climate recs are very robust.
It is also not a criticism that I make as an FP-representative, it just happens to be the case that the goals of FP and of myself are perfectly aligned which is why I joined FP and why I now happen to lead its climate work.
So, that said, my criticism serves two purposes:
Clarify for the community what the current state of knowledge is on climate from an EA perspective.
Offering paths forward for GG to improve so that GG can realize its positive potential.
That said, and with constructive intent, onwards to the disagreements.
I first discuss key-takeaways from the debate and then dive into one particular fundamental area of disagreement that is prominent in GG’s comment—on what we know and can now.
What are the key take-aways from this debate?
You write “Johannes thinks” a lot which, in my view, makes the debate unnecessarily personal but, more importantly, also misrepresents the debate. (I will use “GG” in the following exactly to make it less personal).
I did not add any new points in my original comment, I just expanded on points that were already in Alex’s original post, made them more explicit and explained them in a more stepwise manner, giving more evidence and context to make it easier to follow (though arguably tediously long, apologies!).
But essentially, Alex and I, and, by their comments, Sanjay and John Halstead, as well as many other commentersall agreeon the critiques. In other words, rather than writing “Johannes thinks” it would be more accurate and less personalizing to say that ~all EAs that have expertise in climate charity evaluation disagree with GG on TSM and broadly think the same thing namely that:
1.From the weight of the evidence, it does not seem justified to think that we are unsure whether TSM or CATF is better, the evidence points in the clear direction of CATF > TSM
a. Note that saying “the weight of the evidence points in direction CATF > TSM” is not equivalent to saying “we know that CATF > TSM” or “we know that TSM = CATF”, it just says that, based on what we know now, we should assume that CATF > TSM. This is a more humble position than implied in the comment, about our current state of knowledge—not final truths.
b. This also means, to your point about humility, that further analysis into TSM is welcomed. Our work is never finished. I am open to finding TSM > CATF or that TSM and CATF are incomparable but both certainly good (which seems your position).
It would make my life a lot easier, because people like TSM so much and I could then treat directing people to TSM as high impact rather than convincing them of the opposite. This is another reason for skepticism on TSM being highly impactful, it would be a case of “suspicious convergence” is something that is like the “cat shelters” of climate change in terms of popularity would also be the highest impact option.
c. But it also means, via humility, that even if GG does not believe our criticisms, given that lots of people—many of which with significant experience in climate and on advocacy-charity evaluation—disagree with GG, it would be reasonable for GG to update rather strongly from that.
2.There are serious and, in our view, unaddressed concerns that marginal donations to TSM are net-negative, making the world worse, in expectation (not only about particular realizations, which, as you point out and everyone agrees, will always be true and should not be disqualifying).
3.Even if TSM is not net-negative, there appears near-consensus that—from what we know—it is likely that the impact of marginal TSM donations is very low—maybe positive, maybe negative—and unlikely to be large either way. This is quite different from CATF which means that there is concern that dollars that could go to CATF will go to TSM instead, with the TSM recommendation being net negative via that mechanism even if TSM alone is not.
4.When offsets and TSM are seen as a lot-lower impact than CATF or other high-impact recs, then “dilution” becomes an issue that needs to be modeled and managed to understand how good GG is from an EA perspective, how “dilution” plays out against crowding in additional money and how much that money is optimized.
I do understand thevalue propositionof Giving Green and I can imagine worlds where Giving Green’s approach, “meeting people where they are”, creates a lot of benefit even if these will often be comparatively low-impact options, because a lot of people are not willing to change their mind very much.
Indeed, I can imagine and hope for a world where the approach of Giving Green (optimizing giving of people not willing to change their mind much) and the approach of the wider EA community (incl. FP) focused on the highest-impact options is complementary.
So, I am not suggesting you should de-recommend TSM, I am just saying you should be clear to the EA community and—ideally, but I know this is a stretch—to your wider audience (e.g. on the website) that you have reasons to think that TSM < CATF in terms of expected benefit, just like offsets < policy.
To meet EAs “where they are”, GG should enable EAs that want to maximize positive impact in expectation. The charity that is most likely to lead to that in the climate space right now happens to be CATF.
On the flipside, I think this would also make many people here quite concerned about GG somewhat less concerned with GG if the risk—that impact-maximizing EA climate dollars go to TSM or offsets that are considered clearly worse—would be mitigated.
Also note that our theory of change is not only focused on energy innovation—we also examine other paths that lead to global leverage such as policy leadership and cascading effects (we evaluated CLC from that angle but came to the view that it won’t work).
What do we know and what can we know?
A lot of the discussion and disagreement seems to be about what we know and what we can know.
As I alluded to above, I agree with GG that humility is super-important—the entire endeavor of EA is doomed without epistemic humility.
I do, however, think humility in this context mostly points to GG trusting EA climate mainstream more (see above) for four reasons (which also explain a bit more how FP is thinking about research).
The TL,DR of it is as follows:
1. There is a lot of expert support for the CATF recommendation and there is a lot more uncertainty regarding TSM.
2. CATF looks very good on the theory of change/frame most relevant to effective climate action—maximizing global decarbonization benefit—and the argument for TSM on that frame is not made.
3. Charity evaluation methodology is our friend and allows us to draw useful inferences even in highly uncertain situations.
4. The length and depth of engagement that led to the CATF and similar recommendations should itself be a reason for confidence, more so than the GG comment suggests.
1. What can we learn from which experts?
As GG rightly points out, asking 50 climate (policy) experts for their favorite charities isn’t a scientific approach to identifying effective charities, but of course it is evidence of some kind for something. In this section, I will ask “what for?” as well as what I would consider the relevant expert evidence in support of CATF.
Why typical US climate experts are not experts for the question at hand
For the question “what are the most effective charities to support to have the most positive impact on climate?” – the questions that EAs asks when trying to identify the highest-impact opportunities – the typical US-based climate (policy) expert isnotan expert. This is so for at least three reasons:
1. The US-debate is focused on the US, the typical expert does not ask “what is the most effective thing one can do to have an effect on global decarbonization?” but rather “what seems best in the US?”. As per theory of change discussion, those two things aren’t necessarily related very much at all.
2. The overall climate debate in Western countries is quite biased – and predictably and importantly so. Compared to their real importance, there is an overemphasis on renewables, energy efficiency, and electric cars, there is an over-emphasis on lifestyle changes vs technological changes and there is an underappreciation of the technological challenge – but rather an emphasis on the political challenge and a typical framing that all that is missing is political will.
All of this makes sense (is predictable) if we think about where mainstream environmentalism comes from – ideologically speaking – from an egalitarian philosophy that prefers to solve problems with virtue rather than technology, that prefers “small is beautiful” and “in harmony with nature” over big centralized solutions (such as nuclear power or large dams), that has a moral need to demonize large centralized structures such as the state or capitalism, etc. Of course, not all experts are steeped in this ideology and not all experts that are follow it to the fullest extent. But it is notable that the overall climate debate is very biased by a particular worldview, an egalitarian / green view, that systematically hypes solutions that fit into the ideological priors – such as decentralized renewables and micro-grids and appealing to individual morality – and, equally systematically, disregards others – such as nuclear power build outs or CCS or a large-scale innovation agenda in line with the climate challenge.
Of course, a naïve narrative that solely focuses on technological innovation is equally biased, which is why I stressed so much in my original comment that I did not come to this topic with techno-optimist priors. All of those simplified ideological stories have kernels of truth to them and we need to judge the overall bias of the debate.
But if you look at the overall climate debate in Western countries it is hard to argue that the innovation arguments are getting their fair share – that innovation is emphasized enough compared to its importance (also see theory of change section).
It is not surprising that in this biased conversation Sunrise comes out on top – it serves all the ideological priors of those voices dominating the climate conversation – but that does not make it correct.
[Aside: Indeed, I would argue that because of the unfortunate situation that the innovation arguments tends to be made by those more that do not really want to act, the innovation agenda is neglected compared to its importance.
Climate activists downplay it because they like existing solutions as well as a framing of the problem as one of evil corporations and the need for fundamental societal change (rather than a severe technological challenge that needs strengthened innovation policy and a move away from “100% renewables”), whereas they often suspect – and often rightly so – that arguments for innovation on the right are made in bad faith, as discursive responses justifying inaction.
This leaves us in a terrible situation where we are severely underinvesting in technological innovation despite it being the only solution to get to net zero globally. This is one of the reasons why at FP we focus on funding innovation advocates that are on the center left but that are more serious about innovation than the green mainstream and that are able to work with people making innovation arguments on the right]
3. The typical climate (policy) expert is not an impact-focused philanthropist – indeed they are not philanthropists at all.
The best philanthropic options are those with the highest expected marginal returns to additional funding, something that a typical climate (policy) expert has very little insight into. Rather, the expert – when prompted with that question – will likely answer with something that seems good to them, that is top-of-mind, that fits with the ideological inclinations of the expert, etc. This is not data about high-impact philanthropy, this is data about the discourse.
What is this evidence for?
I am not sure that the fact that CATF does not come out on top of many of those experts is evidence of anything, but, if it is evidence of something, it seems weak evidence in support of CATF—if it were popular with everyone we should not be funding it!
Indeed, this points to the “suspicious convergence” point on TSM v CATF. The fact that TSM is so popular and that CATF is not should make us fairly skeptical, by itself, that we should fund TSM rather than CATF, and that TSM is the most effective thing to fund.
In comparatively non-neglected causes such as climate we should be quite wary of funding the popular things. Indeed, the entire theory of value of CATF and other charities that FP champions is that these are charities that improve societal resource allocation in a field that receives a lot of resources but where those are spent fairly inefficiently.
In other words, it is precisely because CATF is focused on inconvenient and neglected things – CCS, advanced nuclear, zero-carbon fuels, industrial decarbonization, etc. – that we should not expect them to be popular but very much worthy of support.
Who are the relevant experts for the question at hand?
There are two types of experts that seem most relevant for finding high-impact climate philanthropy options:
1. Aligned climate philanthropists
Among aligned climate philanthropists – those that want to have the maximum impact on global decarbonization solutions with their dollars and approach this topic rigorously – CATF and/or the approach it stands for has a lot of supporters.
Most prominently that would be Bill Gates who does not get tired to advocate for innovation and technology pluralism. But there are also many climate philanthropists at US and European foundations that are strong supporters of CATF for exactly the reasons that we support them (not sharing publicly here, but happy to share in PM). As such, there are many relevant experts that would support CATF as a top choice.
2. Experts on sub-topics that help inform the theory of change, identify effective interventions, etc.
Experts are of course essential to form a view of effective interventions in the climate space and of effective organizations pursuing these.
Indeed, as I mentioned, the FP research is informed by hundreds of conversations with experts.
But apart from experts on particular organizations or interventions, the way expert views flow into the research is mostly via getting a sense of the most effective interventions, the bottlenecks, the neglected areas, the overall theory of change. There is a lot of information and expert opinion out there supporting the choice of CATF more generally that does not mention or is even aware of those organizations at all which leads me to my next point, “what is the appropriate theory of change?”
2. What is the appropriate theory of change?
Global decarbonization, not US policy change
GG states that there are several theories of (political) change and this is undoubtedly true.
It is important to recognize, though, that the theories of change that GG focuses on – theories of change in the US political context – are not directly relatedat allto a theory of change that focuses on maximal global decarbonization impact.
While this would be fine if the goal would be to identify the best charities reducing emissions in the US, it becomes problematic when it is claimed based on GG analysis that we are entirely ignorant about the expected value of donations to CATF or TSM.
If we ask “what’s the most effective climate funding opportunity to support at the margin that we know of?”, theories of political change in the US only matter insofar as emissions reductions “travel” – in a causal sense—through US political change.
Given that US emissions are only ~15% of global emissions and that this share is declining as the rest of the world gets richer and the US is decarbonizing (even if on current trajectory), but that the US has roughly half of the world’s energy innovation budget, is generally considered the most capable innovation economy, and that the US is – for the moment, in any case – the leading global power, it stands to reason that the most effective things one can focus on in the US with regards to climate are interventions that maximize indirect effects and that might be quite different from those maximizing US emissions reductions (in expectation).
This is at the heart of the misunderstanding of CATF as “incrementalist”. CATF might be framed in that way if one is focused on US emissions and US political change. OK.
But the reason that most EAs support CATF as high-impact climate org is not because of the reduced emissions in the US but, precisely, because of the large international benefits that innovation policy targeted at neglected technologies brings.
All of the major success stories we have seen in climate over the past 20 years – solar, wind, coal > gas in the US, electric cars and batteries – have been the result of relatively narrow and targeted policies, the kind of which CATF advances for technologies that are less popular with greens for reasons of ideology, not merit.
Indeed, to make this more explicit, a lot of my skepticism about TSMat the margincomes from scenarios where TSM right now is relatively close to its maximum potential – lots of sway on Democrats, little sway on Republicans – but where binding climate legislation or more sweeping legislation still remains off the table and will remain so for relevant periods (getting a binding climate target in the US in 2035 is not that useful).
In that world, TSM will put lots of pressure on Biden and Democrats to advance executive orders (we can already observe that) that are ultimately unstable and open to court challenges rather than seizing the opportunity that presents itself to make much stronger innovation policy happen. This is a real danger and this is why I am very wary of stating certainty of positive impact in expectation of marginal TSM donations.
While one might be quite fatalistic about mandatory climate targets of the kind we have in Europe for the US, the response to that shouldn’t necessarily be – just push harder and harder and some day the Democrats will get rid of the filibuster or find a way to make executive orders robust and then we will win – but rather to make the most of the opportunity that the US presents for climate – an unparalleled energy innovation system that has a serious shot at breakthroughs in CCS, advanced nuclear, etc. Ambitious economy-wide climate legislation in the US is neither a necessary nor a sufficient condition for climate progress globally.
Climate is not, fundamentally, a problem like civil rights or the other analogies where a solution will need a strong legislation passing Congress and for which a large grassroots movement and even one larger than Sunrise right now might be essential.
CATF and similar charities look very good on a theory of change focused on the global picture
And CATF looks very good on a theory of change focused on maximized global decarbonization impact when taking into account some of the most important stylized facts about the climate challenge (widely recognized as median views in the respective expert communities):
1. Global energy demand will grow and restricting energy demand growth is very problematic from a humanitarian perspective.
2. Effective global decarbonization requires a much larger set of technologies than those currently available. Most of those technologies are not on track and many necessary technologies are in early stages.
You can then combine this with two CATF-specific features:
5. CATF is a strong organization that translates money into effective advocacy. This is not controversial within the EA community, something GG agrees on. It was first established in the FP 2018 report and it appears that at least 4 EA orgs had multiple calls with CATF, often dozens, that reaffirmed this conclusion (FP, Legacies Now, SoGive, Giving Green).
6. CATF has very productive funding margins, projects that are currently unfunded and that make a lot of sense from the above stylized facts and the theory of change.
This is all you need to come to CATF as a likely local optimum in effective climate philanthropy.
None of this is controversial and – indeed – each of the claims above about the world in general (1-4) follow directly from median expert views on those respective topics and the CATF-specific claims (5-6) are even entirely uncontroversial across the EA community.
In contrast, motivating TSM as a top-choice requires a lot of controversial claims, such as (a) that we are sure that the impact of marginal TSM donations is not negative in expectation and (b) that additional effort can lead to significant change beyond what is already baked in despite the approach of Sunrise being partisan and thereby, quite plausibly, limited in its ultimate potential given the structure of the Senate and the Electoral College.
What is more, as Robin pointed out, the fact that CATF operates not only in the US is another source of believing in it to be higher impact—CATF can optimize marginal resources across many jurisdictions, many of which are often more fruitful than the US context, which is very fruitful now but might as well dry up significantly.
3. What can we know through charity evaluation methodology?
A lot of the arguments for GG’s skepticism about being able to figure out the goodness of CATF v TSM comes from non-marginal arguments, from the stated inherent incompatibility between different approaches.
While I do disagree with these incomparability claims—see my original comment—even if one agrees with these there is a lot of traction from the methods that “EA-style” (for lack of a better word) charity evaluation provides.
Neglectedness
Neglectedness is a proxy, but it is a very useful one. In a world with millions of funding margins which we can never all evaluate in detail, neglectedness gives crucial information (as a prior, also see my long comment).
If a field is flooded with attention and money, like progressive climate activism, then the probability that there are great things to be funded is low for three fundamental reasons:
Declining marginal returns: A strategic actor will order projects such that the lowest value ones are at the margin and the further we are down that margin, the lower we should expect the value to be (ceteris paribus, which is key, e.g. CATF’s expansion into other geographies, changing its funding margins is an argument in another direction).
Declining probability of true additionality: The more money and attention goes into a particular direction or org the less likely it is that additional funding there will be truly additional—it is more likely that actors “satisfice” and, when the funding environment is fertile, additional dollars will not always be additional, as there might be funding targets or, at the very least, a declining eagerness to fundraise more. That’s the point about 10,000 volunteers and inspiring the public imagination of an entire wing of the dominant American party—if in that situation you are seriously funding-constrained, this would be quite surprising.
3.** Declining probability of money being the impact-constraining resource:** The more money you have and the easier it is to get more money if you need it, the lower the probability that money really is your constraint on higher marginal impact.
While I won’t name names here, there are many great climate charities which I believe are doing incredibly important and good work, which we are not recommending because of such considerations—this is not something that only applies to TSM.
Aggregating expert views
When we have an issue where relevant experts disagree strongly—such as whether a more partisan approach to climate policy is good or bad—epistemic humility pushes us closer to a lower value (zero).
This is where a lot of the TSM skepticism comes from.
Not from saying “we are sure that TSM is bad” but rather “when observing the debate, we notice that this is an area of severe disagreement between reasonable people”—e.g. look at all the debates in the Democratic party between progressives and moderates on which path is more promising to ensure Democratic priorities are implemented (and the Democratic majority survives).
So, this would be an area where we would either need to set the marginal impact of TSM downwards or where more research would be needed to demonstrate that the moderate wing of the Democratic party is entirely wrong (or, more climate-specifically, that more centrist voices are unduly worried about a progressive over-reach that will backfire and lead to deadlock and no progress).
This argument is largely missing in the TSM analysis, instead it feels like an assessment that is based on the most positive expert interpretations of TSM, not the median ones, and not taking into account voices that would be critical of TSM’s impact.
4. Length and depth of engagement that led to CATF and other climate high-impact recs
This is about the “intellectual history” leading to the CATF recommendation, with the TL, DR that the recommendation is based on a lot deeper engagement than the GG comment is giving credit for.
The argument on which CATF comes out as the best recommendation we currently know of goes back to 2016 -- when the argument that EAs interested in climate should focus on (a) innovation for (b) neglected technology, (c) given a world of rising energy demand and real conflicts between climate and poverty alleviation, was first publicly made (to my knowledge). You can watch it here and, somewhat later, but more refined, here. There’s a write-up here (though I never quite finished it because I focused on embedding this argument in the EA debate, which was achieved with John’s and Hauke’s reports, leaving me to prioritize other things).
This argument was already informed by deep engagement with climate for about a decade, reading and talking to hundreds of experts and, arguably—in the process—becoming an expert at being better able to evaluate the strength of different claims as well.
These arguments then found their ways into both John’s 2018 report and Hauke’s Let’s Fund report, both of whom vetted this argument against competing arguments. I don’t know the details about Hauke’s report, but I know that John talked to at least 50 experts for the first report and I know—from, sometimes painful, experience :) -- that John does not accept any argument at face-value before he has not engaged with it himself. Crucially, both Hauke and John made charity recommendations that turned this argument into something actionable and thereby unlocking massive value.
After I joined FP in late 2019, John and I spent a lot of time revisiting our recs (which also led to not continuing some recs), working on a new report that more clearly motivates our theory of change, and identifying new recs.
The robustness of the CATF recommendation is one of the few things that stood the test of time, indeed by more clearly articulating our theory of change—improving societal resource allocation in light of decarbonization priorities, towards innovation in neglected technologies—we became somewhat more convinced of CATF being a “local optimum” (obviously, there might be other great options—I for one would never claim that we will never find something better than CATF, just that until now we haven’t).
This is also true for the EA community at large: while there have been many critical reviews of FP’s work on climate (e.g. concerns about CfRN), the CATF recommendation has stood the test of time and examination.
I will need a bit of time for a full reply, but I wanted to let you and team know that I really appreciate the thoughtful, gracious, and civil reply.
While we do have many disagreements on epistemics and empirics—and I think I also still deeply disagree with many things in the post above (to be explained) -- we are united in the same purpose, making the world better the best we can.
So the point of this comment is just to recognize that and thank you.
Hello everyone. Well, this forum has blown up, and we (GG) have taken some punches. I want to list a few take-aways on my end:
One thing we’re hearing loud and clear is that there is a lot of worry among this group that having recommendations in categories that are not the most cost-effective will do more harm than good. I think this is worth considering, though I don’t totally agree. What I do agree with is that our site could be designed to lead people to the most cost-effective recommendations, and make deviations from this ideal more obvious. Based on this feedback, we’ve made some changes already to the site, changing ordering and adding text to be clear that we believe policy is the most effective donation category. We would like to make deeper changes in the future.
One argument in this thread that particularly resonated with me is that there would really be a lot of value for conducting some more rigorous cost-benefit analysis for the offset recommendations. While we still believe at this moment that our offset recommendations are solid, I think there is value in performing more explicit cost-effectiveness analysis and using this to re-evaluate our choices. This is something we intend to do in the near term.
There’s clearly a lot of disagreement over activism and our recommendation of the Sunrise Movement Education Fund (TSM) in particular. I’m still on the side of defending this recommendation, but hear and respect the arguments put forth and believe they deserve consideration. At the very least, we need to re-work our materials from this recommendation to lay out our assumptions and arguments better. I’ll write briefly more about activism and TSM a bit below.
Do you want to help? We’re currently at a place of limited funding, so we are not actively conducting research at the moment. That makes it hard to do all the improvements we’d like (and have promised above). Our plan was to spend the next couple of months fundraising before turning back to research, but this thread is making us reconsider that a bit. We’d be interested in taking on a volunteer or short term contract from the EA community to help put some of the legwork in shoring up our existing research (and looking at a couple of new things.) Interested? Have some experience in climate issues and thinking through cost-effectiveness in an EA framework? Reach out at givinggreen@idinsight.org
One thing that I’d like to make clear: our mission at Giving Green is to create the most social impact we can through our recommendations. This, notably, does not necessarily mean that we just find the most cost-effective charity and recommend it solely. I truly believe that by offering ‘best in class’ recommendations in different categories, we can reach a wider audience and increase the pie of money that we can influence. That’s why we do offsets. That is (partially) why we felt it important to search for donation options outside of the narrow mechanism of action advocated by Founder’s Pledge (“(i) advocates (ii) focused on accelerating (iii) neglected yet critical decarbonisation and carbon removal technologies” in the US). While we find FP’s arguments clear and compelling, I don’t share the confidence that this is the mechanism of action with the highest expected value of donations (more later on this). This lack of confidence plus the ability to widen the pool of donors we influence is why we think there is a lot of value in exploring other donation categories. I do understand the concern about dilution from the most cost-effective places, but believe this can be mitigated through clear (better than now) design and information on the website.
Now, just a little bit more on the meat of the disagreement, which is the TSM recommendation. If I understand Johannes’ argument, he agrees with the importance of activism for making policy, both for putting electoral pressure on legislators as well as agenda-setting for current legislators. But, if I were to summarize, there are two main arguments against activism: timing, neglectedness, and one more argument against TSM in particular: strategy.
First timing: Johannes declares “Biden has declared climate as one of his top four priorities,” as an argument that activist energy is no longer important in the current situation. First, I think it’s pretty clear the TSM can take huge credit for this being the case. They flexed their political muscles in the last election, won a seat at the Bernie-Biden task force, heavily influenced Biden’s climate strategy, and succeeded in putting it at the top of the agenda. While I understand the argument that this victory is now in place, I don’t think we should take this climate priority as a given and assume that it will stay in place without further activist pressure. Yes Biden has taken some executive actions on climate already, but there’s no concerted push (as far as I can tell yet) for concerted legislation. At one point Johannes asked what success looks like for TSM. My answer: It’s ensuring climate policy is a priority (in executive and regulatory action, as well as legislation), and indeed shifting the Overton window of climate policy. (Agree with Johannes that the “Green New Deal” is not a realistic piece of legislation.) And for that you need continued pressure.
Second, Neglectedness: Ok, we hear you that some of those numbers in our report on activism spending are out of date. Sorry. The late 2020 numbers from the ClimateWorks Foundation helps, but it appears that the quoted category of “Public engagement” spans a lot more than policy-focused activism, so I’m not sure this is so helpful either. We should probably do some work to get some better numbers, but overall I’m not so convinced that the amount of money spent tells us all that much about the marginal benefit of the extra dollar. For TSM in particular, their budget really ballooned in 2020 (financial reports aren’t out but we heard almost $10 million), so that’s a reason to give some pause. That being said, I don’t see a compelling case for diminishing returns at this level of funding, especially since activism relies on sheer numbers to be effective. But yes, I hear this argument and I think it’s something we should consider more closely.
Finally, strategy: A lot of the previous arguments have revolved around TSM potentially having a negative effect due to its polarization of the climate policy. I really think strategy (ie polarization vs compromise) is a deeply difficult issue that is the subject of many debates within US politics. It’s important to realize that this is a feature, not a bug. TSM has chosen to take a “radical” position and linked with other progressive priorities as a way of building a strong and cohesive political movement. It’s a strategy that has historical precedent, and has borne dividends at numerous other times in US History (civil rights movement, tea party). Does that mean it will work this time? Unclear. Is it risky? Sure. But to argue that you need a bipartisan strategy to influence policy in DC I believe ignores political realities. I agree that sometimes the solutions put forth by TSM (no CCS, etc.) are not ideal, but again this is part of the strategy to build a fired-up coalition. Like Johannes said, Sunrise are not going to be the ones at the table hammering out policy anyway. It’s just their job to get people to the table, and put power in the hands of the climate negotiators.
Our recommendation of TSM also stems on our philosophy of supporting multiple theories of change. There are lots of different ways that political change happens, and I think that the best way to enact the desired change is to accept this, be humble about deciding ex-ante which theory is right, and support multiple avenues.
All of this doesn’t really address Johannes’ main concern, which is that one needs to make the case not just for TSM to be “Good”, but to really be better than CATF at the margin. Based on our approach of providing recommendations in multiple categories and subscribing to multiple theories of change, I don’t think we need to show perfectly that TSM is better or equal to CATF. But we should be able to provide a compelling argument that impact is in the same ballpark. Johannes thinks it is not. I think it is. But also I think we need to do a better job of crafting this argument in our documents. That’s something we are going to work on.
Finally, one more thing. I think that there is a little bit of risk as an EA community of over-committing to a single method of action (“(i) advocates (ii) focused on accelerating (iii) neglected yet critical decarbonisation and carbon removal technologies” in the US) and very small set of organizations dedicated to this method. Like I said, I find FP’s arguments compelling and think their central recommendation of CATF is very solid. But I would be careful about overconfidence here. In our research, we interviewed 50+ experts from various sectors of the climate space. At some point in the conversation we always asked “If you could allocate money in the most cost-effective matter to fight climate change, what would you do?” we got a lot of different answers, but notably no one said CATF unprompted. We usually followed up and asked about CATF directly. When prompted, some people said “Oh yeah, they are great.” But others didn’t agree, and pointed out various issues with their approach or influence. I know this is not a scientific approach and don’t want to put too much weight on it, but I do just think it’s important for EAs as a community to approach this space with humility.
Thanks again, Dan & team, for your gracious and constructive comment! This is what I love about this community most. I think there are still lots of misunderstandings on the nature of the criticisms and severe and consequential disagreements on epistemics and empirics to which I reply to below. But before I do so, just a meta-point on why I engaged in this criticism in the first place.
Why I engage in this criticism
I do not enjoy criticizing. The fact that I engage in criticism is impact-related and not personal. Indeed, John (Halstead) and I spent much of the last year criticizing each other’s reasoning which is one of the reasons I believe the FP climate recs are very robust.
It is also not a criticism that I make as an FP-representative, it just happens to be the case that the goals of FP and of myself are perfectly aligned which is why I joined FP and why I now happen to lead its climate work.
So, that said, my criticism serves two purposes:
Clarify for the community what the current state of knowledge is on climate from an EA perspective.
Offering paths forward for GG to improve so that GG can realize its positive potential.
That said, and with constructive intent, onwards to the disagreements. I first discuss key-takeaways from the debate and then dive into one particular fundamental area of disagreement that is prominent in GG’s comment—on what we know and can now.
What are the key take-aways from this debate?
You write “Johannes thinks” a lot which, in my view, makes the debate unnecessarily personal but, more importantly, also misrepresents the debate. (I will use “GG” in the following exactly to make it less personal).
I did not add any new points in my original comment, I just expanded on points that were already in Alex’s original post, made them more explicit and explained them in a more stepwise manner, giving more evidence and context to make it easier to follow (though arguably tediously long, apologies!).
But essentially, Alex and I, and, by their comments, Sanjay and John Halstead, as well as many other commenters all agree on the critiques. In other words, rather than writing “Johannes thinks” it would be more accurate and less personalizing to say that ~all EAs that have expertise in climate charity evaluation disagree with GG on TSM and broadly think the same thing namely that:
1.From the weight of the evidence, it does not seem justified to think that we are unsure whether TSM or CATF is better, the evidence points in the clear direction of CATF > TSM
a. Note that saying “the weight of the evidence points in direction CATF > TSM” is not equivalent to saying “we know that CATF > TSM” or “we know that TSM = CATF”, it just says that, based on what we know now, we should assume that CATF > TSM. This is a more humble position than implied in the comment, about our current state of knowledge—not final truths.
b. This also means, to your point about humility, that further analysis into TSM is welcomed. Our work is never finished. I am open to finding TSM > CATF or that TSM and CATF are incomparable but both certainly good (which seems your position).
It would make my life a lot easier, because people like TSM so much and I could then treat directing people to TSM as high impact rather than convincing them of the opposite. This is another reason for skepticism on TSM being highly impactful, it would be a case of “suspicious convergence” is something that is like the “cat shelters” of climate change in terms of popularity would also be the highest impact option.
c. But it also means, via humility, that even if GG does not believe our criticisms, given that lots of people—many of which with significant experience in climate and on advocacy-charity evaluation—disagree with GG, it would be reasonable for GG to update rather strongly from that.
2.There are serious and, in our view, unaddressed concerns that marginal donations to TSM are net-negative, making the world worse, in expectation (not only about particular realizations, which, as you point out and everyone agrees, will always be true and should not be disqualifying).
3.Even if TSM is not net-negative, there appears near-consensus that—from what we know—it is likely that the impact of marginal TSM donations is very low—maybe positive, maybe negative—and unlikely to be large either way. This is quite different from CATF which means that there is concern that dollars that could go to CATF will go to TSM instead, with the TSM recommendation being net negative via that mechanism even if TSM alone is not.
4.When offsets and TSM are seen as a lot-lower impact than CATF or other high-impact recs, then “dilution” becomes an issue that needs to be modeled and managed to understand how good GG is from an EA perspective, how “dilution” plays out against crowding in additional money and how much that money is optimized.
I do understand the value proposition of Giving Green and I can imagine worlds where Giving Green’s approach, “meeting people where they are”, creates a lot of benefit even if these will often be comparatively low-impact options, because a lot of people are not willing to change their mind very much.
Indeed, I can imagine and hope for a world where the approach of Giving Green (optimizing giving of people not willing to change their mind much) and the approach of the wider EA community (incl. FP) focused on the highest-impact options is complementary.
So, I am not suggesting you should de-recommend TSM, I am just saying you should be clear to the EA community and—ideally, but I know this is a stretch—to your wider audience (e.g. on the website) that you have reasons to think that TSM < CATF in terms of expected benefit, just like offsets < policy.
To meet EAs “where they are”, GG should enable EAs that want to maximize positive impact in expectation. The charity that is most likely to lead to that in the climate space right now happens to be CATF.
On the flipside, I think this would also make many people here quite concerned about GG somewhat less concerned with GG if the risk—that impact-maximizing EA climate dollars go to TSM or offsets that are considered clearly worse—would be mitigated.
Also note that our theory of change is not only focused on energy innovation—we also examine other paths that lead to global leverage such as policy leadership and cascading effects (we evaluated CLC from that angle but came to the view that it won’t work).
What do we know and what can we know?
A lot of the discussion and disagreement seems to be about what we know and what we can know.
As I alluded to above, I agree with GG that humility is super-important—the entire endeavor of EA is doomed without epistemic humility.
I do, however, think humility in this context mostly points to GG trusting EA climate mainstream more (see above) for four reasons (which also explain a bit more how FP is thinking about research).
The TL,DR of it is as follows:
1. There is a lot of expert support for the CATF recommendation and there is a lot more uncertainty regarding TSM.
2. CATF looks very good on the theory of change/frame most relevant to effective climate action—maximizing global decarbonization benefit—and the argument for TSM on that frame is not made.
3. Charity evaluation methodology is our friend and allows us to draw useful inferences even in highly uncertain situations.
4. The length and depth of engagement that led to the CATF and similar recommendations should itself be a reason for confidence, more so than the GG comment suggests.
1. What can we learn from which experts?
As GG rightly points out, asking 50 climate (policy) experts for their favorite charities isn’t a scientific approach to identifying effective charities, but of course it is evidence of some kind for something. In this section, I will ask “what for?” as well as what I would consider the relevant expert evidence in support of CATF.
Why typical US climate experts are not experts for the question at hand
For the question “what are the most effective charities to support to have the most positive impact on climate?” – the questions that EAs asks when trying to identify the highest-impact opportunities – the typical US-based climate (policy) expert is not an expert. This is so for at least three reasons:
1. The US-debate is focused on the US, the typical expert does not ask “what is the most effective thing one can do to have an effect on global decarbonization?” but rather “what seems best in the US?”. As per theory of change discussion, those two things aren’t necessarily related very much at all.
2. The overall climate debate in Western countries is quite biased – and predictably and importantly so. Compared to their real importance, there is an overemphasis on renewables, energy efficiency, and electric cars, there is an over-emphasis on lifestyle changes vs technological changes and there is an underappreciation of the technological challenge – but rather an emphasis on the political challenge and a typical framing that all that is missing is political will.
All of this makes sense (is predictable) if we think about where mainstream environmentalism comes from – ideologically speaking – from an egalitarian philosophy that prefers to solve problems with virtue rather than technology, that prefers “small is beautiful” and “in harmony with nature” over big centralized solutions (such as nuclear power or large dams), that has a moral need to demonize large centralized structures such as the state or capitalism, etc. Of course, not all experts are steeped in this ideology and not all experts that are follow it to the fullest extent. But it is notable that the overall climate debate is very biased by a particular worldview, an egalitarian / green view, that systematically hypes solutions that fit into the ideological priors – such as decentralized renewables and micro-grids and appealing to individual morality – and, equally systematically, disregards others – such as nuclear power build outs or CCS or a large-scale innovation agenda in line with the climate challenge.
Of course, a naïve narrative that solely focuses on technological innovation is equally biased, which is why I stressed so much in my original comment that I did not come to this topic with techno-optimist priors. All of those simplified ideological stories have kernels of truth to them and we need to judge the overall bias of the debate.
But if you look at the overall climate debate in Western countries it is hard to argue that the innovation arguments are getting their fair share – that innovation is emphasized enough compared to its importance (also see theory of change section).
It is not surprising that in this biased conversation Sunrise comes out on top – it serves all the ideological priors of those voices dominating the climate conversation – but that does not make it correct.
[Aside: Indeed, I would argue that because of the unfortunate situation that the innovation arguments tends to be made by those more that do not really want to act, the innovation agenda is neglected compared to its importance.
Climate activists downplay it because they like existing solutions as well as a framing of the problem as one of evil corporations and the need for fundamental societal change (rather than a severe technological challenge that needs strengthened innovation policy and a move away from “100% renewables”), whereas they often suspect – and often rightly so – that arguments for innovation on the right are made in bad faith, as discursive responses justifying inaction.
This leaves us in a terrible situation where we are severely underinvesting in technological innovation despite it being the only solution to get to net zero globally. This is one of the reasons why at FP we focus on funding innovation advocates that are on the center left but that are more serious about innovation than the green mainstream and that are able to work with people making innovation arguments on the right]
3. The typical climate (policy) expert is not an impact-focused philanthropist – indeed they are not philanthropists at all.
The best philanthropic options are those with the highest expected marginal returns to additional funding, something that a typical climate (policy) expert has very little insight into. Rather, the expert – when prompted with that question – will likely answer with something that seems good to them, that is top-of-mind, that fits with the ideological inclinations of the expert, etc. This is not data about high-impact philanthropy, this is data about the discourse.
What is this evidence for?
I am not sure that the fact that CATF does not come out on top of many of those experts is evidence of anything, but, if it is evidence of something, it seems weak evidence in support of CATF—if it were popular with everyone we should not be funding it!
Indeed, this points to the “suspicious convergence” point on TSM v CATF. The fact that TSM is so popular and that CATF is not should make us fairly skeptical, by itself, that we should fund TSM rather than CATF, and that TSM is the most effective thing to fund.
In comparatively non-neglected causes such as climate we should be quite wary of funding the popular things. Indeed, the entire theory of value of CATF and other charities that FP champions is that these are charities that improve societal resource allocation in a field that receives a lot of resources but where those are spent fairly inefficiently.
In other words, it is precisely because CATF is focused on inconvenient and neglected things – CCS, advanced nuclear, zero-carbon fuels, industrial decarbonization, etc. – that we should not expect them to be popular but very much worthy of support.
Who are the relevant experts for the question at hand?
There are two types of experts that seem most relevant for finding high-impact climate philanthropy options:
1. Aligned climate philanthropists
Among aligned climate philanthropists – those that want to have the maximum impact on global decarbonization solutions with their dollars and approach this topic rigorously – CATF and/or the approach it stands for has a lot of supporters.
Most prominently that would be Bill Gates who does not get tired to advocate for innovation and technology pluralism. But there are also many climate philanthropists at US and European foundations that are strong supporters of CATF for exactly the reasons that we support them (not sharing publicly here, but happy to share in PM). As such, there are many relevant experts that would support CATF as a top choice.
2. Experts on sub-topics that help inform the theory of change, identify effective interventions, etc.
Experts are of course essential to form a view of effective interventions in the climate space and of effective organizations pursuing these.
Indeed, as I mentioned, the FP research is informed by hundreds of conversations with experts.
But apart from experts on particular organizations or interventions, the way expert views flow into the research is mostly via getting a sense of the most effective interventions, the bottlenecks, the neglected areas, the overall theory of change. There is a lot of information and expert opinion out there supporting the choice of CATF more generally that does not mention or is even aware of those organizations at all which leads me to my next point, “what is the appropriate theory of change?”
2. What is the appropriate theory of change?
Global decarbonization, not US policy change
GG states that there are several theories of (political) change and this is undoubtedly true.
It is important to recognize, though, that the theories of change that GG focuses on – theories of change in the US political context – are not directly related at all to a theory of change that focuses on maximal global decarbonization impact.
While this would be fine if the goal would be to identify the best charities reducing emissions in the US, it becomes problematic when it is claimed based on GG analysis that we are entirely ignorant about the expected value of donations to CATF or TSM.
If we ask “what’s the most effective climate funding opportunity to support at the margin that we know of?”, theories of political change in the US only matter insofar as emissions reductions “travel” – in a causal sense—through US political change.
Given that US emissions are only ~15% of global emissions and that this share is declining as the rest of the world gets richer and the US is decarbonizing (even if on current trajectory), but that the US has roughly half of the world’s energy innovation budget, is generally considered the most capable innovation economy, and that the US is – for the moment, in any case – the leading global power, it stands to reason that the most effective things one can focus on in the US with regards to climate are interventions that maximize indirect effects and that might be quite different from those maximizing US emissions reductions (in expectation).
This is at the heart of the misunderstanding of CATF as “incrementalist”. CATF might be framed in that way if one is focused on US emissions and US political change. OK.
But the reason that most EAs support CATF as high-impact climate org is not because of the reduced emissions in the US but, precisely, because of the large international benefits that innovation policy targeted at neglected technologies brings.
All of the major success stories we have seen in climate over the past 20 years – solar, wind, coal > gas in the US, electric cars and batteries – have been the result of relatively narrow and targeted policies, the kind of which CATF advances for technologies that are less popular with greens for reasons of ideology, not merit.
Indeed, to make this more explicit, a lot of my skepticism about TSM at the margin comes from scenarios where TSM right now is relatively close to its maximum potential – lots of sway on Democrats, little sway on Republicans – but where binding climate legislation or more sweeping legislation still remains off the table and will remain so for relevant periods (getting a binding climate target in the US in 2035 is not that useful).
In that world, TSM will put lots of pressure on Biden and Democrats to advance executive orders (we can already observe that) that are ultimately unstable and open to court challenges rather than seizing the opportunity that presents itself to make much stronger innovation policy happen. This is a real danger and this is why I am very wary of stating certainty of positive impact in expectation of marginal TSM donations.
While one might be quite fatalistic about mandatory climate targets of the kind we have in Europe for the US, the response to that shouldn’t necessarily be – just push harder and harder and some day the Democrats will get rid of the filibuster or find a way to make executive orders robust and then we will win – but rather to make the most of the opportunity that the US presents for climate – an unparalleled energy innovation system that has a serious shot at breakthroughs in CCS, advanced nuclear, etc. Ambitious economy-wide climate legislation in the US is neither a necessary nor a sufficient condition for climate progress globally.
Climate is not, fundamentally, a problem like civil rights or the other analogies where a solution will need a strong legislation passing Congress and for which a large grassroots movement and even one larger than Sunrise right now might be essential.
CATF and similar charities look very good on a theory of change focused on the global picture
And CATF looks very good on a theory of change focused on maximized global decarbonization impact when taking into account some of the most important stylized facts about the climate challenge (widely recognized as median views in the respective expert communities):
1. Global energy demand will grow and restricting energy demand growth is very problematic from a humanitarian perspective.
2. Effective global decarbonization requires a much larger set of technologies than those currently available. Most of those technologies are not on track and many necessary technologies are in early stages.
3. Attention to many of those technologies is not on par with their importance, there is systematic neglect of key solutions.
4. More active US energy innovation is expected to be a very cost-efficient way to reduce emissions in the US and, crucially, this does not even include the global benefits.
You can then combine this with two CATF-specific features:
5. CATF is a strong organization that translates money into effective advocacy. This is not controversial within the EA community, something GG agrees on. It was first established in the FP 2018 report and it appears that at least 4 EA orgs had multiple calls with CATF, often dozens, that reaffirmed this conclusion (FP, Legacies Now, SoGive, Giving Green).
6. CATF has very productive funding margins, projects that are currently unfunded and that make a lot of sense from the above stylized facts and the theory of change.
This is all you need to come to CATF as a likely local optimum in effective climate philanthropy.
None of this is controversial and – indeed – each of the claims above about the world in general (1-4) follow directly from median expert views on those respective topics and the CATF-specific claims (5-6) are even entirely uncontroversial across the EA community.
In contrast, motivating TSM as a top-choice requires a lot of controversial claims, such as (a) that we are sure that the impact of marginal TSM donations is not negative in expectation and (b) that additional effort can lead to significant change beyond what is already baked in despite the approach of Sunrise being partisan and thereby, quite plausibly, limited in its ultimate potential given the structure of the Senate and the Electoral College.
What is more, as Robin pointed out, the fact that CATF operates not only in the US is another source of believing in it to be higher impact—CATF can optimize marginal resources across many jurisdictions, many of which are often more fruitful than the US context, which is very fruitful now but might as well dry up significantly.
3. What can we know through charity evaluation methodology?
A lot of the arguments for GG’s skepticism about being able to figure out the goodness of CATF v TSM comes from non-marginal arguments, from the stated inherent incompatibility between different approaches.
While I do disagree with these incomparability claims—see my original comment—even if one agrees with these there is a lot of traction from the methods that “EA-style” (for lack of a better word) charity evaluation provides.
Neglectedness
Neglectedness is a proxy, but it is a very useful one. In a world with millions of funding margins which we can never all evaluate in detail, neglectedness gives crucial information (as a prior, also see my long comment).
If a field is flooded with attention and money, like progressive climate activism, then the probability that there are great things to be funded is low for three fundamental reasons:
Declining marginal returns: A strategic actor will order projects such that the lowest value ones are at the margin and the further we are down that margin, the lower we should expect the value to be (ceteris paribus, which is key, e.g. CATF’s expansion into other geographies, changing its funding margins is an argument in another direction).
Declining probability of true additionality: The more money and attention goes into a particular direction or org the less likely it is that additional funding there will be truly additional—it is more likely that actors “satisfice” and, when the funding environment is fertile, additional dollars will not always be additional, as there might be funding targets or, at the very least, a declining eagerness to fundraise more. That’s the point about 10,000 volunteers and inspiring the public imagination of an entire wing of the dominant American party—if in that situation you are seriously funding-constrained, this would be quite surprising.
3.** Declining probability of money being the impact-constraining resource:** The more money you have and the easier it is to get more money if you need it, the lower the probability that money really is your constraint on higher marginal impact.
While I won’t name names here, there are many great climate charities which I believe are doing incredibly important and good work, which we are not recommending because of such considerations—this is not something that only applies to TSM.
Aggregating expert views
When we have an issue where relevant experts disagree strongly—such as whether a more partisan approach to climate policy is good or bad—epistemic humility pushes us closer to a lower value (zero).
This is where a lot of the TSM skepticism comes from.
Not from saying “we are sure that TSM is bad” but rather “when observing the debate, we notice that this is an area of severe disagreement between reasonable people”—e.g. look at all the debates in the Democratic party between progressives and moderates on which path is more promising to ensure Democratic priorities are implemented (and the Democratic majority survives).
So, this would be an area where we would either need to set the marginal impact of TSM downwards or where more research would be needed to demonstrate that the moderate wing of the Democratic party is entirely wrong (or, more climate-specifically, that more centrist voices are unduly worried about a progressive over-reach that will backfire and lead to deadlock and no progress).
This argument is largely missing in the TSM analysis, instead it feels like an assessment that is based on the most positive expert interpretations of TSM, not the median ones, and not taking into account voices that would be critical of TSM’s impact.
4. Length and depth of engagement that led to CATF and other climate high-impact recs
This is about the “intellectual history” leading to the CATF recommendation, with the TL, DR that the recommendation is based on a lot deeper engagement than the GG comment is giving credit for.
The argument on which CATF comes out as the best recommendation we currently know of goes back to 2016 -- when the argument that EAs interested in climate should focus on (a) innovation for (b) neglected technology, (c) given a world of rising energy demand and real conflicts between climate and poverty alleviation, was first publicly made (to my knowledge). You can watch it here and, somewhat later, but more refined, here. There’s a write-up here (though I never quite finished it because I focused on embedding this argument in the EA debate, which was achieved with John’s and Hauke’s reports, leaving me to prioritize other things).
This argument was already informed by deep engagement with climate for about a decade, reading and talking to hundreds of experts and, arguably—in the process—becoming an expert at being better able to evaluate the strength of different claims as well.
These arguments then found their ways into both John’s 2018 report and Hauke’s Let’s Fund report, both of whom vetted this argument against competing arguments. I don’t know the details about Hauke’s report, but I know that John talked to at least 50 experts for the first report and I know—from, sometimes painful, experience :) -- that John does not accept any argument at face-value before he has not engaged with it himself. Crucially, both Hauke and John made charity recommendations that turned this argument into something actionable and thereby unlocking massive value.
After I joined FP in late 2019, John and I spent a lot of time revisiting our recs (which also led to not continuing some recs), working on a new report that more clearly motivates our theory of change, and identifying new recs.
The robustness of the CATF recommendation is one of the few things that stood the test of time, indeed by more clearly articulating our theory of change—improving societal resource allocation in light of decarbonization priorities, towards innovation in neglected technologies—we became somewhat more convinced of CATF being a “local optimum” (obviously, there might be other great options—I for one would never claim that we will never find something better than CATF, just that until now we haven’t).
This is also true for the EA community at large: while there have been many critical reviews of FP’s work on climate (e.g. concerns about CfRN), the CATF recommendation has stood the test of time and examination.
Thank you for great comment—I agree with a lot of this and a lot of the criticism in this thread.
Some more related general thoughts:
I think the marginal vs. average impact is one of the most frequently misunderstood aspects of EA and should be highlighted even more.
The value of grassroots advocacy (vs. grass tops / top down policy advocacy) is an under-explored research topic.
Kudos, Dan & team, for this reply!
I will need a bit of time for a full reply, but I wanted to let you and team know that I really appreciate the thoughtful, gracious, and civil reply.
While we do have many disagreements on epistemics and empirics—and I think I also still deeply disagree with many things in the post above (to be explained) -- we are united in the same purpose, making the world better the best we can.
So the point of this comment is just to recognize that and thank you.